Proceedings
The
Thirteenth
Annual
Waste
Testing
&
Quality
Assurance
Symposium
(
WTQA
'
97)

July
6­
9,
1997
Crystal
Gateway
Marriott
Arlington,
VA
CONTENTS
SPECIAL
SESSION:
ENDOCRINE
DISRUPTERS
24
Developing
A
Method
Used
to
Screen
for
More
Than
400
Pesticides
and
Endocrine
Disrupters.
P.
Wylie,
B.
Quimby
6
16
Reference
Materials
for
Endocrine
Disrupting
Compound
Analysis:
An
Overview.
M.
Re
5
11
Status
of
EPA
Laboratory
Methods
for
Measuring
Endocrine
Disrupters.
J.
Parr,
M.
Bruce,
P.
Winkler
4
11
Multimedia
Analytical
Approaches
to
Monitoring
and
Measuring
Suspect
Endocrine
Disrupting
Compounds.
T.
Jones,
J.
Bumgarner,
D.
Vallero
3
10
CDC's
Analytical
Approach.
L.
Needham
2
3
ACS
Overview
of
the
Analytical
Challenge.
L.
Keith
1
Page
Number
Paper
Number
SPECIAL
SESSION:
ADVANCED
ENVIRONMENTAL
MONITORING
RESEARCH
46
Direct
Monitoring
of
Environmental
Pollutants.
O.
Sadik
12
46
Electrodialytic
NaOH
Eluent
Production
and
Gradient
Generation.
P.
Dasgupta
11
41
Real­
Time
Trace
Dectection
of
Elemental
Mercury
and
Its
Compounds.
R.
Barat
10
41
Solid­
Phase
Microextraction
Coupled
with
Infrared
Spectroscopy
for
the
Determination
of
Organic
Pollutants
in
Water.
D.
Tilotta
9
40
Field
Determination
of
Organics
for
Soil
and
Sludge
Using
Subcritical
Water
Extraction
Coupled
with
Solid­
Phase
Extraction.
S.
Hawthorne
8
35
EPA's
Extramural
Monitoring
Research
Program.
W.
Stelz
7
Page
Number
Paper
Number
INORGANIC
60
Long­
Term
Stability
of
ICP
Spectra
Registration
by
Management
of
the
Models:
Application
to
Quantitation
Using
Multivariate
Analysis.
C.
Hanna,
A.
Ganz
18
55
EPA
Methods
3015A
and
3051A:
Validation
Studies
for
Updated
Microwave
Leach
Methods.
D.
Link,
P.
Walter,
H.
Kingston
17
54
Legally
Defensible
Speciated
Measurements
Using
SIDMS.
H.
Kingston,
D.
Huo,
Y.
Lu,
P.
Walter
16
53
Field
and
Laboratory
Analysis
of
Mercury.
P.
Walter,
H.
Kingston,
H.
Boylan,
Y.
Han
15
49
Flame
Atomic
Absorption
Spectrophotometry
for
the
Determination
of
Arsenic
and
Selenium
in
TCLP
Extracts.
Z.
Grosser
14
49
Practical
Clean
Chemistry
Techniques
for
Trace
and
Ultratrace
Elemental
Analysis.
H.
Kingston,
P.
Walter
13
Page
Number
Paper
Number
iii
ORGANIC
132
On­
site
Analysis
of
Explosives
in
Soil:
Evaluation
of
Thin­
Layer
Chromatography
for
Confirmation
of
Analyte
Identity.
M.
Stutz,
S.
Nam,
D.
Leggett,
T.
Jenkins
42
131
Fast
Prescreening
of
Water
and
Soil
Samples
Using
Solid­
Phase
Microextraction
(
SPME).
T.
Schumacher
41
131
Proposed
U.
S.
EPA
Method
8320:
A
Risk
Assessment
Method
for
Secondary
Explosives.
W.
Murray,
Y.
Tondeaur,
C.
Enterline,
H.
Gruelich,
J.
Roach
40
125
Comprehensive,
Quantitative,
Congener­
Specific
PCB
Analysis:
When
Is
It
Required
and
What
Is
Necessary
To
Achieve
It?
G.
Frame
39
113
Determination
of
Nitroaromatic,
Nitramine,
and
Nitrate
Ester
Explosives
in
Water
Using
Solid­
Phase
Extraction
and
GC­
ECD.
M.
Walsh,
T.
Ranney
38
106
Accidental
Chemistry.
M.
Bruce,
R.
Risden,
K.
Richards,
R.
Ruyechan,
P.
Winkler
37
105
A
Comparison
of
Microwave
Extraction
Solvent
Sytsems:
Nonpolar
versus
Nonpolar/
Polar,
General
Differences
from
Soil
Samples.
P.
Walter,
G.
Lusnak,
H.
Kingston
36
104
A
Comparison
of
ASE
with
Soxhlet,
SFE
and
Sonication
for
the
Extraction
of
Explosives
from
Contaminated
Soils.
B.
Richter,
J.
Ezzell,
F.
Hoefler
35
104
Automated
Small­
Volume
Extraction
of
Semivolatiles
Followed
by
Large­
Volume
GC/
MS
Injection.
F.
Feyerherm,
R.
McMillin,
D.
Gregg,
M.
Daggett
34
98
Estimating
the
Total
Concentration
of
Volatile
Organic
Compounds
in
Soil
Samples.
A.
Hewitt,
N.
Lukash
33
98
Summary
of
Stability
of
Volatile
Organics
in
Environmental
Soil
Samples.
D.
Bottrell
32
98
Overview
of
RCRA
Organic
Methods
Program.
B.
Lesnik
31
97
Full
Evaluation
of
a
Microwave­
Assisted
Process
(
MAP
 
)
Method
for
the
Extraction
of
Contaminants
Under
Closed­
Vessels
Conditions.
B.
Lesnik,
J.
Paré,
J.
Bélanger,
R.
Turpin,
R.
Singhvi,
C.
Chiu,
R.
Turle
30
96
Congener­
Specific
PCB
GC
Analysis:
A
Fundamental
Approach.
D.
Gere
29
96
SPME
Preparative
Applications
in
Analysis
of
Organics
in
Radioactive
Waste.
J.
Young
28
85
Improved
Extraction
Efficiency
of
Polychlorinated
Biphenyls
from
Contaminated
Soil
Using
a
Total
Halogen
Screening
Method.
W.
Schutt­
Young,
A.
Lynn,
T.
Lynn,
M.
Krumenacher
27
82
Determination
of
Oil
Contaminated
Soils
and
Sludges.
A.
Majid,
B.
Sparks
26
81
A
Field­
Useable
Method
for
Toxicity
Screening
of
Waste
Streams
Generated
During
Destruction
of
Chemical
Warfare
Agents.
K.
Morrissey,
T.
Connell,
H.
Dupont
Durst
25
74
Determination
of
Chlorinated
Hydrocarbon
Concentrations
in
Soil
Using
a
Total
Organic
Halogen
Method.
T.
Lynn,
J.
Kneece,
B.
Meyer,
A.
Lynn
24
67
The
Development
of
an
Ion
Chromatography
Method
to
Monitor
Organic
and
Inorganic
Indicators
of
Intrinsic
Bioremediation
at
Hazardous
Waste
Sites.
R.
Rediske,
A.
Stiop,
D.
Van
Dyke,
P.
Durda
23
66
Easier
and
Faster
GC/
ECD
Analyses
of
Pesticides
and
PCBs.
S.
Brillante
22
66
A
New
Pulsed
Flame
Photometric
Detector
for
the
Analysis
of
Pesticieds.
R.
Snelling
21
65
Solid­
Phase
Extraction
Applications
in
the
Sampling
of
Organic
Components
in
Radioactive
Wastes.
S.
Crump,
J.
Young,
D.
Hobbs,
M.
Jamison
20
65
Solid­
Phase
Microextraction
Preparative
Applications
in
the
Analysis
of
Organic
Components
in
Radioactive
Wastes.
J.
Young,
S.
Crump
19
Page
Number
Paper
Number
iv
ENDOCRINE
DISRUPTORS
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
1
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
2
ENVIRONMENTAL
ENDOCRINE
DISRUPTORS:
AN
ACS
OVERVIEW
OF
THE
ANALYTICAL
CHALLENGE
L.
H.
Keith
Radian
International
LLC,
P.
O.
Box
201088,
Austin,
TX
78720­
1088
INTRODUCTION
Environmental
Endocrine
Disruptors
(
EEDs),
endocrine
modifying
chemicals
which
are
also
environmental
pollutants,
are
the
subject
of
a
special
session
at
the
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
sponsored
by
the
U.
S.
Environmental
Protection
Agency
(
EPA)
and
the
American
Chemical
Society
(
ACS)
Division
of
Environmental
Chemistry.
While
most
recent
technical
symposia
on
this
topic
have,
and
continue,
to
focus
on
the
effects
of
EEDs,
the
focus
of
this
special
session
is
on
the
analysis
of
EEDs.
The
analysis
of
EEDs
in
environmental
matrices
(
e.
g.,
water,
air,
soil,
wastes,
and
biota
including
both
plants
and
animals)
is
critical
to
future
regulatory
monitoring
of
them
as
well
as
to
an
understanding
of
their
occurrence,
transport
and
migration
within
the
environment,
and
their
ultimate
degradation.

In
order
to
be
able
to
analyze
for
EEDs
several
prerequisites
are
necessary:

The
EEDs
subject
to
monitoring
and/
or
regulation
need
to
be
identified*;

2.
Methods
for
EEDs
need
to
be
effective
at
concentration
levels
desired
for
monitoring**;
and
3.
Analytical
reference
materials
must
be
available
to
calibrate
the
instruments
used
for
the
analyses.***

DEFINITION
OF
ENDOCRINE
DISRUPTING
CHEMICALS
The
endocrine
system
refers
to
the
complex
system
that
involves
the
brain
and
associated
organs
and
tissues
of
the
body.
These
include
the
pituitary,
thyroid,
and
adrenal
glands
and
the
male
and
female
reproductive
systems,
all
of
which
release
hormones
into
the
bloodstream.
In
particular
the
sex
hormones
include
estrogens
in
females
and
androgens
in
males.
Endocrine
disrupting
chemicals
(
EDCs)
consist
of
synthetic
and
naturally
occurring
chemicals
that
affect
the
balance
of
normal
hormonal
functions
in
animals.
Depending
on
their
activity
they
may
be
characterized
as
estrogen
modulators
or
androgen
modulators.
They
may
mimic
the
sex
hormones
estrogen
or
androgen
(
thereby
producing
similar
responses
to
them)
or
they
may
block
the
activities
of
estrogen
or
androgen.
(
i.
e.,
be
anti­
estrogens
or
anti­
androgens).
1
There
are
three
categorical
sources
of
EDCs:
2
1.
Pharmaceuticals
­
One
of
the
first
recognized
synthetic
EDCs
was
diethylstilbestrol
(
DES),
a
pharmaceutical
product
given
to
pregnant
women
from
1948
to
1972
to
help
prevent
miscarriages.
It
caused
clear­
cell
carcinoma
in
the
vagina,
reproductive
abnormalities
in
female
offspring,
and
a
much
higher
than
normal
rate
of
genital
defects
in
male
babies.
___________________________________________________

*
Developing
screening
methods
is
an
important
effort
that
hopes
to
bypass
initial
needs
to
identify
specific
individual
EEDs.
However,
as
discussed
below,
screening
methods
for
EEDs
will
have
to
accommodate
a
wider
variety
of
diverse
chemicals
than
have
ever
been
subjected
to
screening
methods
before.
This
is
an
extremely
complex
and
difficult
challenge.

**
Concentration
levels
at
which
EEDs
are
presumed
to
be
effective,
and
therefore
potentially
monitored,
are
at
those
of
hormones.
These
are
typically
a
thousand
or
more
times
below
levels
at
which
most
current
analytical
methods
are
able
to
measure
analytes
in
the
environment
with
reasonable
rates
(
e.
g.,
<
10%
at
a
95%
confidence
level)
of
false
positive
and
false
negative
identifications.
Thus,
EED
levels
of
concern
may
be
at
parts
per
trillion
(
10­
12)
or
below.

***
Analytical
reference
materials
with
certifications
of
identity,
purity,
and
homogeneity
are
necessary
for
both
qualitative
and
quantitative
analysis.
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
3
2.
Naturally
Occurring
EDCs
­
This
source
of
EDCs,
collectively
called
"
phyto­
estrogens,"
includes
foods
such
as
soybeans,
apples,
cherries,
wheat,
and
peas.

3.
Environmental
EDCs
­
The
third
group
of
EDCs
are
some
environmental
pollutants.
These
environmental
endocrine
disruptors
(
EEDs)
are
the
subject
of
this
symposium.

ANALYTICAL
CHALLENGES
The
analytical
challenges
can
be
summarized
as:

1.
Determining
which
analyses
are
EEDs
so
that
their
effects
can
be
studied
and
those
of
importance
can
be
monitored
and/
or
regulated
in
the
environment
and
in
food.

2.
Developing
screening
analytical
methods
that
will
accommodate
a
wide
variety
of
analytical
functional
groups
at
extremely
low
detection
levels.

3.
Developing
qualitative
and
quantitative
analytical
methods
to
confirm,
as
necessary,
the
identity
and
quantity
of
EEDs
in
the
environment
and
in
food.

4.
Assuring,
through
appropriate
use
of
analytical
reference
materials
and
QA/
QC
procedures,
an
acceptable
level
of
false
positive
and
false
negative
determinations
at
regulatory
levels.

Analytical
Challenge
#
1
is
the
subject
of
this
paper.
The
other
authors
in
this
symposium
will
address
the
other
three
analytical
challenges.

CHALLENGE
#
1
­
DETERMINING
WHICH
POLLUTANTS
ARE
ENVIRONMENTAL
ENDOCRINE
DISRUPTORS
One
of
the
major
problems
we
face
is
determining
which
chemicals
in
the
environment
should
be
labeled
as
environmental
endocrine
disruptors.
This
is
a
critical
question
because
these
materials
will
be
the
subject
of
future
regulations.
Obtaining
information
on
them
will
also
require
very
significant
expenditures
of
time
and
money.
In
addition,
traditional
environmental
analytical
methods
require
specific
known
analytes
for
identification
and
quantification.
If
you
don't
know
what
chemical(
s)
to
analyze
for
then
you
can't
analyze
for
them
­­
a
trivial
statement
but
a
difficult
problem
with
EEDs.
Thus,
identifying
which
chemicals
are
EEDs
is
the
first
of
four
analytical
challenges.
Currently,
the
process
for
deciding
if
a
chemical
is
an
environmental
endocrine
disruptor
is
to
determine
the
effects
of
that
chemical
on
the
endocrine
systems
of
humans
and
other
animals.

Much
of
the
research
on
EEDs
to
date
has
focused
on
the
effects
of
EEDs
on
wildlife.
There
are
many
studies
involving
certain
pesticides,
polychlorobiphenyls
(
PCBs)
and
polychlorodibenzo­
p­
dioxins
(
e.
g.,
2,3,7,8­
TCDD)
that
link
them
to
birth
defects
and
aberrant
sexual
behavior.
Laboratory
tests
have
also
produced
genital
defects,
reduced
testicular
weights
and
low
sperm
counts
in
rats
fed
with
DDE,
PCBs,
Vinclozolin,
and
2,3,7,8­
TCDD.
3
It
is
also
believed
by
some
researchers
that
EEDs
may
be
the
cause
of
similar
types
of
recent
observations
in
humans.
The
CDC
in
Atlanta
has
performed
surveys
that
show
that
the
average
US
resident
has
hundreds
of
chemicals
accumulated
in
their
fat
tissues
including
polychlorodibenzo­
p­
dioxin
and
polychlorodibenzofuran
isomers
("
dioxins"
and
"
furans").
4
Some
chemicals,
including
many
EEDs,
can
"
bioaccumulate"
or
build
up
in
animals.
Once
they
are
incorporated
into
the
tissues
and
fat
of
animals
and
humans,
they
can
remain
there
for
long
periods
of
time
until
they
are
ultimately
metabolized.
Thus,
an
embryo,
the
most
sensitive
stage
of
life,
can
be
damaged
by
chemicals
the
mother
was
exposed
to
weeks
or
years
earlier.
5
Another
factor
that
is
important
with
EEDs
is
that
timing
of
exposure
to
them
can
be
critical.
In
fact,
timing
of
exposure
may
be
more
important
than
the
dose
or
concentration
of
their
exposure.
A
single
exposure
at
a
vulnerable
moment
for
a
developing
embryo
has
the
potential
to
cause
damage
and,
of
course,
long
term
exposure
to
relatively
small
amounts
of
them
also
could
cause
damage.
5
It
has
also
been
discovered
that
some
combinations
of
two
or
three
EEDs
can
be
many
times
more
potent
than
any
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
4
one
of
them
by
themselves.
Thus,
in
addition
to
the
timing
of
exposure,
it
appears
that
the
combinations
of
EEDs
that
one
is
exposed
to
can
cause
a
synergistic
effect
that
can
magnify
the
damage
they
can
cause.
These
are
all
unresolved
issues
that
will
have
to
be
worked
out
with
time
and
additional
research.
1
Lastly,
there
is
the
problem
of
the
naturally­
occurring
EDCs.
How
does
the
exposure
of
EEDs
relate
to
exposure
to
naturally­
occurring
EDCs?
How
do
humans
and
animals
metabolize
or
neutralize
the
effects
of
each
categorical
source
of
EDCs?
These
effects
are
briefly
described
later.

Table
1
reflects
the
enormity
of
the
first
analytical
challenge:
deciding
which
environmental
pollutants
are
EDCs.
There
are
at
least
103
suspect
EEDs
identified
to
date
by
various
organizations;
no
doubt
this
will
change.
The
EPA
lists
60
suspect
EEDs,
25
of
which
are
targeted
by
the
EPA
NERL
Endocrine
Disruptor
Exposure
Team
for
multi­
media
environmental
analysis
this
year.
6
The
Centers
for
Disease
Control
&
Prevention
(
CDC)
in
Atlanta,
GA
has
identified
48
suspect
EEDs
that
are
of
interest
to
that
agency.
7
Finally,
the
World
Wildlife
Fund
Canada
(
WWF)
has
expanded
on
the
approximately
50
suspect
EEDs
listed
in
Our
Stolen
Future8
and
now
lists
68
suspect
EEDs.
4
Many
of
these
chemicals
are
on
all
three
lists,
some
are
on
only
two
of
the
three
lists,
and
others
are
on
only
one
of
the
three
lists.
In
addition,
some
are
listed
by
other
known
names
or
synonyms
so
it
is
very
important
to
characterize
any
list
of
these
chemicals
by
their
(
almost)
unique
Chemical
Abstract
Service
(
CAS)
numbers.
Even
using
CAS
numbers
two
of
the
chemicals
were
discovered
to
have
two
different
CAS
numbers
and
"
parathion"
was
interpreted
by
this
author
to
be
the
ethyl
ester
rather
than
the
methyl
ester.

USAGE
CLASSIFICATIONS
OF
EEDS
One
of
the
characteristics
of
EEDs
mentioned
in
the
introduction
is
the
wide
variability
of
their
chemical
class
characteristics.
This
is
also
reflected
in
a
classification
of
their
uses.
Some
suspected
EEDs
are
various
types
of
pesticides,
others
are
common
metals,
and
many
fall
into
the
classification
of
useful
organic
industrial
chemicals.
The
net
result
provides
a
real
challenge
for
developing
screening
methods.

There
are
at
least
nine
different
usage
classifications
of
ECDs
and
these
are
listed
below
and
also
in
Table
1.

1.
Biocides.
2.
Insecticides,
3.
Herbicides,
4.
Nematocides,
5.
Fungicides,
6.
Industrial
Chemicals
(
e.
g.,
solvents,
plasticizers,
etc.),
7.
Metals
8.
PCBs
(
i.
e.,
Specific
PCB
isomers),
and
9.
No
Commercial
Use
(
i.
e.,
compounds
that
are
a
degradation
product
or
impurity
of
other
chemicals).

EFFECTS
OF
EEDS
Earlier
it
was
mentioned
that
the
determination
of
which
pollutants
are
environmental
endocrine
disruptors
depends
on
their
effects
on
the
endocrine
systems.
However,
the
exact
effect
of
hormone
exposure,
both
natural
and
unnatural,
is
greatly
dependent
on
factors
such
as
species,
age,
and
gender.

Generally,
the
offspring
of
exposed
adults
are
the
most
vulnerable
to
these
effects.
Fetuses
and
newborns
are
especially
susceptible
to
environmental
contaminants.
In
addition
to
regulating
sexual
differentiation
during
fetal
development,
sex
hormones
play
a
role
in
the
organization
of
specific
areas
of
the
brain.
Less
is
known
about
this
action,
but
studies
have
shown
a
correlation
between
levels
of
estrogen
and
brain
morphology,
as
well
as
with
sexual
behavior
in
male
rats
and
mice.
The
brain
and
central
nervous
system
continue
development
throughout
the
fetal
stage
and
early
natal
period
making
them
particularly
susceptible
to
chemical
exposure.
5
EEDs
affect
humans
as
well
as
wildlife.
There
are
many
documented
effects
on
humans
and
even
more
suspected
effects.
These
include
decrease
in
male
fertility,
defects
in
male
sexual
development,
increases
in
prostate
cancer,
female
reproductive
problems,
increases
in
breast
cancer,
endometriosis,
immune
system
damage,
increased
incidence
of
goiters,
and
behavioral
and
developmental
problems
in
children.
Some
examples
of
these
are
described
below.
1
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
5
Male
Fertility
The
most
likely
effect
of
endocrine
disruption
in
men
may
be
a
reduction
in
sperm
production
and
also
in
the
sperm's
ability
to
fertilize
an
egg.
In
normal
human
males,
the
number
of
sperm
produced
per
ejaculate
is
normally
close
to
the
level
required
for
fertility.
Thus,
even
a
small
reduction
in
daily
sperm
production
can
lead
to
infertility.
5
Sperm
production
by
the
average
man
in
western
countries,
including
the
U.
S.,
today
is
reported
by
some
to
be
half
of
what
it
was
in
1940.
One
report
indicates
that
average
sperm
count
has
declined
42%
and
average
volume
of
semen
diminished
by
20%.
9,5
Another
report
showed
an
increase
in
infertility
in
the
last
twenty
years
concluding
that
one
in
twenty
men
are
either
subfertile
or
infertile.
10,5
However,
these
may
be
an
over
simplification
and
later
reports
have
questioned
these
kind
of
conclusions.
1
Male
Sexual
Development
Defects
and
Cancer
There
appears
to
be
an
increase
in
sexual
development
defects
in
recent
years
and
some
of
these
may
be
linked
to
exposure
to
EEDs.
For
example,
more
baby
boys
have
to
undergo
operations
to
correct
undescended
testicles
("
cryptorchidism")
now
than
30
years
ago;
the
rate
appears
to
have
increased
2­
to
3­
fold
during
the
past
30
years.
A
birth
defect
called
"
hypospadias,"
in
which
the
male
urinary
canal
is
open
on
the
underside
of
the
penis,
also
is
increasing.
"
Inter­
sex"
features
in
baby
boys,
where
the
penis
is
covered
with
a
layer
of
fat
and
genitals
have
a
cleft
resembling
female
features,
also
appear
to
be
increasing.
In
some
cases
where
pregnant
mothers
were
exposed
to
very
high
levels
of
toxic
chemicals,
the
mothers'
boys
have
shorter
than
normal
penises,
similar
to
Lake
Apopka's
alligators
in
Florida.
Boys
born
to
women
who
were
exposed
to
PCB­
poisoned
rice
bran
cooking
oil
in
1978­
79
in
central
Taiwan,
the
so­
called
"
Yucheng"
boys,
were
found
to
have
significantly
shorter
penis
lengths
at
ages
11
to
14.11,5
Studies
in
some
industrialized
western
nations
show
that
cancer
of
the
testicles,
relatively
more
common
in
young
men
than
older
men,
has
increased
at
least
3­
fold
in
the
past
30
years.
Another
possible
effect
of
exposure
to
estrogen­
like
contaminants
is
prostate
enlargement
in
older
men.
This
condition
affects
80%
of
men
70
years
and
older.
The
exact
cause
of
prostate
enlargement,
however,
is
often
unknown.
Prostate
cancer
in
men
also
has
increased
by
80%
in
the
last
20
years.
12,5
Female
Reproductive
Effects
Women
normally
are
exposed
to
estrogen,
but
the
effects
of
EEDs
on
females
are
more
difficult
to
track
due
to
the
estrous
cycle
and
the
resulting
huge
differences
in
circulating
hormone
concentrations
at
different
stages
of
the
cycle.
The
presence
of
estrogen
mimicking
compounds
in
adult
women
can
impair
reproductive
capacity
by
interfering
with
natural
hormone
cycles,
potentially
rendering
women
unable
to
conceive
or
to
maintain
pregnancy.
5
Female
Breast
Cancer
Breast
cancer
may
also
have
links
to
the
estrogenic
contaminants.
5
Women
in
the
US
and
Canada
who
live
to
age
85
have
a
one
in
nine
risk
of
contracting
breast
cancer
in
their
lifetime,
double
the
risk
in
1940.13
Furthermore,
breast
cancer
mortality
since
the
1940s
has
increased
by
1%
per
year.
14
Two
leading
theories
of
the
primary
risk
factors
for
breast
cancer
are
exposure
to
estrogen
and
high
fat
diets.
5
It
also
may
be
possible
that
some
chemicals
are
promoters
or
inducers
of
cancer
rather
than
being
direct
carcinogens.
This
theory
is
supported
by
the
findings
that
some
EEDs
have
estrogenic
properties,
and
that
estrogen
is
known
to
promote
abnormal
cell
growth.
If
estrogen
exposure
after
maturation
plays
a
role
in
the
full
expression
of
early
developmental
changes
then
this
could
provide
an
explanation
for
both
the
increased
risk
of
breast
cancer
to
women
exposed
to
estrogens
in
utero
and
the
rare
cancers
initiated
at
maturation
in
the
women
whose
mothers
took
DES.
5
Endometriosis
Recent
animal
studies
strongly
suggest
that
human
exposure
to
dioxin
may
be
linked
to
endometriosis,
a
painful
disease
currently
affecting
10%
of
reproductive­
age
women.
Endometriosis
causes
bits
of
uterine
lining
to
migrate
generally
to
other
pelvic
organs
and
can
cause
infertility,
internal
bleeding
and
other
serious
problems.
The
disease
appears
to
becoming
more
common
and
afflicting
women
at
younger
ages.
5
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
6
Immune
System
Damages
Associations
between
endocrine­
disrupting
pollutants
and
immune
system
damages
in
wildlife
are
well
established.
15
Similar
associations
are
being
discovered
in
humans.
5
Goiters
Another
effect
of
endocrine
disruption
in
both
adult
males
and
females
may
be
thyroid
gland
enlargement,
more
commonly
known
as
goiter.
The
thyroid
gland
controls
growth
hormones
and
the
hormones
that
regulate
metabolism,
and
enlargement
of
the
thyroid
gland
can
disrupt
metabolism.
EEDs
that
have
been
implicated
in
a
syndrome
known
as
the
"
wasting
syndrome"
include
PCBs,
dioxin,
DDT,
toxaphene
and
lead.
5
Hyperactivity,
Learning,
and
Attention
Problems
With
Children
Recent
studies
found
a
dose­
response
relationship
between
the
quantity
of
contaminated
Great
Lakes
fish
consumed
by
the
mother
and
such
measures
in
newborn
infants
as
abnormally
weak
reflexes,
reduced
responsiveness,
motor
coordination
and
muscle
tone.
16,17,5
Further
studies
indicated
that
more
highly
exposed
children
had
slower
reaction
times
to
visual
stimuli,
made
more
errors
on
a
memory
test
and
took
longer
to
solve
problems.
Hyperactivity
and
learning
deficits
are
among
the
likely
effects
in
children
exposed
in
utero
to
endocrine­
disrupting
chemicals,
based
on
many
related
studies.
If
only
a
small
part
of
the
learning
and
behavioral
problems
of
children
can
be
attributed
to
endocrine,
immune,
or
nervous
system
damages
caused
by
maternal
or
childhood
exposure
to
EEDs,
the
implications
are
profound.
5
SUMMARY
In
summary,
the
analytical
challenges
are
complex
and
depend
first
on
determining
which
pollutants
are
to
be
labeled
as
environmental
endocrine
disruptors
and
thus
studied,
analyzed,
and
perhaps
regulated.
Currently,
the
only
means
of
determining
which
pollutants
are
EEDs
is
from
observing
their
effects
on
the
endocrine
systems
of
humans
and
other
animals
but
that
also
has
a
separate
set
of
complex
factors
that
affect
those
decisions
(
this
is
where
most
research
is
currently
focused
but
it
is
only
the
first
stage
of
the
process).
Once
chemicals
are
selected
for
analysis
as
EEDs
then
the
problems
of
developing
appropriate
screening
and/
or
individual
chemical
confirmatory
analyses
must
be
solved.
Methods
must
be
developed
and
validated
for
identification
and
quantification
of
EEDs
at
concentration
levels
a
thousand
times
lower
than
most
environmental
methods
can
currently
function.
At
those
levels
there
will
be
more
interferences
and
thus
greater
possibilities
for
false
positive
and
false
negative
conclusions
from
the
data.
Thus,
programs
that
incorporate
appropriate
QA/
QC
data
and
reliable
analytical
reference
materials
for
both
qualitative
(
identification)
and
quantitative
analysis
will
be
critical
in
order
to
avoid
basing
decisions
and
regulations
on
data
of
unknown
or
unreliable
quality.
EPA's
Data
Quality
Objective
(
DQO)
process
and
associated
data
quality
assessments
will
be
important
tools
to
facilitate
the
measurement
and
use
of
reliable
analytical
data
as
environmental
chemists
cross
a
new
threshold
of
analytical
challenges.
It
won't
be
easy.

REFERENCES
1.
Keith,
L.
H.,
Environmental
Endocrine
Disruptors
­
A
Handbook
of
Properties,
John
Wiley
&
Sons,
Inc.,
In
Press,
1997.
2.
Keith,
L.
H.,
Environmental
Endocrine
Disruptors
­
A
New
Analytical
Challange,
EnvirofACS,
Vol.
44,
No.
2,
p.
5,
December
1996.
(
ACS
Division
of
Environmental
Chemistry
Newsletter,
c/
o
L.
H.
Keith,
Radian
International
LLC,
P
O
Box
201088,
Austin,
TX
78720­
1088).
3.
Hileman,
Bette,
Chemical
&
Engineering
News,
p.
28,
May
13,
1996.
4.
World
Wildlife
Fund
Canada
[
Online]
Available
http://
www.
wwfcanada.
org/
hormone­
disruptors/,
December
15,
1996.
5.
Schmidt,
Wayne
A.,
"
Hormone
Copycats"
[
Online]
Available
http://
www.
greatlakes.
nwf.
org:
80/
toxics/
hcc1­
int.
htm,
Feb.
11,
1997.
6.
Jones,
Tammy,
U.
S.
EPA,
Las
Vegas,
NV,
Private
Communication,
December
12,
1996.
7.
Needham,
Larry,
National
Center
for
Environmental
Health,
Centers
for
Disease
Control
&
Prevention,
Public
Health
Service,
DHHS,
Atlanta,
GA,
Private
Communication,
August
25,
1996.
8.
Colborn,
T.,
D.
Dumanoski,
and
Meyers,
J.
P.,
Our
Stolen
Future,
Penguin
Books,
New
York,
NY,
1996.
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
7
9.
Carlsen,
E.
et
al.
"
Evidence
for
Decreasing
Quality
of
Semen
During
Past
50
Years,"
British
Med.
J.
305:
609­
613,
1992.
10.
Sharpe,
R.
M.
"
Are
Environmental
Chemicals
a
Threat
to
Male
Fertility?"
Chemistry
&
Industry
3:
87­
94,
1992.
11.
Guo,
Y.
L.,
T.
J.
Lai,
S.
H.
Ju,
Y.
C.
Chen
and
C.
C.
Hsu.,
"
Sexual
Developments
and
Biological
Findings
in
Yucheng
Children,"
Dioxin
'
93:
13th
International
Symposium
on
Chlorinated
and
Related
Compounds,
Vol.
14:
235­
238,
Vienna,
Austria,
Sept.
1993.
12.
Beardsley,
T.
A
war
not
won,
Scientific
Am.
270:
130­
138,
1994.
13.
Pollner,
F.
"
A
Holistic
Approach
to
Breast
Cancer
Research,"
Environ.
Health
Perspectives
101:
116­
120,
1993.
14.
Davis,
D.
L.,
H.
L.
Bradlow,
M.
Wolff,
T.
Woodruff,
D.
G.
Hoel
and
H.
Anton­
Culver.
"
Medical
Hypothesis:
Xenoestrogens
as
Preventable
Causes
of
Breast
Cancer,"
Environ.
Health
Perspectives
101:
372­
377,
1993.
15.
Klein,
P.
A.,
"
Immunology
and
Biotechnology
for
the
Study
and
Control
of
Infectious
Diseases
in
Wildlife
Populations."
J.
Zoo
&
Wildl.
Medicine
24:
346­
351,
1993.
16.
Fein,
G.
G.,
J.
L.
Jacobson,
S.
W.
Jacobson,
P.
M.
Schwartz
and
J.
K.
Dowler.
"
Prenatal
Exposure
to
Polychlorinated
Biphenyls:
Effects
on
Birth
Size
and
Gestational
Age,"
J.
Pediatrics
105:
315­
320,
1984.
17.
Jacobson,
S.
W.,
G.
G.
Fein,
J.
L.
Jacobson,
P.
M.
Schwartz,
J.
K.
Dowler,
"
The
Effects
of
PCB
Exposure
on
Visual
Recognition
Memory."
Child
Development
56:
853­
860,
1985.

Table
1.
Comparative
List
of
Environmental
Endocrine
Disruptors
56­
35­
9
900­
95­
8
76­
87­
9
53­
96­
3
83­
32­
9
120­
12­
7
56­
55­
3
50­
32­
8
205­
99­
2
207­
08­
9
119­
61­
9
80­
05­
7
104­
51­
8
85­
68­
7
25013­
16­
5
128­
37­
0
218­
01­
9
120­
83­
2
84­
61
­
7
84­
66­
2
103­
23­
1
117­
81­
7
84­
75­
3
84­
74­
2
131­
18­
0
131­
16­
8
193­
39­
5
29082­
74­
4
99­
99­
0
25154­
52­
3
1336­
36­
3
87­
86­
5
85­
01­
8
129­
00­
0
100­
42­
5
56­
35­
9
1746­
01­
6
X
XXXX
XXXXXXXXX
XX
XX
XX
X
X
X
XXX
X
X
x*

XXXXXX
X*

XX
X
X*

XXX
X
X
X*
X*
X*

XX
B,
F
BBCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCC
C#
Tributyl
tin
chloride
Triphenyl
tin
acetate
Triphenyl
tin
hydroxide
2­
Acetylaminofluorene
Acenaphthene
Anthracene
Benz(
a)
anthracene
Benzo(
a)
pyrene
Benzo(
b)
fluoranthene
Benzo(
k)
fluoranthene
Benzophenone
Bisphenol­
A
n­
Butyl
benzene
Butyl
benzyl
phthalate
Butylated
hydroxyanisole
(
BHA)
Butylated
hydroxytoluene
(
BHT)
Chrysene
2,4­
Dichlorophenol
Dicyclohexyl
phthalate
Diethyl
phthalate
Diethylhexyl
adipate
Diethylhexyl
phthalate
Dihexyl
phthalate
Di­
n­
butyl
phthalate
Di­
n­
pentyl
phthalate
Dipropyl
phthalate
Indeno(
1,2,3­
cd)
pyrene
Octachlorostyrene
p­
Nitrotoluene
p­
Nonylphenol
PCBs
Pentachlorophenol
Phenanthrene
Pyrene
Styrene
Tributyltin
oxide
2,3,7,8­
TCDD
CAS
Number
WWF
CDC
EPA
Type
Compound
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
8
593­
74­
8
17804­
35­
2
1897­
45­
6
118­
74­
1
8018­
01­
7
12427­
38­
2
9006­
42­
2
82­
68­
8
12122­
67­
7
137­
30­
4
15972­
60­
8
61­
82­
5
1912­
24­
9
94­
75­
7
51218­
45­
2
21087­
64­
9
1836­
75­
5
122­
34­
9
93­
76­
5
1582­
09­
8
309­
00­
2
584­
79­
2
319­
84­
6
319­
85­
7
63­
25­
2
57­
74­
9
2921­
88­
2
13121­
70­
5
52315­
07­
8
115­
32­
2
60­
57­
1
115­
29­
7
72­
20­
8
66230­
04­
4
51630­
58­
1
76­
44­
8
1024­
57­
3
115­
32­
2
143­
50­
0
58­
89­
9
121­
75­
5
16752­
77­
5
72­
43­
5
2385­
85­
5
27304­
13­
8
72­
5­
8
72­
55­
9
50­
29­
3
56­
38­
2
52645­
53­
1
NA
8001­
35­
2
39765­
80­
5
50471­
44­
8
7440­
38­
2
7440­
43­
9
7440­
50­
8
7439­
92­
1
X
XXXX
XXXX
XX
XX
XX
XXX
XXXX
XXXXXXXXXXXXXXXXXXXXX
X
X
X
XXXX
XXXX
XX
XX
XX
XXX
XXX
XX
X
XXXXXXXX
XXX
X
X
X
X
X*

X
X*

XX
X
XXXXX
X*
X*
X
X
X*
X*

X
X*

XX
X*
X*

X
XX
X*
X*
X*
X*
C#

FFFFFFFFF
HH
HHHHHHHHIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIMMMM
Dimethyl
mercury
Benomyl
Chlorothalonil
Hexachlorobenzene
Mancozeb
Maneb
Metiram
Pentachloronitrobenzene
Zineb
Ziram
Alachlor
Amitrole
Atrazine
2,4­
D
Metolachlor
Metribuzin
Nitrofen
Simizine
2,4,5­
T
Trifluralin
Aldrin
Allethrin
alpha­
BHC
beta­
BHC
Carbaryl
Chlordane
Chlorpyrifos
Cyhexatin
Cypermethrin
Dicofol
Dieldrin
Endosulfan
Endrin
Esfenvalerate
Fenvalerate
Heptachlor
Heptachlor
epoxide
Kelthane
Kepone
Lindane
(
gamma­
BHC)
Malathion
Methomyl
Methoxychlor
Mirex
Oxychlordane
p,
p'­
DDD
p,
p'­
DDE
p,
p'­
DDT
Parathion
(
ethyl)
Permethrin
Pyrethroids
(
synthetic)
Toxaphene
trans­
Nonachlor
Vinclozolin
Arsenic
Cadmium
Copper
Lead
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
9
68
48
60
103
Total
Number
in
Column
7439­
96­
5
7439­
97­
6
7440­
31­
5
96­
12­
8
116­
06­
3
26601­
64­
9
25429­
29­
2
32598­
13­
3
X
XX
X
XX
X*
X*
X*

X*
X*
X*
MMMNN
PCB
PCB
PCB
Manganese
Mercury
Tin
1,2­
Dibromo­
3­
chloropropane
Aldicarb
3,3',
4,4',
5,5'­
Hexachlorobiphenyl
3,3',
4,4',
5­
Pentachlorobiphenyl
3,3',
4,4'­
Tetrachlorobiphenyl
Abbreviations
Used
in
Table
1:

EPA
=
NERL
Endocrine
Exposure
Team
List,
October
24,
1996
from
Tammy
Jones
CDC
=
List
from
Larry
Needham
@
CDC
8/
25/
96
WWF
=
List
from
World
Wildlife
Fund
Canada
from
the
Internet
8/
27/
96
X
=
Present
on
this
list
B
=
Biocide
I
=
Insecticide
H
=
Herbicide
N
=
Nematocide
F
=
Fungacide
C
=
Industrial
Organic
Chemical
M
=
Metal
PCB
=
Specific
PCB
isomer
*
=
Analytes
that
the
EPA
NERL
Endocrine
Disruptor
Exposure
Team
plans
to
examine
in
a
multi­
media
environment
in
1997.
#
=
No
Commercial
Use;
compound
is
a
degradation
product
or
impurity
of
other
chemicals.

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
CDC'S
ANALYTICAL
APPROACH
Larry
Needham
Center
for
Disease
Control
and
Prevention,
Atlanta,
GA
The
Toxicology
Branch
of
the
Centers
for
Disease
Control
and
Prevention
(
CDC)
is
collaborating
with
other
Federal
and
international
investigators
to
determine
the
relationship
between
human
exposure
to
selected
environmental
toxicants
and
health
effects
involving
the
endocrine
system.
Several
case­
control
studies
are
underway;
in
general,
these
studies
involve
the
use
of
serum
that
was
given
for
another
purpose
in
the
past
(
e.
g.,
1970s)
by
healthy
adultssome
of
these
adults
have
since
developed
certain
diseases
(
e.
g.,
breast
cancer)
while
others
have
not.
Those
cases
and
controls
are
then
matched
(
based
on
certain
demographic
criteria),
and
the
serum
from
each
is
analyzed
for
the
environmental
toxicants,
such
as
polychlorinated
dibenzo­
p­
dioxins,
furans,
and
biphenyls
as
well
as
chlorinated
pesticides.
The
results
are
then
analyzed
to
determine
if
the
cases
have
significantly
higher
levels
of
these
compounds
than
the
controls.
In
order
to
perform
these
assessments,
we
have
developed
methods
to
measure
these
toxicants
either
separately
or
sequentially
from
one
milliliter
of
human
serum.
These
studies,
the
analytical
methods,
and
some
of
the
findings
to
date
will
be
discussed.
Analytical
approaches
for
assessing
human
exposure
to
other
potential
endocrine
disruptors,
such
as
the
phthlate
esters
and
bisphenol­
A
will
also
be
discussed.
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
10
MULTIMEDIA
ANALYTICAL
APPROACHES
TO
MONITORING
AND
MEASURING
SUSPECT
ENDOCRINE
DISRUPTING
COMPOUNDS
T.
L.
Jones1,
J.
E.
Bumgarner2,
D.
A.
Vallero2
1U.
S.
Environmental
Protection
Agency,
Office
of
Research
and
Development,
National
Exposure
Research
Laboratory,
Las
Vegas,
Nevada
89130
2U.
S.
Environmental
Protection
Agency,
Office
of
Research
and
Development,
National
Exposure
Research
Laboratory,
Research
Triangle
Park,
North
Carolina,
27711
To
develop
a
quantitative
health
and
environmental
risk
assessment
of
endocrine
disrupting
compounds
(
EDCs),
information
on
exposures
is
essential.
Since
the
late
1980'
s,
considerable
literature
has
evolved
concerning
EDCs
and
the
role
they
may
be
playing
in
decreasing
fertility
of
mammalian
and
reptilian
species
and
in
increasing
the
incidence
of
breast
and
other
reproductive­
tract
cancers.
Over
45
environmental
contaminants
or
classes
of
agents
have
been
reported
to
cause
changes
in
reproductive
and
hormonal
systems.
The
U.
S.
Environmental
Protection
Agency
(
EPA)
needs
to
be
able
to
identify
these
suspect
agents
and
estimate
the
exposure
of
susceptible
populations
to
them.
A
full
exposure
assessment
has
complex
requirements
that
require
preliminary
information
to
direct
further
research
in
this
area.
Such
research
begins
with
determining
the
levels
of
suspect
EDCs
in
the
environment.
The
EDCs
can
be
broadly
classified
into
two
categories:
those
that
can
be
analyzed
by
conventional
means
(
e.
g.,
gas
chromatography
for
organics
and
various
elemental
analyzers
for
inorganics),
and
those
that
are
non­
volatile/
non­
extractable/
thermally
labile
(
unconventional).
Although
many
of
the
EDCs
can
be
measured
in
dilute
standards,
few,
if
any,
EPA
methods
exist
for
their
measurement
in
biota
or
complex
environmental
media.
Current
methodologies
also
may
not
be
sensitive
enough
to
measure
low
levels
of
the
EDCs
in
ambient
environmental
media.
New
analytical
methodologies
will
be
needed
to
deal
with
the
monitoring
and
measurement
of
the
EDCs
in
ambient
multimedia
environment.

Notice:
The
U.
S.
Environmental
Protection
Agency
(
EPA),
through
its
Office
of
Research
and
Development
(
ORD),
funded
this
research
and
approved
this
abstract
as
a
basis
for
an
oral
presentation.
The
actual
presentation
has
not
been
peer
reviewed
by
EPA.

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
STATUS
OF
EPA
LABORATORY
METHODS
FOR
MEASURING
ENDOCRINE
DISRUPTORS
Jerry
Parr
Quanterra
Environmental
Services,
4955
Yarrow
Street,
Arvada,
Colorado
80002
303/
421­
6611,
<
parrj@
quanterra.
com>

ABSTRACT
Over
the
past
20
years,
EPA
has
published
hundreds
of
analytical
laboratory
procedures
for
measuring
pollutants
in
air,
water,
and
soil.
Unfortunately,
these
procedures
may
not
be
applicable
to
the
current
need
of
providing
baseline
data
for
assessing
the
fate
and
transport
of
endocrine
disruptor
chemicals
(
EDCs)
in
the
environment.
This
presentation
will
focus
on
the
capability
of
existing
EPA
procedures
to
provide
measurement
data
for
EDCs
in
environmental
samples.
A
suggested
EDC
analyte
list
will
be
presented
listing
existing
EPA
method
performance
data
(
sensitivity
and
accuracy).
This
table
can
be
used
to
determine
gaps
in
EPA
methods.

INTRODUCTION
Within
the
last
decade,
the
first
generation
of
children
whose
parents
were
first
exposed
to
organochlorine
pesticides
during
fetal
development
have
become
adults
and
reached
their
reproductive
stage
of
life.
There
is
growing
evidence
that
a
number
of
synthetic
chemicals
may
be
disrupting
the
endocrine
(
hormonal)
systems
of
mammals.
1­
4
These
endocrine
disruptor
chemicals
(
EDCs)
may
cause
a
variety
of
problems
with
development,
behavior,
and
reproduction.
Thus,
the
current
reproductive
human
population
may
be
at
risk
due
to
environmental
activities
conducted
50
years
ago.

The
endocrine
system
helps
guide
development,
growth,
reproduction,
behavior
and
other
bodily
functions.
It
is
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
11
comprised
of
endocrine
glands
and
hormones.
Some
of
the
major
endocrine
glands
are
the
pituitary,
thyroid,
pancreas,
adrenal,
and
the
male
and
female
gonads
(
testes
and
ovaries).
Endocrine
glands
produce
hormones
and
secrete
them
directly
into
the
bloodstream.
Hormones
(
estrogen,
testosterone,
adrelin,
etc.)
act
as
chemical
messengers,
traveling
through
the
blood
to
distant
tissues
and
organs,
where
they
can
bind
to
specific
cell
sites
called
receptors.
By
binding
to
receptors,
hormones
trigger
various
responses
in
the
tissues
containing
the
receptors.
For
example,
PCBs
can
bind
to
the
receptor
for
estrogen,
yielding
an
estrogenic
effect
such
as
feminization.

An
endocrine
disruptor
chemical
has
been
defined
as
"
an
exogenous
agent
that
interferes
with
the
synthesis,
secretion,
transport,
binding,
action,
or
elimination
of
natural
hormones
in
the
body
that
are
responsible
for
the
maintenance
of
homeostasis,
reproduction,
development,
and/
or
behavior."
5
A
variety
of
chemicals,
including
certain
pesticides,
have
been
found
to
cause
endocrine
disruption
in
laboratory
animal
studies.
Observed
effects
have
included
disruption
of
female
and
male
reproductive
function
(
such
as
disruption
of
normal
sexual
differentiation,
ovarian
function,
sperm
production,
and
pregnancy)
and
effects
on
the
thyroid
gland.

Some
of
the
research
indicates
that
effects
can
occur
at
trace
levels.
For
example,
levels
of
bisphenol
A
in
the
low
microgram
range
have
been
shown
to
have
an
estrogenic
effect
in
animals.
6
At
this
time,
conclusive
evidence
does
not
exist
which
can
establish
a
causal
relationship
between
ambient
levels
of
EDC
exposure
and
observed
adverse
effects.
5
However,
all
of
the
research
indicates
that
much
more
information
is
needed.

Relative
to
environmental
monitoring
activities,
the
following
questions
need
to
be
answered:
w
What
specific
chemicals
are
of
concern?

w
What
matrices
need
to
be
measured?

w
How
low
do
the
measurements
need
to
be
made?

w
What
quality
of
data
are
needed?

These
issues
are
discussed
in
more
detail
below.

ANALYTE
LIST
Existing
research
indicates
that
many
synthetic
chemicals
may
adversely
affect
the
endocrine
system.
Since
studies
for
this
type
of
effect
have
not
been
routinely
performed,
it
is
likely
that
the
list
of
analyses
of
concern
will
substantially
grow.

Table
1
provides
a
list
of
chemicals
which
reflects
the
current
research.
The
list
was
compiled
from
the
literature
sources
in
the
bibliography,
including
the
Internet
sources
listed.

The
table
also
has
a
column
showing
relative
risk,
as
established
by
the
Illinois
EPA.
7
MATRICES
For
the
past
25
years,
the
environmental
testing
industry
has
focused
primarily
on
environmental
media­­
air,
water
soil.
This
focus
was
due
to
concerns
relative
to
determining
the
concentration
of
pollutants
in
the
ambient
environment.

With
the
new
emphasis
on
EDCs,
the
focus
will
likely
shift
to
ingestion
pathways
and
to
a
better
understanding
of
ambient
levels
in
the
exposed
population.
Thus,
we
will
need
to
establish
rugged
methods
for
measuring
EDCs
at
trace
levels
in
foods
and
tissue.

As
shown
in
Table
1,
EPA
methods
have
not
been
established
for
many
EDC
analyses.
However,
many
of
the
existing
methods
can
likely
be
adapted
to
provide
reliable
measurements.
Other
technology
widely
available
(
e.
g.
LC/
MS)
may
be
necessary
for
some
analyses.
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
12
Based
on
the
current
list,
there
appear
to
be
no
"
impossible"
analyses.

DETECTION
LEVELS
At
this
time,
the
levels
at
which
measurements
need
to
be
made
are
not
known.
Current
research
indicates
that
levels
at
or
below
our
current
technology
will
be
needed.

Table
1
list
typical
levels
of
measurement
which
are
achievable
using
current
EPA
methods.

Better
sample
preparation
techniques,
better
cleanup
techniques,
and
more
sensitive
instrumentation
may
be
needed.

DATA
QUALITY
Given
the
importance
of
EDC
measurements,
I
believe
we
need
to
move
away
from
a
concept
of
"
data
of
known
quality"
to
a
concept
which
establishes
a
data
quality
goal.
In
my
opinion,
we
will
need
methods
which
can
provide
quantitative
results.
One
measure
of
this
level
of
data
quality
is
an
accuracy
of
65­
135%.
8­
9
As
shown
in
Table
1,
for
many
EDC
analyses
for
which
EPA
methods
exist,
the
accuracy
of
the
methods
is
at
best
semi­
quantitative.
The
accuracy
of
the
methods
can
be
improved
by
procedural
changes
such
as
better
instrument
calibration,
isotope
dilution,
etc.

SUMMARY
The
concern
over
EDCs
in
the
environment
will
require
knowledge
 
What
chemicals
are
present,
and
at
what
concentrations,
in
what
species,
and
via
what
pathways?

The
importance
of
these
measurements
is
such
that
we
must
establish
rugged,
reliable
methods
for
measuring
EDCs
in
a
variety
of
challenging
matrices
at
very
low
levels
with
very
good
data
quality.

I
call
on
EPA
and
all
of
those
who
develop
methods
to
begin
to
fill
in
the
gaps
in
Table
1.

REFERENCES
1.
Colborn,
T.
and
C.
Clement
(
1992)
Chemically
Induced
Alterations
in
Sexual
and
Functional
Development:
The
Wildlife/
Human
Connection.
Princeton,
N.
J.
Princeton
Scientific
Publishing.
2.
Colburn,
T.,
F.
vom
Saal
and
A.
M.
Soto
(
1993)
"
Developmental
Effects
of
Endocrine­
Disrupting
Chemicals
in
Wildlife
and
Humans,"
Environmental
Health
Perspectives,
Vol
101,
Number
5
3.
Lyons,
G.
(
1995)
"
Phthalates
in
the
Environment,"
World
Wildlife
Fund
UK.
4.
"
Assessing
the
Risks
of
Adverse
Endocrine­
Mediated
Effects,"
Endocrine/
Estrogen
Letter,
February
24,
1997.
5.
"
Special
Report
on
Environmental
Endocrine
Disruption:
an
Effects
Assessment
and
Analysis,"
EPA/
630/
R­
96­
012,
February
1997
6.
"
Endocrine
Disruption
is
Caused
by
Low
Doses
of
Bisphenol
A,"
Endocrine/
Estrogen
Letter,
February
17,
1997
7.
"
Endocrine
Disruptors
Strategy,"
Illinois
EPA,
February
1997
8.
Quality
Assurance
of
Chemical
Measurements,
John
Taylor,
Lewis
Publishers,
1987
9.
"
Developing
a
Uniform
Approach
for
Complying
with
EPA
Methods";
Jerry
Parr,
Peggy
Sleevi,
Deborah
Loring'
Nancy
Rothman;
Seventh
Annual
Quality
Assurance
and
Waste
Testing
Conference;
Washington
D.
C.;
July
1991.

INTERNET
SITES
http://
www.
mst.
dk
/
liste.
htm
http://
www.
libertytree.
org/
Trenches/
endo/
endo.
html
http://
www.
epa.
gov/
opptintr/
opptendo/
index.
htm
http://
www.
OSF­
facts.
com
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
13
Table
1.
Capabilities
of
Existing
EPA
Methods
to
Determine
EDCs
in
the
Environment
SQ
10
0.05
504
GC/
ECD
­
EDB
NA
NA
NA
NA
NA
S
Di­
n­
propyl
phthalyte
NA
NA
NA
NA
NA
S
Di­
n­
pentyl
phthalate
NA
NA
NA
NA
NA
S
Di­
n­
hexyl
phthalate
SQ
200
10
625,
8270
GC/
MS
S
Di­
n­
butyl
phthalate
SQ
0.2
0.001
8290,
1613
GC/
HRMS
K
Dioxins/
furans
SQ
2
0.1
608,
8080
GC/
ECD
K
Dieldrin
NA
NA
NA
NA
NA
S
Dicyclohexyl
phtholate
NA
NA
NA
8081
GC/
ECD
K
Dicofol
(
kelthane)
NA
NA
NA
NA
NA
S
Di
(
2­
Ethylhexyl)
Adipate
SQ
200
10
615,
8270
GC/
MS
S
2,4­
Dichlorophenol
NA
NA
NA
NA
NA
K
DES*
SQ
2
0.1
608,
8080
GC/
ECD
K
DDT
SQ
2
0.1
608,
8080
GC/
ECD
K
DDE
SQ
2
0.1
608,
8080
GC/
ECD
K
DDD
SQ
100
1
524,
8260
GC/
MS
K
DBCP
NA
NA
NA
NA
NA
Dacthal
SQ
20
1
615,
8150
GC/
ECD
P
2,4­
D
NA
NA
NA
NA
NA
S
Cypermethrin
SQ
1
0.05
608,8080
GE/
ECD
K
Chlordane
SQ
200
10
8270
GC/
MS
S
Carbaryl
Q
2
0.02
200.8,
6020
ICP/
MS
P
Cadmium
NA
NA
NA
NA
NA
S
p­
tert­
Butyl
phenol
NA
NA
NA
NA
NA
S
p­
sec­
Butyl
phenol
NA
NA
NA
NA
NA
S
tert­
Butylheydroxy
anisole
SQ
200
10
625,
8270
GC/
MS
S
Butyl
benzyl
phthalate
SQ
100
5
524,
8260
GC/
MS
­
n­
Butyl
benzene
SQ
200
10
625,
8270
GC/
MS
P
bis
(
2­
Ethylhexyl)
phthalate
NA
NA
NA
NA
NA
P
Bisphenol­
A
SQ
1
0.05
608,
8080
GC/
ECD
K/
P
BHCs
NA
NA
NA
NA
NA
­
Benzophenone
NA
40
2
8321
LC/
MS
P
Benomyl
SQ
2
0.1
525
GC/
MS
K
Atrazine
NA
NA
NA
NA
NA
P
Amitrole
SQ
1
0.05
8080
GC/
ECD
P
Aldrin
NA
100
5
8321
LC/
MS
S
Aldicarb
SQ
20
1
525
GC/
MS
P
Alachlor
Solids
Water
Accuracy
Sensitivity
(
ppb)
Example
EPA
Methods(
s)
Methodology
IEPA
Category
EDC
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
14
NA
NA
NA
NA
NA
P
Tifluralin
NA
NA
NA
NA
NA
K
Tributyl
tin
1
0.05
619
GC/
NPD
­
Triazines
NA
10
0.5
525
GC/
MS
­
Transnonachlor
200
5
608,
8080
GC/
ECD
K
Toxaphene
20
1
615,
8150
GC/
ECD
P
2,4,5­
T
50
5
624,
8260
GC/
MS
P
Styrene
NA
NA
NA
NA
NA
­
Pyrimidine
carbinol
NA
NA
NA
8081
GC/
ECD
S
Permethrin
NA
NA
NA
NA
NA
S
p­
tert­
Pentyl
phenol
NA
NA
NA
NA
NA
S
p­
iso­
Pentyl
phenol
SQ
1000
50
625,
8270
GC/
MS
P
Pentachlorophenol
SQ
20
1
608,
8080
GC/
ECD
K
PCBs
NA
NA
NA
NA
NA
P
PBBs
SQ
200
10
615,
8270
GC/
MS
S
PAHs
NA
NA
NA
NA
NA
­
Oxychlordane
NA
NA
NA
NA
NA
P
p­
Octylphenol
NA
NA
NA
NA
NA
S
Octachlorostyrene
NA
NA
NA
NA
NA
K
p­
Nonylphenol
SQ
200
10
8330
HPLC
­
4­
Nitrotoluene
NA
NA
NA
8081
GC/
ECD
S
Nitrofen
NA
NA
NA
8081
GC/
ECD
P
Mirex
NA
NA
NA
NA
NA
S
Metribuzin
NA
NA
NA
NA
NA
P
Metiram
SQ
5
0.1
614,
8140
GC/
FPD
P
Methyl
parathian
SQ
10
0.5
608,
8080
GC/
ECD
K
Methoxychlor
NA
100
5
8321
LC/
MS
S
Methomyl
Q
20
0.2
245.2,
7470
CVAA
P
Mercury
NA
NA
NA
NA
NA
P
Maneb
NA
NA
NA
NA
NA
P
Mancozeb
SQ
5
0.1
614,
8140
GC/
FPD
S
Malathion
Q
2
0.02
200.8,
6020
ICP/
MS
P
Lead
NA
NA
NA
8270
GC/
MS
K
Kepone
SQ
200
10
625,
8270
GC/
MS
P
Hexachlorobenzene
SQ
2
0.05
608,
8080
GC/
ECD
P
Heptachlor
epoxide
SQ
2
0.05
608,
8080
GC/
ECD
P
Heptachlor
NA
NA
NA
NA
NA
S
Fenvalerate
SQ
5
0.1
614,
8140
GC/
NPD
P
Ethylparathion
NA
NA
NA
NA
NA
S
Esfenvalerate
SQ
2
0.1
608,
8080
GC/
ECD
P
Endrin
SQ
5
0.1
608,
8080
GC/
ECD
K
Endosulfans
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
15
NA
NA
NA
NA
NA
S
Ziram
NA
NA
NA
NA
NA
P
Zineb
NA
NA
NA
NA
NA
P
Vinclozolin
*
Best
documented
evidence
of
endocrine
effect
in
humans.
Unlikely
compound
for
monitoring
due
to
ban
on
use
and
manufacture.

NOTES:
Sensitivity
and
accuracy
values
are
rough
averages
of
data
in
EPA
Methods
CVAA:
Cold
Vapor
Atomic
Adsorption
GC/
ECD:
Gas
Chromatography
Electron
Capture
Detection
GC/
PPD:
Gas
Chromatography
Flame
Photometric
Detector
GC/
HRMS:
Gas
Chromatography
High
Resolution
Mass
Spectrometry
GC/
MS:
Gas
Chromatography
Mass
Spectrometry
GC/
NPD:
Gas
Chromatography
Nitrogen
Phosphorous
Detector
HPLC:
High
Pressure
Liquid
Chromatography
ICP/
MS:
Inductively
Coupled
Plasma
Mass
Spectrometry
LC/
MS:
Liquid
Chromatography
Mass
Spectrometry
Q:
Quantitative;
accuracy
is
generally
+/­
35%
SQ:
Semiquantitative;
accuracy
is
generally
+/­
50%
NA:
No
routine
EPA
method
exists
for
this
analyte;
no
published
performance
data
IEPA
Categories:
K­
Known;
P
­
Probable;
S
­
Suspect
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
REFERENCE
MATERIALS
FOR
ENDOCRINE
DISRUPTING
COMPOUND
(
EDC)
ANALYSIS:
AN
OVERVIEW
M.
A.
Re,
Technical
Director
Analytical
Reference
Materials
(
ARM)
Division,
Radian
International
LLC,
P.
O.
Box
201088,
Austin,
TX
78720.

ABSTRACT
Analytical
reference
materials
will
be
a
critical
part
of
Endocrine
Disrupting
Compound
(
EDC)
analysis.
Aspects
of
research
and
testing
for
EDCs,
including
toxicology
studies,
environmental
testing,
screening
of
chemicals
for
endocrine
disrupting
effects,
and
in
food
analysis,
will
require
accurate
high
quality
reference
materials.
Neat
reference
materials
must
be
of
high
purity
and
correct
identity,
and
standard
solutions
must
be
accurate,
consistent,
and
reliable.
Sensitivity
and
reliability
of
analytical
methods
will
be
important.
Isotope
dilution
methodology
has
emerged
as
the
analytical
method
of
choice
for
many
environmental
pollutants.
This
technique
provides
many
advantages
resulting
in
higher
quality
data.
The
availability
of
stable
isotope­
labeled
reference
materials
will
help
to
make
this
methodology
a
viable
alternative
to
existing
analytical
methods
for
EDC
analysis.

INTRODUCTION
There
is
a
rising
interest
in
environmental
pollutants
that
affect
hormonal
functions
in
animals.
Many
of
these
chemicals
appear
to
adversely
affect
the
endocrine
system
by
either
mimicking
hormones
or
by
blocking
their
activity.
The
activity
of
these
chemicals
may
cause
a
number
of
defects
including:
malformations
of
newborns,
abnormal
sperm
or
low
sperm
counts,
feminization
of
males,
masculinization
of
females,
thyroid
dysfunction,
goiter,
learning
disabilities,
and
hyperactivity
in
children.
1
The
effects
appear
to
be
strongest
on
the
developing
fetus,
and
the
time
of
exposure
is
more
critical
than
the
concentration
level.

This
new
class
of
chemicals,
called
endocrine
disrupting
chemicals
(
EDCs),
consists
of
many
synthetic
and
naturally
occurring
substances.
Some
of
these
materials
are
ubiquitous
and
have
been
known
to
exist
in
the
environment
for
a
number
years.
The
sources
of
possible
EDCs
include:
naturally
occurring
substances,
pharmaceutical
products,
synthetic
industrial
chemicals,
synthetic
pesticides,
and
other
environmental
contaminants
(
such
as
banned
pesticides,
PCBs,
and
dioxins).
The
World
Wildlife
Fund
(
WWF),
Centers
for
Disease
Control
(
CDC),
and
the
U.
S.
Environmental
Protection
Agency
(
EPA)
have
each
identified
compounds
as
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
16
possible
EDCs.
A
comprehensive
list
is
shown
in
Table
1.
The
list
does
not
include
many
of
the
hundreds
of
registered
pesticides
and
thousands
of
industrial
chemicals
currently
being
produced.

The
interest
in
EDCs
is
widespread,
and
the
U.
S.
Government
is
funding
over
400
research
projects
to
explore
their
chemical
properties
and
effects.
2
Several
challenges
in
EDC
research
include
development
of:
toxicological
research
programs
to
understand
the
complexities
of
endocrine
disruptors;
screening
tests
to
identify
potential
EDCs;
and
analytical
methods
to
measure
EDCs
in
the
environment.
3,4
Recent
legislation
passed
by
the
U.
S.
Congress
provided
the
EPA
with
a
timeline
for
developing
a
screening
program.
3
Toxicological
studies
and
development
of
screening
methods
are
complicated
because
there
does
not
appear
to
be
a
common
structural
link
between
suspected
EDCs.
Because
the
compounds
appear
to
exhibit
their
effects
at
very
low
levels,
lower
detection
limits
may
be
necessary
in
environmental
testing
and
may
present
analytical
challenges.

As
toxicological
studies,
screening
methods,
and
analytical
protocols
are
developed,
the
need
for
reference
materials
will
increase
dramatically.
Highly
pure
authentic
compounds
and
accurate
standard
reference
solutions
(
SRSs)
will
be
an
integral
part
of
all
aspects
of
EDC
research.
Meaningful
toxicological
results
and
development
of
screening
tests
will
require
using
compounds
with
confirmed
identity
and
free
from
impurities.
Establishing
sensitive
analytical
testing
methods
will
not
be
possible
without
the
availability
of
appropriate
and
accurate
SRSs.
A
number
of
factors
which
should
be
considered
in
developing
reference
materials
include:
types
(
neat,
SRS,
or
other
matrix)
needed;
analytical
specifications
for
each
type;
and
applicability
for
intended
uses.

CHARACTERISTICS
OF
NEAT
REFERENCE
MATERIALS
EDC
research
and
testing
will
require
high
purity
neat
materials.
In
toxicological
studies
and
screening
method
development,
neat
materials
will
be
needed
for
both
in
vivo
and
in
vitro
assays.
In
environmental
testing,
neat
materials
will
be
needed
for
preparation
of
quantitative
analytical
solutions.
Each
area
will
have
a
variety
of
specifications
for
neat
materials.
For
example,
a
material
used
for
toxicological
testing
should
be
of
very
high
purity
(
99+%)
and
impurities
that
may
adversely
affect
a
sensitive
assay
should
be
minimized.
Some
impurities
may
also
exhibit
endocrine
disrupting
properties
that
would
lead
to
false
conclusions
about
the
original
compound.
Purity
requirements
for
a
neat
material
used
to
prepare
an
SRS
for
environmental
analysis
may
not
be
as
stringent
since
minor
impurities
typically
do
not
affect
analytical
results.
In
these
cases,
98%
chemical
purity
may
be
acceptable.
An
analytical
method
may
require
a
specific
isotopically
labeled
material
that
would
not
be
necessary
in
developing
screening
assays.
The
two
common
factors
for
all
neat
reference
material
needs
are
that
the
substance
must
be
of
known
identity
and
purity.

Identity
of
a
neat
reference
material
should
be
unequivocally
established
to
ensure
studies
are
not
performed
on
the
wrong
compound.
A
common
occurrence
is
the
misidentification
of
isomeric
compounds.
Isomers
(
for
example,
cis
and
trans
isomers)
often
have
very
different
toxicological
and
analytical
properties,
and
using
the
incorrect
isomer
would
yield
erroneous
results
and
conclusions.

One
potential
EDC
is
the
banned
pesticide
chlordane.
Technical
chlordane
is
actually
a
mixture
of
a
number
of
similar
chlorinated
compounds.
Cis­
and
trans­
chlordane
are
two
major
components
of
the
technical
mixture.
These
two
materials
have
very
similar
physical
properties
but
may
have
very
different
toxicological
properties.
Positive
structure
identification
is
critical
in
reference
material
production.
Figure
1
illustrates
very
similar
mass
spectra
of
cis­
and
trans­
chlordane
and
unequivocal
structure
determination
is
not
possible.
Detailed
analysis
of
the
1H
NMR
spectra
(
Figure
2)
indicates
differences
in
coupling
constants
of
cis
and
trans
hydrogens
and
allows
for
positive
identification
of
isomers.

The
chlordane
example
illustrates
the
value
of
performing
multiple
analytical
methods
to
determine
identity,
and
that
relying
only
on
a
single
method
may
lead
to
misidentification
of
the
reference
material.
Common
techniques
used
in
combination
to
establish
identity
include:
gas
chromatography­
mass
spectrometry
(
GC­
MS),
fourier
transform
infrared
spectroscopy
(
FT­
IR),
nuclear
magnetic
resonance
spectroscopy
(
NMR),
melting
point
determination,
refractive
index
determination,
and
gas
chromatography
(
GC).

Determining
purity
is
also
critical
to
the
quality
of
the
reference
material.
Purity
specifications
must
be
adequate
for
the
material's
intended
use
as
illustrated
above.
Some
of
the
common
impurities
found
in
neat
reference
materials
include:
residual
solvents;
water;
inert
materials
or
surfactants;
and
reaction
by­
products,
intermediates,
and
starting
materials.
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
17
Simazine
is
a
herbicide
and
a
suspected
EDC
that
was
procured
for
use
as
a
neat
reference
material.
Purity
analysis
by
GC
with
a
flame
ionization
detector
(
FID)
indicated
a
purity
of
99%.
A
second
purity
assay
by
thin
layer
chromatography
(
TLC)
indicated
a
spot
with
an
Rf
of
0.76
and
an
additional
spot
at
the
baseline
using
neutral
alumina
and
a
chloroform
mobile
phase.
Additionally,
elemental
analysis
for
carbon
and
hydrogen
did
not
agree
with
theoretical
values
(
C,
theoretical­
41.69%,
found­
38.65%;
H,
theoretical­
6.00%,
found­
5.42%).
By
chromatographic
purification,
it
was
determined
that
the
simazine
contained
greater
than
10%
of
an
impurity.
In
this
case
using
only
gas
chromatographic
methods
for
purity
determination
would
have
resulted
in
an
incorrect
purity
assay.
The
impurity
may
have
contributed
to
a
false
conclusion
in
a
toxicological
study.

Table
2
contains
four
representative
compounds
that
are
suspected
EDCs.
These
materials
were
procured
for
use
as
reference
materials.
Each
compound
was
analyzed
for
chemical
purity
by
multiple
methods
including
GC­
FID
and
elemental
analysis
(
EA).
Carbaryl
is
an
example
of
a
material
that
required
no
additional
purification
and
purity
data
was
consistent
by
all
methods.
Nitrofen
is
an
example
where
GC
data
indicated
impurities,
but
EA
data
was
acceptable.
In
this
case
further
purification
was
required.
Using
only
EA
would
have
resulted
in
an
inaccurate
purity
assignment.
Parathion
and
vinclozolin
are
examples
where
GC
data
was
acceptable
but
EA
data
was
not.
The
two
materials
contained
residual
solvent
and/
or
water
and
further
drying
was
necessary.
Gravimetric
preparation
using
impure
neat
materials
would
have
resulted
in
solution
concentrations
with
a
low
bias.

As
with
identity,
the
above
examples
illustrate
that
multiple
methods
must
be
used
for
purity
assays.
Some
of
the
common
techniques
used
for
purity
determination
include:
GC­
FID,
high
performance
liquid
chromatography
(
HPLC),
differential
scanning
calorimetry
(
DSC),
EA,
TLC,
and
melting
point
or
refractive
index
determinations.
Recent
publications
include
more
detailed
discussions
of
neat
reference
materials
and
their
properties.
5,6
CHARACTERISTICS
OF
STANDARD
REFERENCE
SOLUTIONS
(
SRSs)

Standard
reference
solutions
of
EDCs
will
primarily
be
used
for
applications
in
analytical
method
development
and
environmental
testing
and
quantitation.
Traditionally,
SRSs
have
been
required
for:
calibration
of
instrumentation;
calibration
checks;
and
internal,
surrogate,
matrix
spiking,
and
quality
control
(
QC)
standards.
The
USEPA
has
published
methods
including
the
SW­
846
series
that
describe
in
detail
the
various
types
and
uses
of
standards.
Many
of
these
standards
are
commercially
available.

Accuracy,
traceability,
consistency,
homogeneity,
and
stability
should
always
be
considered
when
preparing
and
verifying
an
SRS.
Accuracy
is
the
nearness
of
a
solution's
true
concentration
to
its
nominal
or
theoretical
concentration.
Traceability
is
the
documented
unbroken
chain
of
comparisons
back
to
a
recognized
national
or
international,
or
an
ab
initio
source.
Consistency
is
a
measure
of
batch­
to­
batch
variability.
Homogeneity
is
a
measure
of
variability
within
a
batch.
Stability
refers
to
the
variability
of
analyte
concentration
in
a
solution
over
time.

Obtaining
an
accurate
SRS
begins
with
verification
of
the
neat
material
as
described
above.
Analytes
should
be
weighed
using
balances
calibrated
with
NIST
traceable
weights
to
provide
gravimetric
verification.
Dilutions
should
be
performed
using
Class
A
glassware
at
a
minimum.
Further
verification
of
accuracy
should
include
comparison
to
independently
prepared
calibration
solutions,
or
at
minimum
to
a
second
independently
prepared
solution.

Consistency
and
homogeneity
of
an
SRS
can
be
verified
by
comparison
studies.
Comparing
each
new
batch
of
standard
to
the
previous
batch
will
demonstrate
consistency.
When
a
batch
of
standard
solution
is
packaged
into
multiple
containers,
homogeneity
should
be
verified.
This
can
be
done
by
removing
random
containers
from
the
batch
and
analyzing
for
concentration
differences.

Stability
issues
should
be
considered
to
help
determine
a
valid
shelf
life
of
the
SRS.
Some
analyses
may
decompose
over
time
or
react
with
the
solvent,
moisture,
or
oxygen.
On­
going
stability
studies
will
help
identify
these
problems.
Testing
can
be
achieved
by
storing
archive
samples
over
time
and
then
comparing
to
a
new
solution.
Accelerated
stability
studies
are
often
performed
by
placing
a
test
standard
at
elevated
temperatures
and
comparing
to
a
new
solution.
Many
suspected
EDCs
such
as
PCBs
and
polycyclic
aromatic
hydrocarbons
(
PAHs)
should
be
quite
stable
in
solution.
Others
(
e.
g.,
pesticides)
may
have
stability
problems
in
solution.
A
more
detailed
discussion
of
standard
solution
verification
has
recently
been
published.
7
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
18
REFERENCE
MATERIALS
AND
ANALYTICAL
METHOD
DEVELOPMENT
Establishing
specifications
for
SRSs
used
in
EDC
analysis
will
be
dependent
on
the
analytical
method.
For
example,
a
gas
chromatography/
electron
capture
detector
(
GC­
ECD)
method
will
require
a
different
set
of
calibration
standards
than
a
GC­
MS
method.
While
there
are
a
number
of
existing
analytical
methods
for
many
analyses
currently
listed
as
potential
EDCs,
there
will
be
a
challenge
to
develop
protocols
for
additional
analyses
and
to
meet
the
requirements
for
lower
detection
limits.
Historically,
analytical
method
development
has
been
somewhat
dependent
on
the
development
of
analytical
equipment
and
the
availability
of
reference
materials.
This
is
exemplified
by
Clement's
discussion
of
method
development
for
trace
analysis
of
chlorinated
dibenzo­
p­
dioxins
and
dibenzofurans8
and
the
development
of
EPA
method
1613
discussed
below.

Requirements
for
low
detection
limits
in
EDC
analysis
will
lead
to
new
method
development.
One
technique
that
has
proven
to
be
very
effective
in
environmental
analysis
is
stable
isotope
dilution
mass
spectrometry.
8,9
Theoretical
aspects
of
this
methodology
are
discussed
in
detail
by
Pickup
and
McPherson.
10
The
basis
of
the
technique
is
that
stable
isotope­
labeled
compounds
are
used
as
internal
standards
in
quantitation
of
analyses.

Internal
standards
are
commonly
used
in
analytical
testing.
A
brief
description
of
the
method
follows.
Calibration
solutions
are
prepared
containing
known
amounts
of
the
analyses
of
interest.
A
known
amount
of
internal
standard
is
added
and
a
response
factor
is
calculated
based
on
the
response
of
the
internal
standard
and
the
analyte
of
interest.
Samples
are
then
"
spiked"
with
a
known
amount
of
internal
standard
and
analyses
of
interest
can
be
quantitated
using
the
response
of
the
unknown
and
the
response
factor
for
that
analyte.
The
challenge
of
the
technique
is
to
identify
appropriate
internal
standards
that
have
similar
properties
to
the
analyses
of
interest
and
do
not
interfere
with
the
analysis.

In
isotope
dilution
mass
spectrometry
the
internal
standard
is
a
stable
isotope­
labeled
analog
of
the
analyte
of
interest.
The
internal
standard
and
the
analyte
of
interest
differ
only
in
their
molecular
mass.
Stable
isotopes
such
as
2H,
13C,
and
15N
are
commonly
used.
This
technique
addresses
many
of
the
challenges
of
selecting
an
internal
standard
since
the
two
compounds
will
exhibit
very
similar
properties.
It
also
takes
advantage
of
mass
spectrometry
which
has
seen
many
advances
in
the
last
few
years.
Using
high
resolution
mass
spectrometry
(
HRMS)
in
the
selective
ion
monitoring
(
SIM)
mode
provides
a
very
sensitive
technique.
A
disadvantage
to
using
isotope­
dilution
methods
is
that
a
labeled
analog
is
needed
for
each
analyte
of
interest,
whereas
only
representative
compounds
are
needed
for
traditional
internal
standard
techniques.
This
disadvantage
has
hindered
the
development
of
isotope­
dilution
method
development.
A
good
example
of
this
is
in
the
development
of
EPA
Method
1613
for
analysis
of
dioxins
and
furans.
Figure
3
represents
a
timeline
in
the
development
of
this
method.
Isotope
dilution
was
discussed
in
detail
in
the
mid
1970s.
10
Gas
chromatographic
separations
were
greatly
improved
by
the
development
of
capillary
column
technology
in
the
early
1980s.
Even
though
the
analytical
tools
were
in
place,
Method
1613
was
not
developed
for
another
decade.
The
method
could
only
be
established
after
13C­
labeled
analogs
of
all
17
chlorinated
dioxin
and
furan
toxic
congeners
were
made
available
in
the
late
1980s.

Isotope
dilution
mass
spectrometry
is
ideally
suited
for
environmental
analysis
of
EDCs
because
of
its
sensitivity
and
past
use
in
trace
analysis.
It
may
be
possible
to
adapt
existing
methods
to
many
EDCs.
While
obtaining
isotope­
labeled
analogs
for
all
potential
EDCs
may
be
a
challenge,
many
labeled
analogs
of
compounds
in
Table
1
are
currently
available
and
are
listed
in
Table
3.
As
additional
EDCs
are
identified,
labeled
analogs
will
most
certainly
be
developed
to
meet
the
demands
for
method
development
and
testing.

CONCLUSION
The
adverse
effects
of
some
chemicals
on
endocrine
systems
of
wildlife
and
humans
appears
to
be
real,
and
research
and
development
for
toxicology,
screening
methods
and
environmental
testing
are
rapidly
growing.
The
availability
of
reference
materials
will
greatly
assist
these
studies.
Appropriate
reference
materials
must
be
available
for
each
area
of
research
and
testing.

Neat
reference
materials
will
be
needed
for
all
aspects
of
EDCs
research
and
testing.
They
must
be
fully
characterized,
and
data
for
identity
and
purity
should
accompany
each
material.
Multiple
complimentary
analytical
methods
should
be
used
to
establish
purity
and
identity.
Relying
on
a
single
method
may
lead
to
errors
and
bias
results.
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
19
Standard
reference
solutions
will
be
used
extensively
in
environmental
testing.
Preparing
and
verifying
SRSs
should
include
components
for
accuracy,
traceability,
consistency,
homogeneity,
and
stability.

Developing
analytical
methods
for
EDC
testing
will
be
dependent
on
the
availability
of
reference
materials.
To
gain
the
necessary
sensitivity,
methods
such
as
isotope
dilution
mass
spectrometry
should
be
employed.
This
will
be
possible
because
current
methods
may
be
adapted
and
isotope­
labeled
reference
materials
are
currently
available
for
use
in
developing
new
methods.

REFERENCES
1.
Colborn,
T.;
Dumanoski,
D.,
Myers,
J.
P.
Our
Stolen
Future;
Dutton:
New
York,
1996.
2.
Johnson,
J.
Environmental
Science
&
Technology
1996,
30,
476A­
77A.
3.
Patlak,
M.
Environmental
Science
&
Technology
1996,
30,
540A­
44A.
4.
Yosie,
T.
Environmental
Science
&
Technology
1996,
30,
498A.
5.
Clement,
R.
E.;
Keith,
L.
H..
Siu,
K.
W.
M.
Reference
Materials
for
Environmental
Analysis;
CRC
Press:
Boca
Raton,
1997,
3­
14.
6.
Re,
M.
A.
Analytical
Reflections;
Radian
International
LLC,
1996,
September,
6­
7.
7.
Hannon,
C.
L.
Analytical
Reflections;
Radian
International
LLC,
1996,
December,
1­
2.
8.
Clement,
R.
E.
Analytical
Chemistry
1991,
63,
1130A­
37A.
9.
Clement,
R.
E.;
Keith,
L.
H.;
Siu,
K.
W.
M.
Reference
Materials
for
Environmental
Analysis;
CRC
Press:
Boca
Raton,
1997,
15­
23.
10.
Pickup,
J.
F.;
McPherson,
K.
Analytical
Chemistry
1976,
48,
1885­
90.

Table
1.
Possible
Endocrine
Disrupting
Chemicals
(
EDCs)

21087­
64­
9
Metribuzin
104­
51­
8
n­
Butylbenzene
51218­
45­
2
Metolachlor
128­
37­
0
Butylate
hydroxytoluene
(
BHT)
9006­
42­
2
Metiram
25013­
16­
5
Butylated
hydroxyanisole
(
BHA)
72­
43­
5
Methoxychlor
80­
05­
7
Bisphenol­
A
16752­
77­
5
Methomyl
319­
85­
7
beta­
BHC
7439­
97­
6
Mercury
319­
84­
6
alpha­
BHC
7439­
96­
5
Manganese
119­
61­
9
Benzophenone
12427­
38­
2
Maneb
207­
08­
9
Benzo(
k)
fluoranthene
8018­
07­
7
Mancozeb
205­
99­
2
Benzo(
b)
fluoranthene
121­
75­
5
Malathion
50­
32­
8
Benzo(
a)
pyrene
58­
89­
9
Lindane
(
gamma­
BHC)
56­
55­
3
Benz(
a)
anthracene
7439­
92­
1
Lead
17804­
35­
2
Benomyl
143­
50­
0
Kepone
1912­
24­
9
Atrazine
115­
32­
2
Kelthane
7440­
38­
2
Arsenic
193­
39­
5
Indeno(
1,2,3­
c,
d)
pyrene
120­
12­
7
Anthracene
26601­
64­
9
3,3',
4,4',
5,5'­
Hexachlorobiphenyl
61­
82­
5
Amitrole
118­
74­
1
Hexachlorobenze
584­
79­
2
Allethrin
1024­
57­
3
Heptachlor
epoxide
309­
00­
2
Aldrin
76­
44­
8
Heptachlor
116­
06­
3
Aldicarb
51630­
58­
1
Fenvalerate
15972­
60­
8
Alachlor
66230­
04­
4
Esfenvalerate
53­
96­
3
2­
Acetylaminofluorene
72­
20­
8
Endrin
83­
32­
9
Acenaphthene
CAS
No
Compound
CAS
No
Compound
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
20
137­
30­
4
Ziram
115­
29­
7
Endosulfan
12122­
67­
7
Zineb
131­
16­
8
Dipropyl
phthalate
50471­
44­
8
Vinclozalin
131­
18­
0
Di­
n­
pentyl
phthalate
76­
87­
9
Triphenyl
tin
hydroxide
84­
74­
2
Di­
n­
butyl
phthalate
900­
95­
8
Triphenyl
tin
acetate
593­
74­
8
Dimethyl
mercury
1582­
09­
8
Trifluralin
84­
75­
3
Dihexyl
phthalate
56­
35­
9
Tributyltin
chloride
117­
81­
7
Diethylhexyl
phthalate
7440­
31­
5
Tin
103­
23­
1
Diethylhexyl
adipate
8001­
35­
2
Toxaphene
84­
66­
2
Diethyl
phthalate
32598­
13­
3
3,3'
4,4'­
Tetrachlorobiphenyl
60­
57­
1
Dieldrin
CAS
No
Compound
CAS
No
Compound
1746­
01­
6
2,3,7,8­
TCDD
84­
61­
7
Dicyclohexyl
phthalate
93­
76­
5
2,4,5­
T
115­
32­
2
Dicofol
100­
42­
5
Styrene
120­
83­
2
2,4­
Dichlorophenol
122­
34­
9
Simazine
96­
12­
8
1,2­
Dibromo­
3­
chlorpropane
129­
00­
0
Pyrene
50­
29­
3
p.
p'­
DDT
85­
01­
8
Phenanthrene
72­
55­
9
p,
p'­
DDE
52645­
53­
1
Permethrin
72­
5­
8
p,
p'­
DDD
87­
86­
5
Pentachlorophenol
94­
75­
7
2,4­
D
82­
68­
8
Pentachloronitrobeneze
52315­
07­
8
Cypermethrin
25429­
29­
2
3,3'
4,4',
5­
Pentachlorobiphenyl
13121­
70­
5
Cyhexatin
56­
38­
2
Parathion
(
ethyl)
7440­
50­
8
Copper
27304­
13­
8
Oxychlordane
218­
01­
9
Chrysene
29082­
74­
4
Octachlorostyrene
2921­
88­
2
Chlorpyrifos
25154­
52­
3
p­
Nonylphenol
1897­
45­
6
Chlorothanlonil
39765­
80­
5
trans­
Nonachlor
57­
74­
9
Chlordane
(
tech)
99­
99­
0
4­
Nitrotoluene
63­
25­
2
Carbaryl
1836­
75­
5
Nitrofen
7440­
43­
9
Cadmium
2385­
85­
5
Mirex
85­
68­
7
Butyl
benzyl
phthalate
Table
2.
Purity
Assays
by
GC­
FID
and
Elemental
Analysis
of
Some
Suspected
EDCs
­­­
2.53
4.91
3.12
­­­
50.77
41.31
50.39
5.47
2.53
4.98
3.01
71.66
50.61
41.75
49.40
5.51
2.48
4.84
3.17
71.63
50.73
41.24
50.38
­­­
99.1
99.8
99.6
99.9
95.1
99.7
99.7
Carbaryl
Nitrofen
Parathion
Vinclozolin
%
H
%
C
%
H
%
C
%
H
%
C
B2
A1
Compound
Found
B2
Found
A1
Theoretical
GC­
FID
(%
Purity)

1Analysis
before
purification
2Analysis
after
purification
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
21
Table
3.
Available
Stable­
Isotope
Labeled
Compounds
for
EDC
Analysis
p,
p'­
DDT­
13
C
12
3,3',
4,4'­
Tetrachlorobiphenyl­
13
C
12
p,
p'­
DDE­
13
C
12
Styrene­
D
8
p,
p'­
DDD­
D
8
Pyrene­
13
C
3
2,4­
D­
13
C
6
Phenanthrene­
13
C
6
Chrysene­
13
C
6
Pentachlorophenol­
13
C
6
Chlordane­
13
C
1
(
random)
Pentachloronitrobenzene­
13
C
6
Butyl
benzyl
phthalate­
D
4
3,3',
4,4',
5­
Pentachlorobiphenyl­
13
C
12
Bisphenol­
A­
13
C
12
Parathion­
D
10
Bis(
2­
ethylhexyl)
phthalate­
D
4
4­
Nitrotoluene­
13
C
6
g
­
BHC­
13
C
6
[
Lindane­
13
C
6]
Mirex­
13
C
8
(
±
)
­
a
­
BHC­
13
C
6
Kepone
®
­
13
C
8
b
­
BHC­
13
C
6
Indeno(
1,2,3­
c,
d)
pyrene­
13
C
6
Benzophenone­
D
10
3,3',
4,4',
5,5'­
Hexachlorobiphenyl­
13
C
12
Benzo(
k)
fluoranthene­
13
C
6
Hexachlorobenzene­
13
C
6
Benzo(
b)
fluoranthene­
13
C
6
Heptachlor
epoxide­
13
C1
Benzo(
a)
pyrene­
13
C
4
Heptachlor­
13
C
4
Benz(
a)
anthracene­
13
C
6
Endosulfan
I­
D
4
(
a
isomer)
Atrazine­
13
C
3
Diethyl
phthalate­
D
4
Anthracene­
13
C
6
Dieldrin­
13
C
4
Aldrin­
13
C
4
2,4­
Dichlorophenol­
13
C
6
Alachlor­
13
C
6
Di­
n­
butyl
phthalate­
D
4
Acenaphthene­
13
C
6
Figure
1.
Mass
Spectra
of
cis­
and
trans­
Chlordane
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
22
Figure
2.
60
mHz
1H
NMR
of
cis­
and
trans­
Chlordane
Figure
3.
Timeline
for
Development
of
Isotope
Dilution
Techniques
for
Trace
Dioxin
Analysis
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
23
DEVELOPING
A
METHOD
USED
TO
SCREEN
FOR
MORE
THAN
400
PESTICIDES
AND
ENDOCRINE
DISRUPTERS
Philip
L.
Wylie,
Sr.
Applications
Chemist
and
Bruce
D.
Quimby,
Sr.
Applications
Chemist
Little
Falls
Analytical
Division,
Hewlett­
Packard
Co.,
2850
Centerville
Rd.,
Wilmington,
Delaware
19808­
1610
ABSTRACT
A
gas
chromatographic
(
GC)
method
has
been
developed
that
can
be
used
to
screen
for
more
than
400
pesticides
and
suspected
endocrine
disrupters.
In
principle,
it
can
be
used
to
screen
for
any
GC­
amenable
pesticide,
metabolite,
or
endocrine
disrupter.
The
method
relies
on
a
new
technique
called
retention
time
locking,
a
procedure
that
allows
the
chromatographer
to
reproduce
analyte
retention
times
independent
of
GC
system,
column
length,
or
detector
so
long
as
columns
with
same
stationary
phase,
nominal
phase
ratio
and
diameter
are
used.
Because
retention
time
locking
increases
retention
time
precision
and
predictability,
raw
retention
times
can
be
used
as
a
more
reliable
indicator
of
compound
identity.
The
chromatographer
first
locks
the
GC
method
so
that
all
retention
times
match
those
listed
in
a
408­
compound
pesticide
retention
time
database.
Using
GC/
AED
or
other
element­
selective
detectors,
the
analyst
enters
a
peak's
retention
time
and
known
elemental
content
(
presence
or
absence
of
heteroatoms)
into
a
dialog
box.
Software
then
searches
the
pesticide
database
for
those
compounds
that
elute
at
the
correct
retention
time
and
have
the
right
elemental
content.
The
software
usually
finds
from
one
to
five
compounds
that
meet
these
criteria.
Confirmation
is
performed
by
GC/
MS
or
by
calculation
of
elemental
ratios
using
GC/
AED
data.
With
retention
time
locking,
pesticides
have
the
same
retention
time
on
all
GC
systems;
this
makes
GC/
MS
confirmation
much
easier
because
the
analyte'
s
retention
time
is
already
known.

INTRODUCTION
More
than
700
pesticides
are
currently
registered
for
use
in
the
world1
and
many
more
continue
to
persist
in
the
environment,
even
after
being
banned.
For
the
protection
of
human
health
and
the
environment,
acceptable
limits
in
food
and
water
have
been
set
by
governmental
bureaus
such
as
the
United
States
Environmental
Protection
Agency
(
USEPA)
and
the
Codex
Alimentarius
Commission2.
Numerous
methods
have
been
developed
to
screen
for
pesticide
contamination
in
food3­
7
and
the
environment8­
10
to
ensure
that
these
standards
are
met.

Certain
pesticides
and
other
synthetic
chemicals
have
been
suspected
of
behaving
as
pseudo
hormones,
disrupting
normal
functions
of
the
endocrine
system
in
wildlife
and
humans.
Maladies
such
as
birth
defects,
behavioral
changes,
breast
cancer,
lowered
sperm
counts,
and
reduced
intelligence
have
been
blamed
on
exposure
to
endocrine
disrupters11.
The
1996
publication
of
Our
Stolen
Future,
a
book
by
Colborn,
Dumanoski,
and
Myers11,
brought
these
concerns
to
the
attention
of
the
public.
Recently­
passed
legislation
in
the
US
calls
for
more
testing
of
suspected
endocrine
disrupters
and
monitoring
of
them
in
food12
and
water13
supplies.
In
order
to
facilitate
more
research
into
the
endocrine
disrupter
issue,
methods
are
needed
to
detect
suspected
compounds
at
trace
levels.

Because
so
many
pesticides
are
in
use,
it
is
usually
impractical
to
screen
for
large
numbers
of
them
individually
and,
therefore,
multiresidue
methods
are
preferred.
Most
laboratories
that
analyze
for
pesticides
in
foods,
screen
for
only
a
few
dozen
compounds
because
it
is
often
very
difficult
to
screen
for
more.
Recently
however,
methods
have
been
developed
using
gas
chromatography
with
mass
spectral
detection
(
GC/
MS),
that
can
screen
for
more
than
200
4
or
even
300
5
pesticide
residues.

Still,
there
is
no
universal
method
to
analyze
for
all
GC­
amenable
pesticides.
While
GC/
MS
methods
are
gaining
in
popularity,
there
are
still
some
limitations.
When
methods
employ
selected
ion
monitoring
(
SIM)
or
tandem
mass
spectrometry
(
MS/
MS),
method
development
is
more
tedious
and
any
shift
in
GC
retention
times
requires
that
individual
analyte
retention
time
windows
be
shifted
accordingly.
These
methods
are
only
capable
of
detecting
compounds
on
the
target
list,
there
are
still
hundreds
of
pesticides,
metabolites,
and
suspected
endocrine
disrupters
that
could
be
missed.
On
the
other
hand,
methods
based
on
scanning
GC/
MS
alone
require
more
sample
cleanup
to
avoid
interferences
from
coextracted
indigenous
compounds.
Typically,
these
methods
do
not
screen
for
many
pesticide
metabolites,
endocrine
disrupters,
or
other
environmental
contaminants.
A
method
that
could
be
used
to
screen
for
endocrine
disrupters
and
almost
all
of
the
volatile
pesticides
and
metabolites
would
offer
a
better
means
of
monitoring
the
food
supply
and
the
environment.
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
24
This
paper
describes
a
universal
method
that,
in
principle,
could
be
used
to
screen
for
any
pesticide,
metabolite,
or
endocrine
disrupter
that
can
be
eluted
from
a
gas
chromatograph.
As
a
first
test
of
the
concept,
a
method
was
developed
to
screen
for
408
pesticides.
The
method
is
being
expanded
to
include
virtually
all
of
the
volatile
pesticides,
metabolites,
and
suspected
endocrine
disrupters.
The
screening
procedure
relies
on
a
new
gas
chromatographic
technique
called
"
retention
time
locking"
14
with
database
searching
based
on
retention
time
and
elemental
content.
This
technique
is
used
to
narrow
an
analyte's
identity
to
a
few
possibilities.
Confirmation
is
performed
by
GC/
MS
or
by
calculation
of
a
compound's
elemental
ratio
using
GC
with
atomic
emission
detection
(
GC/
AED).

EXPERIMENTAL
Samples
Fruit
and
vegetable
extracts
were
obtained
from
the
Florida
Department
of
Agriculture
and
Consumer
Services
(
Tallahassee)
and
from
the
Canadian
Pest
Management
Regulatory
Agency,
Laboratory
Services
Subdivision
(
Ottawa).
Samples
from
Florida
were
extracted
using
the
Luke
procedure15­
17
while
those
from
Canada
were
prepared
according
to
the
method
described
by
Fillion,
et
al.
5
Instrumentation.
Table
1
lists
the
instrumentation
and
chromatographic
conditions
used
for
GC/
AED
screening
and
GC/
MS
confirmation.

Table
1.
Instrumentation
and
conditions
of
analysis.

14
psi
(
constant
pressure)
Inlet
pressurea
2
µ
L
Injection
volume
Split/
splitless,
250
°
C
Inlet
30
m
X
0.25
mm
X
0.25
µ
m
HP­
5MS
Column
HP
G1701AA
Version
A.
03.00
running
on
MS
Windows
95
Software
HP
Vectra
XU
6/
200
Computer
for
data
acquisition
&
analysis
HP
5973
MSD
Mass
spectral
detector
HP
6890
Series
Automatic
Sampler
Automatic
sampler
HP
6890
Gas
chromatograph
GC/
MS
System
Group
1:
C
496,
Cl
479,
Br
478
Group
2:
C
193,
S
181,
N
174
Group
3:
P
178
AED
elements
&
wavelengths
(
nm)
260
°
C
AED
cavity
temperature
260
°
C
AED
transfer
line
temperature
50
°
C
(
1.13
min),
30
°
C/
min
to
150
°
C
(
2
min),
3
°
C/
min
to
205
°
C
(
0
min),
10
°
C/
min
to
250
°
C
(
20
min)
Oven
temperature
program
60
psi
(
2.01
min),
10
psi/
min
to
27.9
psi
(
hold)
Inlet
pressure
program
(
pulsed
splitless)
a
27.6
psi,
constant
pressure
for
2­
µ
L
injections
Inlet
pressure
(
splitless)
a
2
µ
L
splitless
or
5
µ
L
pulsed
splitless
Injection
volumes
Split/
splitless,
250
°
C
or
280
°
C
GC
Inlet
30
m
X
0.25
mm
X
0.25
µ
m
HP­
5MS
Column
HP
G2360AA
GC/
AED
Software
running
on
MS
Windows
3.11
Software
HP
Vectra
XU
Series
4
5/
150
Computer
for
data
acquisition
&
analysis
HP
G2350A
Atomic
Emission
Detector
Atomic
Emission
detector
HP
6890
Series
Automatic
Sampler
Automatic
sampler
HP
6890
Gas
chromatograph
GC/
AED
System
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
25
Transfer
line
=
280
°
C,
MS
quad
=
150
°
C,
MS
source
=
230
°
C
Temperatures
Scan
(
35­
550
amu)
Acquisition
mode
MSD
Parameters
Same
as
GC/
AED
Oven
temperature
program
a)
The
column
head
pressures
shown
are
typical
values.
Exact
values
were
determined
as
part
of
the
retention
time
locking
procedure.

Software
for
Method
Translation
Software
for
use
in
translating
the
GC
method
from
one
GC
column
to
another
column
(
same
phase
but
different
dimensions)
was
obtained
from
Hewlett­
Packard
Co.
(
Wilmington,
DE);
the
software
is
available
on
the
world
wide
web
at
the
following
address:
http://
www.
dmo.
hp.
com/
apg/
servsup/
usersoft/
main.
html
RESULTS
AND
DISCUSSION
Retention
Time
Locking
Key
to
the
development
of
this
method
is
a
new
concept
in
gas
chromatography
called
retention
time
locking.
This
is
a
procedure
that
allows
the
chromatographer
to
match
analyte
retention
times
from
run
to
run,
independent
of
the
GC
system,
detector,
or
manufacturing
variations
in
column
dimensions;
the
only
requirement
is
that
the
columns
used
have
the
same
stationary
phase
and
the
same
nominal
diameter
and
phase
ratio.
For
example,
with
retention
time
locking,
it
is
possible
to
match
analyte
retention
times
on
a
GC/
AED
and
a
GC/
MS
even
though
the
column
outlet
pressures
are
much
different:
1.5
psi
above
ambient
pressure
for
the
AED
and
vacuum
for
the
MSD.
The
procedure
also
compensates
for
differences
in
GC
column
length
resulting
from
variations
in
manufacturing
or
from
column
cutting
required
during
routine
maintenance.

Retention
time
locking
is
accomplished
by
adjusting
the
GC
column
head
pressure
until
a
given
analyte,
such
as
an
internal
standard,
has
the
required
retention
time.
When
this
is
done,
all
other
analyses
in
the
chromatogram
will
have
the
correct
retention
times
as
well.
Software
has
been
developed
that
can
be
used
to
determine
the
column
head
pressure
that
will
correctly
lock
the
retention
times
after
a
single
"
scouting''
run.

With
retention
time
locking,
it
is
possible
to
measure
pesticide
retention
times
using
a
given
GC
method
and
then
reproduce
those
retention
times
in
subsequent
runs
on
the
same
or
different
instruments.
With
this
increased
retention
time
precision
and
predictability,
raw
retention
times
become
a
far
more
useful
indicator
of
analyte
identity.
For
many
years
relative
retention
times3,6
or
retention
indices18­
19
have
been
used
to
identify
compounds;
these
techniques
were
developed
to
compensate
for
the
fact
that
retention
times
were
not
generally
predictable
from
day
to
day,
column
to
column,
or
instrument
to
instrument.
With
modern
instrumentation
and
retention
time
locking,
it
seemed
that
raw
retention
times
could
be
used
for
compound
identification
in
much
the
same
way
that
retention
indices
have
been
used
in
the
past,
albeit
with
much
less
effort.
The
chromatographer
could
simply
scan
a
table
of
pesticide
retention
times,
eliminating
all
possibilities
but
those
with
close
elusion
times
under
the
same
locked
GC
conditions.

Pesticides
almost
always
contain
heteroatoms
and
often
have
several
in
a
single
molecule;
the
most
frequently
encountered
heteroatoms
are
O,
P,
S,
N,
Cl,
Br,
and
F.
GC/
AED
has
been
shown
to
be
a
useful
tool
for
pesticide
screening
because
it
is
selective
for
all
of
the
elements
found
in
these
compounds20­
22.
Thus,
GC/
AED
screening
provides
valuable
information
about
the
elemental
content
of
an
unknown
molecule.
By
including
this
elemental
information
along
with
the
retention
time,
it
should
be
possible
to
narrow
pesticide
"
hits"
to
just
a
few
possibilities.

Pesticide
Retention
Time
Table
To
test
this
pesticide
screening
concept,
a
table
of
pesticide
retention
times
and
molecular
formulas
was
required.
Stan
and
Linkerhagner7
recently
published
a
list
of
408
pesticides
with
their
molecular
formulas
and
their
GC/
AED
retention
times
using
a
25
m
X
0.32
mm
X
0.17
µ
m
HP­
5
column.
While
their
retention
time
table
and
GC
method
could
have
been
used
as
published,
their
column
was
not
an
ideal
choice
for
GC/
MS.
Therefore,
the
GC
method
and
all
of
the
pesticide
retention
times
were
translated
to
a
30
m
X
0.25
mm
X
0.25
µ
m
HP­
5MS
column
which
could
be
used
for
both
GC/
AED
screening
and
GC/
MS
confirmation.
These
conversions
were
made
using
software
for
GC
method
translation
developed
by
Blumberg23­
24.
Further
adjustments
to
the
408
retention
times
were
made
by
curve
fitting
the
actual
retention
times
for
60
known
compounds
and
applying
the
correction
to
the
table.

Pesticide
Screening
Method
Figure
1
diagrams
the
pesticide
screening
method.
First,
retention
time
locking
was
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
26
used
to
match
GC/
AED
and
GC/
MS
analyte
retention
times
to
those
listed
in
the
translated
pesticide
database.
Prototype
software
for
retention
time
locking
was
used
to
determine
the
column
head
pressure
needed
to
produce
a
retention
time
of
25.216
min
for
p,
p'­
DDE.
Using
the
GC/
AED,
element
selective
chromatograms
were
obtained
for
C,
Cl,
Br,
N,
S,
P,
and
sometimes
F
and
I.
Prototype
software
was
then
used
to
search
the
database
by
retention
time
and
elemental
content.

Figure
1.
Diagram
of
the
pesticide
screening
method
that
uses
retention
time
locking
and
retention
time
database
searching
Figure
2
is
a
screen
capture
from
this
software
showing
the
dialog
box
used
to
input
the
search
criteria.
One
must
choose
a
search
time
window
wide
enough
to
be
sure
to
include
the
correct
analyte,
but
narrow
enough
to
eliminate
as
many
extraneous
"
hits"
as
possible.
A
value
of
0.8
min
was
chosen
because
tests
with
several
dozen
compounds
showed
that,
under
locked
condi­
tions,
pesticide
retention
times
always
fell
within
+/­
0.3
min
of
the
tabulated
value.
This
time
window
would
be
smaller
if
one
were
to
use
a
database
generated
on
the
same
column
under
locked
conditions.
Of
course,
a
narrower
time
window
would
generate
fewer
hits
and
a
more
accurate
screening
method.

Figure
2.
Dialog
box
used
in
the
pesticide
database
searching
software.

From
the
GC/
AED
chromatograms
it
is
usually
possible
to
determine
which
heteroatoms
are
present
or
absent
in
the
suspected
pesticide
peak.
All
available
information
is
added
to
the
dialog
box
and
this
is
used
to
focus
the
search
on
those
pesticides
that
fall
within
the
retention
time
window
and
have
the
specified
elemental
content.
The
search
produces
a
list
of
pesticides
that
meet
these
criteria.

Confirmation
is
usually
done
by
GC/
MS
under
locked
conditions
so
that
all
GC/
MS
retention
times
match
the
GC/
AED
values.
Alternatively,
when
there
is
adequate
signal
to
quantitate
the
analyte
in
multiple
AED
element­
selective
chromatograms,
it
is
often
possible
to
confirm
a
pesticide's
identity
simply
by
calculating
its
heteroatom
ratio.
GC/
AED
software
for
element
ratioing
facilitates
this
procedure.
Although
not
yet
tried,
it
should
also
be
possible
to
use
a
second
column
with
a
different
phase
for
confirmation.
This
would
require
developing
a
new
retention
time
database
on
the
confirmation
column,
but
that
is
not
yet
available.
Figure
3
shows
a
set
of
GC/
AED
element­
selective
chromatograms
obtained
for
a
strawberry
extract.
Peaks
in
the
S,
N,
P,
and
Cl
chromatograms
suggest
the
presence
of
several
pesticides.
The
peak
at
22.031
min
contains
S,
N,
and
Cl
but
does
not
appear
to
have
any
P,
F,
Br,
or
I.
When
the
database
was
searched
only
on
the
basis
of
the
peak's
retention
time
using
a
0.8
min
window,
20
possibilities
were
reported
(
Figure
4).
However,
when
the
elemental
information
was
included
in
the
search,
only
three
of
the
408
pesticides
in
the
database
met
the
criteria
(
Figure
5).
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
27
Figure
3.
GC/
AED
elementselective
chromatograms
obtained
for
a
strawberry
extract.
The
GC/
AED
method
was
locked
to
the
pesticide
database.

Figure
4.
Database
search
results
for
the
peak
at
22.031
min
(
Figure
3).
The
database
search
used
only
retention
time;
no
elemental
information
was
entered
into
the
search
dialog
box.

Figure
5.
Database
search
results
for
the
peak
in
the
strawberry
extract
found
at
22.031
min
(
Figure
3)
when
both
retention
time
and
elemental
content
were
entered
into
the
dialog
box
as
shown
in
Figure
2
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
28
Confirmation
was
first
done
by
calculating
the
Cl:
N:
S
ratio
in
the
molecule
using
the
GC/
AED
elemental
ratioing
software.
Using
chlorpyrifos
(
C9H11Cl3NO3PS)
as
the
element­
specific
calibration
standard,
the
Cl:
N:
S
ratio
in
the
unknown
peak
was
calculated
to
be
3.07:
0.95:
1.00.
Though
approximate,
this
heteroatom
ratio
is
only
consistent
with
captan
(
Figure
5).

The
same
sample
was
then
analyzed
by
GC/
MS
under
locked
conditions
so
that
all
suspect
pesticides
would
have
GC/
MS
retention
times
very
close
to
their
GC/
AED
values.
Captan
was
found
at
21.979
min,
just
0.052
min
away
from
its
GC/
AED
retention
time
and
close
to
the
database
value
of
22.23
min
(
Figure
6).
With
retention
time
locking,
it
was
possible
to
make
the
GC/
AED,
GC/
MS,
and
database
retention
times
all
agree
to
within
0.25
min.
This
made
it
much
easier
to
find
the
suspected
pesticide
in
the
total
ion
current
chromatogram
(
TIC),
because
the
compound's
retention
time
was
already
known.
Moreover,
since
the
possibilities
had
been
narrowed
to
just
three
compounds,
their
characteristic
ions
could
be
extracted
to
reduce
the
background
contribution
from
coextracted
indigenous
materials.

Figure
6.
Scanning
GC/
MS
analysis
of
the
strawberry
extract
shown
in
Figure
3.
The
GC/
MS
method
was
locked
so
that
retention
times
would
match
both
the
pesticide
database
and
the
GC/
AED.
Captan
was
identified
at
21.979
min.
Its
retention
times
in
the
GC/
AED
and
pesticide
database
were
22.031
min
and
22.23
min,
respectively.

Figure
7
shows
S­,
N­,
P­,
and
Cl­
selective
chromatograms
of
an
orange
extract.
When
the
database
was
searched
for
the
chlorine­
containing
compound
labeled
1,
two
possibilities
were
listed
­
aldrin
and
PCB­
152
(
Table
2).
However
GC/
MS
could
not
confirm
either
of
these
possibilities
nor
were
any
other
compounds
suggested.
Several
explanations
are
possible
for
this
discrepancy:
a)
the
compound
is
not
a
pesticide
or
PCB,
b)
the
compound
is
a
pesticide
or
metabolite
that
does
not
appear
in
either
the
retention
time
or
mass
spectral
databases.
or
c)
one
or
both
of
the
suggested
compounds
is
not
contained
in
the
mass
spectral
libraries.

Figure
7.
GC/
AED
element­
selective
chromatograms
for
an
orange
extract.
The
GC/
AED
method
was
locked
to
the
pesticide
database
retention
times.
Peaks
labeled
1­
4
are
identified
in
Table
2.
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
29
Table
2.
Pesticides
initially
suggested
by
searching
the
pesticide
database
on
the
basis
of
retention
times
and
elemental
content.
Identities
as
determined
by
GC/
MS
are
shown
along
with
pesticide
concentrations
determined
using
compound­
independent
calibration.
Peak
numbers
refer
to
peaks
labeled
in
Figure
7.

2.28
Thiabendazole
(
not
in
database)
Thiophenate­
methyl
Thiophenate­
ethyl
Dimethametrynel
21.717
4
0.26
Chlorpyrifos
Chlorpyrifos
Dicapthon
19.957
3
0.8
Malathion
Malathion
19.519
2
0.6
(
Cl
only)
not
found
Aldrin
PCB­
152
19.152
1
Concentration
Using
CIC
(
ppm)
MS
Confirmation
Database
Search
Results
Using
AED
Element
Information
GC/
AED
Retention
Time
Peak
Number
Malathion
(
Table
2)
was
the
only
choice
suggested
for
peak
2
(
Figure
7)
and
this
was
easily
confirmed
by
MS
library
searching.
On
the
basis
of
retention
time
and
elemental
content,
chlorpyrifos
and
dicapthon
were
suggested
for
peak
3.
Chlorpyrifos
was
confirmed
by
GC/
MS.
Three
possibilities
were
offered
for
peak
4
­
thiophenate­
methyl,
thiophenate­
ethyl,
and
dimethametrynel.
However,
at
that
retention
time,
GC/
MS
found
thiabendazole.
Thiabendazole
was
not
in
the
original
retention
time
database
and
this
is
why
it
was
not
given
as
a
possibility.
The
database
searching
software
allows
one
to
add
or
edit
entries,
so
thiabendazole
was
added
to
list.
Even
though
thiabendazole
was
not
in
the
original
retention
time
database,
it
was
still
much
easier
to
find
and
identify
this
compound
in
the
TIC
because
retention
time
locking
ensured
that
its
GC/
MS
and
GC/
AED
retention
times
were
nearly
the
same.

Compound­
independent
calibration
(
CIC)
is
a
GC/
AED
technique
that
allows
one
to
use
a
single
analyte
as
a
calibration
standard
for
all
others
that
contain
the
same
elements.
It
is
the
first
step
in
calculating
elemental
ratios.
Using
chlorpyrifos
as
the
element­
selective
calibration
standard,
the
concentrations
of
malathion,
chlorpyrifos,
and
thiabendazole
in
the
orange
extract
were
determined
(
Table
2).
Even
though
peak
1
could
not
be
identified,
CIC
could
still
be
used
to
determine
the
concentration
of
Cl
(
0.6
ppm).
For
regulatory
purposes
this
might
be
very
useful.
If
the
Cl
level
is
found
to
be
very
low
further
investigation
may
not
be
required;
however,
if
the
Cl
level
is
high,
it
may
be
necessary
to
work
harder
at
compound
identification.

SUMMARY
Most
pesticide
screening
procedures
are
capable
of
finding
only
a
fraction
of
the
pesticides
that
are
registered
for
use.
This
procedure
has
the
capability
of
screening
for
virtually
any
volatile
pesticide,
metabolite,
or
endocrine
disrupter.
Although
confirmation
is
usually
required,
GC/
MS
analysis
is
made
much
easier
and
more
reliable
because
the
pesticide's
retention
time
is
already
known.

The
pesticide
database
used
to
demonstrate
the
feasibility
of
this
method
was
originally
developed
by
Stan
and
Linkerhagner
on
a
different
gas
chromatograph
(
HP
5890
Series
II),
using
a
different
column,
and
without
the
benefit
of
retention
time
locking.
Experience
in
this
laboratory
shows
that
retention
time
locking
and
the
new
generation
of
electronic
pneumatic
control
should
allow
one
to
prepare
a
database
with
even
more
precise
retention
times.
Moreover,
retention
time
locking
would
allow
chromatographers
to
match
those
retention
times
with
far
more
accuracy
than
was
routinely
possible
in
the
past.
By
creating
a
database
from
the
beginning
on
the
preferred
GC
column
under
locked
conditions,
it
should
be
possible
to
narrow
the
required
search
window
which
would
result
in
fewer
hits
and
greater
accuracy.
A
project
is
underway
in
this
laboratory
to
create
such
a
database
on
two
different
GC
columns
that
will
contain
several
hundred
volatile
pesticides,
metabolites,
and
suspected
endocrine
disrupters.

While
GC/
AED
has
been
shown
to
be
an
ideal
tool
for
element­
selective
pesticide
screening20­
22,
many
laboratories
rely
on
a
combination
of
other
selective
detectors.
It
should
still
be
possible
to
apply
this
method
if
each
GC
system
runs
the
same
method
under
the
same
locked
conditions.
Whatever
elemental
data
that
is
available
could
be
entered
into
the
search
dialog
box.
It
may
often
be
possible
to
identify
a
pesticide
by
using
two
different
GC
columns
each
configured
with
the
same
type
of
element­
selective
detector.
For
example,
a
pair
of
flame
photometric
detectors
could
be
used
to
isolate
organophosphorus
pesticides
on
two
different
GC
columns.
Two
databases
could
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
30
be
searched
and
only
those
compounds
that
appear
in
both
lists
would
be
possibilities.
Future
work
in
this
laboratory
will
test
this
possibility.

Retention
time
locking
with
database
searching
could
easily
be
applied
to
similar
types
of
analyses.
For
example,
one
might
use
the
procedure
to
identify
polychlorinated
biphenyls,
polynuclear
aromatics,
or
flavor
and
fragrance
compounds.

ACKNOWLEDGMENTS
The
authors
wish
to
thank
Joanne
Cook
and
Julie
Fillion
for
supplying
extracts
used
in
the
development
and
testing
of
this
method.
The
authors
also
wish
to
thank
Matthew
Klee
and
Leonid
Blumberg
for
many
useful
discussions.

REFERENCES
1.
Tomlin,
Clive,
ed
(
1994)
The
Pesticide
Manual,
Tenth
Edition.
British
Crop
Protection
Council,
Surry,
UK
2.
Miller,
R.
W.
This
is
Codex
Alimentarius
Secretariat
of
the
Joint
FAO/
WHO
Food
Standards
Programme,
Food
and
Agriculture
Organization
of
the
United
Nations,
Rome
3.
McMahon,
B.
M.
&
Hardin,
N.
F.
eds.
(
1994)
Pesticide
Analytical
Manual,
Vol.
I,
Third
Edition
U.
S.
Food
and
Drug
Administration,
Washington,
D.
C.
4.
Lee,
S.
M.,
Papathakis,
M.
L.,
Feng,
H.­
M.
C.,
Hunter,
G.
G.,
&
Carr,
J.
E.
(
1991)
Fresenius'
A.
Anal.
Chem.
339,
376­
383
5.
Fillion,
J.,
Hindle,
R.,
Lacroix,
M.,
&
Selwyn,
J.
(
1995)
J.
AOAC
Int.
78,
1252­
1266
6.
Working
Group
on
Development
and
Improvement
of
Residue­
analytical
Methods
(
1996)
Analytical
Methods
for
Pesticide
Residues
in
Foodstuffs
General
Inspectorate
for
Health
Protection,
Ministry
of
Public
Health,
Welfare
&
Sport,
The
Netherlands
7.
Luke,
M.
A.,
Froberg,
J.
E.,
Doose,
G
M.,
Masumoto,
H.
T.
(
1981)
J.
Assoc.
Off.
Anal.
Chem.
64,
1187­
1195
8.
Stan,
H.­
J.,
ed.
(
1995)
Analysis
of
Pesticides
in
Ground
and
Surface
Water
II
Springer­
Verlag,
Berlin,
Germany
9.
Wagner,
R.
E.,
Kotas,
W.,
&
Yogis,
G.
A.,
eds.
(
1994)
Guide
to
Environmental
Analytical
Methods,
2nd
Edition
Genium.
Schenectady,
NY
10.
U.
S.
Environmental
Protection
Agency
Test
Methods
for
Evaluating
Solid
Waste,
SW­
846,
Draft
Method
8085:
Pesticides
by
GC/
AED
11.
Colborn,
T.,
Dumanoski,
D.,
&
Myers,
J.
P.
(
1996)
Our
Stolen
Future
Penguin,
NY,
NY
12.
Food
Quality
Protection
Act
of
1996
Public
Law
104­
170,
Congressional
Record
pp.
H8127­
H8141
13.
Safe
Drinking
Water
Act
Amendments
of
1996
Public
Law
104­
182,
Congressional
Record
pp.
H9680­
H9700
14.
Wylie,
P.
L.
&
B.
D.
Quimby
Abstracts
of
the
1st
European
Pesticide
Residue
Workshop
Alkmaar,
The
Netherlands,
June
10­
12,
1996,
Paper
#
O­
023.
15.
Luke,
M.
A.,
Froberg,
J.
E.,
Doose,
G.
M.,
&
Masumoto,
H.
T.
(
1981)
J.
Assoc.
Off.
Anal.
Chem.
64,
1187­
1195
16.
Luke,
M.
A.,
&
Doose,
G.
M.
(
1983)
Bull.
Environ.
Contamin.
Toxicol.
30,
110­
116
17.
Sawyer,
L.
D.
(
1985)
J.
Assoc.
Off:
Anal.
Chem.
68,
64­
71
18.
Lee,
M.
L.,
Yang,
F.
J.,
&
Bartle,
K.
D.
(
1984)
Open
Tubular
Column
Gas
Chromatography,
Theory
and
Practice
Wiley,
NY,
NY,
pp.
217­
218
19.
Bicchi,
C.,
D'Amato,
A.,
&
Binello,
A.
(
1996)
J.
High
Resol.
Chromatogr.
19,
80­
84
20.
Lee,
S.
M.
&
Wylie,
P.
L.
(
1991)
J.
Agric.
Food
Chem.
39,
2192­
2199
21.
Wylie,
P.
L.
&
Oguchi,
R
J.
Chrom.
(
1990)
517
131­
142
22.
Olson,
N.
L.,
Carrell,
R.,
Cummings,
R.
K.,
&
Rieck,
R.
(
1994)
LC­
GC
12,
142­
154
23.
Snyder,
W.
D.
&
Blumberg,
L.
M.
Proceedings
of
the
Fourteenth
International
Symposium
on
Capillary
Chromatography,
Baltimore,
MD,
May
25­
29,
1992,
pp.
28­
38
24.
Hewlett
Packard
Co.
Software
for
Method
Translation,
Available
on
the
World
Wide
Web
at:
http://
www.
dmo.
hp.
com/
apg/
servsup/
usersoft/
main.
html
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
31
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
32
ADVANCED
ENVIRONMENTAL
MONITORING
RESEARCH
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
33
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
34
EPA'S
EXTRAMURAL
MONITORING
RESEARCH
PROGRAM
William
Stelz
Office
of
Research
and
Development,
U.
S.
Environmental
Protection
Agency,
Washington,
DC
For
Fiscal
Year
1996,
the
Environmental
Protection
Agency
(
EPA)
increased
funding
for
its
investigator­
initiated
research
grants
program
and
identified
research
on
field
analytical
methods
and
continuous
monitoring
methods
as
ones
that
are
important
to
the
future
of
EPA
and
therefore
worthy
of
funding.
The
purpose
of
the
program
is
to
advance
measurement
science
by
stimulating
research
on
radically
new
approaches
to
solving
environmental
monitoring
problems.
EPA
solicited
proposals
in
these
areas
from
the
academic,
state
government,
and
non­
profit
research
communities.
Approximately
78
proposals
were
received
In
the
first
phase
of
the
proposal
review
process,
EPA
convened
a
panel
of
outside
experts
representing
academia,
industry,
other
government
agencies,
and
other
research
institutions
to
evaluate
and
rate
the
proposals
based
on
originality
and
creativity
of
the
proposed
research,
training
and
demonstrated
experience
of
the
investigators,
availability
and
adequacy
of
the
facilities
and
equipment
to
conduct
the
proposed
work,
and
responsiveness
of
the
proposal
to
the
indicated
research
needs
that
were
set
forth
in
the
solicitation.
Only
those
proposals
that
receive
a
rating
of
excellent
or
very
good,
move
to
the
second
phase.

In
the
second
phase,
EPA
monitoring
experts
focused
on:
how
well
does
the
proposed
study
fit
into
EPA's
overall
research
strategy
or
regulatory
program
in
this
area;
did
the
study
have
the
potential
to
strengthen
the
scientific
basis
for
risk
assessmen/
risk
management
in
the
subject
area
by
addressing
uncertainties
in
the
supporting
science;
would
the
research
enhance
or
complement
EPA's
in­
house
research
program;
would
the
results
of
the
study
have
broad
applicability
or
impact
large
segments
of
the
population;
and
would
the
study
eventually
produce
data
or
methods
which
can
be
utilized
by
the
public,
states,
and
EPA
to
better
assess
or
manage
environmental
problems?

Based
on
the
results
of
these
reviews,
EPA
awarded
grants
to
11
researchers
for
a
total
of
$
2.5
M.
In
this
paper,
the
author
will
discuss
the
extramural
research
program
and
review
the
specific
to
be
conducted.
The
author
will
emphasize
research
dealing
with
potential
applicability
to
RCRA
and
CERCLA
monitoring.

INTRODUCTION
EPA,
as
a
regulatory
agency,
is
charged
with
protecting
human
health
and
the
environment
through
regulatory
programs
and
other
avenues.
In
order
to
effectively
carry
out
its
mission,
EPA
relies
on
the
application
of
sound
science
in
the
assessment
of
environmental
problems
and
solutions.

As
a
means
of
helping
the
nation's
academic
institutions
to
focus
the
creativity
of
their
physical,
biological
and
social
scientists
and
engineers
on
developing
new
environmental
methods,
techniques,
and
information,
Agency
has
undertaken
a
major
expansion
of
its
extramural
environmental
science
research
program.

The
FY
96
EPA
academic
and
not­
for­
profit
research
support
program
identified
nine
priority
areas
of
environmental
study:

w
Ecological
assessment
w
Exposure
of
children
to
pesticides
w
Air
quality
w
Analytical
and
Monitoring
Methods
w
Drinking
water
w
Environmental
fate
and
treatment
of
toxics
and
hazardous
wastes
w
Environmental
statistics
w
High­
performance
computing
w
Exploratory
research,
including
Early
Career
Research
Awards
Because
quality
analytical
and
monitoring
data
is
critical
to
the
decision
making
process,
in
this
age
of
increasingly
scarce
resources,
a
significant
challenge
that
the
environmental
community
faces
is
how
to
gather
the
needed
date
in
the
most
efficient
manner.
To
that
end,
the
Agency
identified
analytical
and
monitoring
science
as
an
important
area
of
research.
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
35
While
it
is
obvious
that
one
needs
accurate
monitoring
data
in
order
to
assess
the
quality
of
the
environment
and
the
effectiveness
of
Agency
programs,
the
importance
of
analytical
and
monitoring
methods
cuts
across
almost
all
of
the
areas
in
which
EPA
conducts
research.
For
instance,
ecological
assessment
cannot
be
successfully
undertaken
if
high
quality
data
are
not
available
on
the
concentrations
of
toxic
chemicals
in
the
environment.
Exposure
of
children
to
pesticides
cannot
be
adequately
determined,
and
therefore
not
properly
controlled,
unless
accurate
measurement
of
pesticide
concentrations
are
available.
Air
quality
cannot
be
assessed
or
improved
unless
there
is
the
capability
to
measure
reliably
the
current
and
future
pollutants
in
the
ambient
air,
as
well
as
emissions
that
affect
air
quality.
Consequently,
analytical
methods
are
the
hingepin
upon
which
many
of
EPA'
s
other
programs
pivot.

ANALYTICAL
MONITORING
PRIORITIES
The
primary
purpose
of
the
Program
is
to
advance
measurement
science
by
stimulating
research
on
radically
new
approaches
to
solving
real­
world
environmental
monitoring
problems.
To
that
end,
for
the
FY
1996
solicitation,
EPA
identified
two
general
areas
where
current
environmental
monitoring
technology
are
known
to
be
inadequate.

w
Field
portable
methodology,
and
w
Continuous
measurement
methods
In
both
of
these
areas,
there
is
a
critical
need
for
equipment
and
methods
that
are
inexpensive
to
procure,
operate,
and
maintain.
Additionally,
EPA
was
seeking
technologies
that
did
not
use
toxic
chemicals,
and/
or
did
not
generate
hazardous
waste.
Given
that
decision
making
can
only
be
as
good
as
the
quality
of
the
data
on
which
it
is
based,
the
methods
should
produce
measurement
data
for
which
EPA,
the
regulated
community,
and
the
public
can
have
a
high
degree
of
confidence
in
the
results.
EPA
favors,
in
this
Program,
proposals
that
are
new
and
innovative
rather
than
proposals
that
are
essentially
improvements
upon
existing
methodology.
However,
existing
methods
adapted
from
the
non­
environmental
arena
were
also
considered.

Field
Portable
Measurement
Techniques
­
There
is
a
serious
need
for
generating
analytical
data
as
close
to
real­
time
as
possible.
To
this
end,
methodology
that
can
develop
reliable
analytical
data
in
the
field
has
become
a
priority.
Having
data
available
immediately
allows
personnel
to
make
management
and
operating
decisions,
without
having
to
wait
days
to
weeks
for
laboratory
analysis.
Such
waits
frequently
necessitate
additional
site
visits,
and
the
repetition
of
complex
sampling
events
(
such
as
moblilizing/
demobilizing
for
the
boring
of
contaminated
soil
samples
or
stack
sampling).
Additionally,
the
development
of
analytical
data
in
the
field
will
save
the
cost
of
transporting
samples,
avoid
problems
with
holding
times,
and
alleviate
chain­
of­
custody
requirements.
Further;
if
analytical
data
obtained
in
the
field
appear
suspect,
additional
confirmatory
analysis
can
be
run
immediately.
To
make
field­
adaptable
monitoring
methods
reality
these
methods
must
use
equipment
that
is
relatively
light,
rugged,
and
will
yield
reliable
data
at
the
required
sensitivity
levels
quickly
(
preferably
within
4
hours).
Examples
of
field­
portable,
rapid
results
monitoring
technology
that
have
now
become
commercially
available
include
immunochemistry­
based
assays
(
e.
g.,
for
analyzing
for
petroleum
contaminants
or
solvents
in
soils),
and
hand­
held
x­
ray
fluorescence
spectrophotometers
for
wastewater
analyses.

Continuous
Measurement
Methods
­
Periodic
analyses
of
contaminants
provides
only
a
"
snapshot"
of
the
condition
of
the
sampled
stream.
Even
though
EPA­
approved
analytical
methods
applied
to
the
periodic
samples
can
provide
extremely
analytical
accurate
data,
there
is
no
guarantee
that
the
"
snapshot"
data
are
representative
of
discharges
that
take
place
continuously.
The
ability
to
monitor
continuously
would
also
allow
for
trend
analysis,
which
in
turn
might
allow
for
the
correlation
of
emissions
with
process
changes.
Specific
areas
where
EPA
believes
continuous
sampling
would
benefit
the
public
are:

w
analysis
of
organic
and
inorganic
contaminants
in
municipal
and
industrial
waste
water
to
support
the
design
of
appropriately
sized
treatment
facilities;
w
metal
emissions
(
particularly
mercury)
from
high
temperature
processes,
such
as
incinerators,

w
release
of
volatile
organic
compounds
from
complex
point
or
area
source,
such
as
tanks,
pipes,
valves,
landfills,
and
contaminated
soils,
w
concentrations
of
toxic
chemicals
in
the
ambient
air
near
hazardous
waste
sites
and
industrial
facilities,

w
deposition
or
emission
flux
of
toxic
air
pollutants,
especially
semi­
volatile
substances
that
exist
both
in
the
gas
phase
and
attached
to
particulate
matter,
and
w
the
mass
of
inhalable
particulate
matter
(
both
PM2.5
and
PM10),
including
semivolatile
organics,
ammonium
nitrate
(
i.
e.,
semivolatile
inorganics),
and
particle­
bound
water.
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
36
This
paper
will
next
discuss
the
research
grant
selection
process
and
review
those
grants
that
focus
on
field
portable
methodology
and
continuous
measurement
methods.

PROPOSAL
SELECTION
PROCESS
The
focus
of
the
grant
selection
process
is
to
ensure
that
the
research
that
is
funded
represents
both
high
quality
of
science
and
will
have
a
significant
positive
impact
on
the
nation's
environmental
programs
if
successful.
To
that
end,
EPA
follows
a
rigorous
protocol
for
selecting
award
recipients.
The
protocol
has
three
major
components.

In
the
first
phase,
experts
from
EPA's
Regulatory
Program
Offices,
Regions,
and
the
Office
of
Research
and
Development
identify
environmental
issues
or
health
problems
that
are
of
the
highest
priority
to
the
Agency.
These
issues
are
based
on
the
regulatory
and
strategic
research
needs
of
the
Agency
as
well
as
the
areas
that
the
Agency
feels
methodology
or
data
gaps
limits
the
Agency's
ability
to
address
the
issue
or
problem.
Based
on
this
premise,
a
technical
working
group
is
established
in
order
to
develop
a
grant
solicitation
for
response
that
represents
the
goals
and
needs
of
the
Agency;
as
in
the
case
of
this
paper,
the
area
of
concern
involves
a
grant
solicitation
that
focuses
on
Air,
RCRA
and
CERCLA
monitoring
problems.

Once
the
proposals
are
received
EPA
conducts
a
peer
review
by
convening
a
panel
of
outside
experts
to
evaluate
the
proposals.
These
experts
represent
academia,
industry
and
other
government
agencies
and
research
institutions.
They
evaluate
and
rate
the
proposals
based
on:

w
originality
and
creativity
of
the
proposed
project,

w
training
and
demonstrated
experience
of
the
investigators,

w
availability
and
adequacy
of
the
facilities
and
equipment
with
which
the
research
will
be
conducted,
and
w
responsiveness
of
the
proposal
to
the
indicated
research
needs
that
were
set
forth
in
the
solicitations
Only
those
proposals
that
are
deemed
excellent
or
very
good
by
the
outside
peer
review
pane!
receive
further
evaluation
The
next
step
in
the
selection
process
consists
of
an
in­
house
relevancy
review.
The
focus
of
this
review,
which
is
conducted
by
a
panel
of
appropriate
EPA
experts,
is
to
rate
the
proposals
on
their:

w
fit
into
EPA's
overall
research
strategy
or
regulatory
program,

w
potential
to
strengthen
the
scientific
basis
for
risk
assessment/
risk
management
by
addressing
uncertainties
in
the
supporting
science,
w
ability
to
enhance
or
complement
EPA's
in­
house
research
program,

w
applicability
to
broad
segments
of
the
population,
and
w
ability
to
produce
data/
information
that
can
be
utilized
by
the
public,
states,
and
the
federal
government
to
better
assess
and
manage
environmental
programs.

The
result
of
this
review
is
a
ranking
of
the
proposals
as
to
their
relevance
and
potential
to
have
a
significant
impact
on
environmental
monitoring.
At
this
point
an
assessment
is
made
as
to
the
total
amount
of
resources
that
is
available,
whether
any
of
the
highly
ranked
proposals
are
duplicative
of
one
another
or
of
already
ongoing
research
EPA
is
conducting
and
whether
the
requested
level
of
EPA
funding
is
reasonable.
Funds
are
then
allocated
to
the
proposals
in
order
of
their
final
ranking
after
taking
these
factors
into
account.

In
response
to
the
FY
1996
solicitation,
EPA
received
approximately
78
proposals
in
the
Analytical
and
Monitoring
Methods
research
category.
The
proposals
covered
several
relevant
areas
of
research.
Five
of
these
areas
that
received
a
significant
number
of
proposals
were:
sampling
(
14
proposals),
particle
monitoring
(
9),
biosensors
(
5),
spectroscopy
(
22),
and
electrochemistry
(
11).
The
outside
peer
review
was
conducted
by
a
panel
of
27
persons.
Based
on
their
review,
17
proposals
which
received
a
rating
of
either
"
excellent"
or
"
very
good"
were
then
submitted
to
the
EPA
panel
for
relevancy
review.
0f
the
17
"
very
good
to
excellent"
proposals,
for
FY
1996,
EPA
awarded
eleven
research
grants
for
analytical
and
monitoring
methods
totaling
about
$
3.2
million.
Most
of
the
grants
fund
research
that
will
take
several
years
to
complete.
Six
of
these
eleven
proposals
relate
directly
to
field
portable
methodology
and/
or
continuous
measurement
for
RCRA
and
CERCLA­
type
pollutants,
or
air
emissions
that
may
exist
at
RCRA/
CERCLA
sites.
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
37
If
one
looks
at
totals,
the
overall
success
rate
for
proposals
was
14%
in
the
analytical
and
monitoring
methods
category.
For
the
proposals
that
received
high
marks
from
the
peer
review
process,
the
success
rate
was
very
high
being
almost
65%;
for
instance
all
the
proposals
that
received
an
excellent
rating
were
funded.
The
6
proposals
dealing
with
RCRA
and
CERCLA­
related
concerns
will
be
discussed
next.

SUMMARY
OF
SIX
ENVIRONMENTAL
MONITORING
PROJECTS
1.
Development
of
a
New
Biosensing
System
for
Environmental
Monitoring
of
PCBs,
Heavy
Metals,
and
Atrazines,
in
Aqueous
Streams,
Using
Conductive
Electroactive
Polymer
(
CEP)­
Based
Sensing
(
Dr.
Omowunmi
Sadik
­
State
University
of
New
York
at
Binghamton)

Conventional
chromatographic­
based
approaches
to
immunoassay
screening
analyses
of
aqueous
contaminants
are
not
very
sensitive,
and
require
substantial
time
(
minutes
to
hours)
to
complete.
The
proposed
project
will
develop
a
CEP­
based
selective
methodology
to
give
rapid
(
i.
e.,
real­
time)
detection
of
PCBs,
heavy
metals,
and
atrazines,
with
the
potential
for
application
to
other
chemicals.
In
addition,
the
equipment
and
procedures
involved
are
purported
to
be
less
expensive,
more
compact,
and
more
user
friendly
than
current
immunoassay
techniques
The
principle
upon
which
this
technology
is
based
is
that
the
electronic
conductivity
of
CEP
materials
allows
electrical
potentials
to
be
applied.
Measurement
of
the
reaction
to
these
potentials
in
combination
with
enzyme­
based
biological
reagents
then
allows
the
chemical
properties
of
the
samples
to
be
determined.

This
project
proposes
to
develop
the
CEP­
based
monitoring
technique,
and
will
describe
the
application
of
the
technology
for
field
demonstration.

2.
Real­
Time
Trace
Detection
of
Elemental
Mercury
and
Its
Compounds
(
Dr.
Robert
Barat
­
New
Jersey
Institute
of
Technology)

Currently,
commercial
devices
are
not
sensitive
enough
to
detect
Mercury
(
Hg)
in
ambient
air
or
gaseous
emissions
at
all
levels
of
interest
(
1
­
1000
µ
g/
m3)
in
real­
time,
and
certainly
there
is
no
technology
that
will
rapidly
speciate
the
various
Hg
compounds.
This
project
will
research
the
capabilities
for
measurement
of
Hg
and
its
compounds,
using
a
novel
hybrid
instrument
combining
Doppler­
shifted
Resonant
Fluorescence
(
RF)
and
Photofragment
Fluorescence
(
PFF)
with
test
gases
expanded
in
a
supersonic
jet.
The
instrument
will
be
called
a
Supersonic
Jet
Spectrometer
(
SJS).
This
project
will
build
the
SJS,
and
evaluate
its
detection
capabilities
on
Hg­
containing
air
streams.
The
detection
limits,
range
of
linearity,
relative
accuracy,
and
response
time
will
be
evaluated.
If
performance
is
satisfactory,
the
SJS
can
subsequently
be
developed
into
a
relatively
low
cost
instrument
for
field
use.

3.
Field­
Usable
Compact
Capillary­
Based
Ion/
Liquid
Chromatographs.
Real­
Time
Gas/
Aerosol
Analyzers
(
Purnendu
Dasupta
­
Texas
Tech
University)

Currently,
ion
and
liquid
chromatography,
two
widely
used
techniques
for
analysis
of
organic
constituents
in
the
environment
(
aqueous
streams
as
well
as
atmospheric
samples),
remain
relegated
to
the
laboratory
due
to
lack
of
true
portability.
This
project
proposes
to
develop
a
briefcase­
sized
(<
20
pound)
packed
capillary
and
open
tubular
ion/
liquid
chromatographic
instrument,
including
suppressed
conductometric
and
optical
detectors,
as
well
as
accompanying
software
for
a
laptop
computer.
The
instrument
is
expected
to
be
far
more
sensitive
than
its
predecessors
(
in
the
sub­
ppt
range
for
gases
and
pg/
m3
range
for
most
aerosol
constituents).
The
project
will
construct
two
complete
systems,
and
will
be
field­
tested
for
pesticide
analysis
in
runoffs
from
cotton
fields
local
to
the
University.

4.
Rapid
Determination
of
Organic
Contaminants
in
Water
by
Solid
Phase
Microextraction
(
SPME)
and
lnfrared
Spectroscopy
(
IR)
(
David
Tilotta
­
University
of
North
Dakota)

The
objective
of
this
study
is
to
develop
a
simple,
sensitive,
and
rapid
method
for
the
field
determination
of
organic
contaminants
in
water.
The
method
will
combine
SPME
and
Fourier
transformed
IR
techniques.
Current
conventional
analytical
techniques
are
typically
time­
consuming,
labor­
intensive,
and
require
expensive
instrumentation.
Hence,
they
are
not
readily
field­
adaptable.
The
proposed
research
will
construct
a
reusable
aluminum
SPME
"
dip­
stick",
which
is
coated
with
a
sorbent
film,
such
as
poly(
dimethylsilaxane)
and
paraffin­
impregnated
poly(
butylene).
The
sorbent
is
expected
to
have
relatively
large
partitioning
coefficients
(
2
­
2000)
for
many
environmentally­
important
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
38
compounds.
The
dipstick
will
be
placed
in
the
contaminated
water,
and
will
selectively
partition
the
organics
into
the
coating.
Initial
studies
show
that
the
SPME
has
a
partitioning
time
as
fast
as
40
minutes,
and
that
the
combined
SPME/
IR
methodology
is
sensitive
(
ppm
­
ppb
range)
as
well
as
reproducible
(
RSDs
of
1
­
11%).
In
addition,
the
method
is
expected
to
cost
less
than
20c
per
film,
and
because
it
is
a
solid
phase
extraction,
will
not
generate
solvent
waste.

The
study
proposes
to
identify
5
to
10
solid
phases
for
use
with
the
dipstick,
and
will
test
the
methodology
on
volatile
organic
compounds,
crop
protection
chemicals,
and
PAHs.
The
proposed
method
will
then
be
compared
to
standard
methods
for
determining
target
organics
in
water,
and
will
then
be
applied
in
the
field,
at
actual
contaminant
sites
5.
Field
Determination
of
Organics
for
Soil
and
Sludge
Using
Sub­
Critical
Water
Extraction
Coupled
with
Solid
Phase
Extraction
(
Dr.
Steven
Hawthorne
­
(
University
of
North
Dakota)

Currently,
there
are
no
field­
portable
water­
based
extraction
methods
available
for
quantitative
analysis
of
both
polar
and
non­
polar
organics
from
soils
and
sludges.
Currently,
the
well­
known
sorption
methods,
solid
phase
extraction
(
SPE)
and
solid
phase
microextraction
(
SPME),
are
only
applicable
to
water
samples.
This
proposal
will
investigate
the
use
of
subcritical
water
(
up
to
250
°
C
and
only
a
few
bars
[
or
atmospheres]
pressure)
to
quantitatively
extract
organics
into
water
in
preparation
for
SPE
or
SPME.
Preliminary
work
suggests
that
subcritical
water
can
solubilize
from
soils
and
sludges
even
non­
polar
organics,
such
as
PAHs,
more
than
106­
fold
as
compared
to
relatively
low
temperature
water.
Additionally,
with
proper
internal
standards,
quantitative
results
for
PAHs,
PCBs,
and
aromatic
amines
have
been
obtained
with
total
sample
preparation
time
of
about
30
minutes
(
including
subcritical
extraction
from
the
solid
sample
and
SPE
or
SPME
sorption),
yielding
detection
limits
of
less
than
ppb­
Ievel
sensitivity.

The
study
proposes
to
further
investigate
the
use
of
subcritical
water
extraction
combined
with
SPE
and
SPME
for
the
rapid
and
quantitative
determination
of
organics
in
soils
and
sludges.
It
intends
to
optimize
the
selectivity
of
the
extractions
(
based
on
the
temperature
of
the
water
used
for
extraction,
and
on
SPE/
SPME
sorbent
selectivity).
It
will
compare
analysis
of
the
extracts
in
the
field
as
opposed
to
sending
the
SPE/
SPME
devices
back
to
a
laboratory,
and
will
also
compare
this
method
with
conventional
EPA
extraction/
analysis
methods
for
soils
and
sludges.

6.
Supercritical
Fluid
Chromatography
Directly
Coupled
to
Dynamic
Nuclear
Polarization
(
SFC/
DNP)
(
Harry
Dorn
­
Virginia
Polytechnic
Institute
and
State
University)

The
objective
of
this
research
will
be
to
develop
an
SFC/
DNP
instrument
which
will
be
used
for
analysis
of
nonvolatile
organic
and
inorganic
contaminants
in
soils
and
aqueous
sources.
The
technology
will
directly
couple
supercritical
fluid
chromatography
(
SFC)
with
dynamic
nuclear
polarization
(
DNP).
DNP
is
a
variant
of
nuclear
magnetic
resonance
(
NMR).
The
DNP
will
allow
much
greater
sensitivity
than
(
NMR).
In
ultimate
applications,
the
organic
and
inorganic
contaminants
will
be
collected
using
adsorbent
traps
(
Tenax,
XAD,
etc.).
The
extract
from
the
adsorbents
will
then
be
injected
onto
a
supercritical
fluid
(
carbon
dioxide)
chromatographic
column
with
a
DNP
detector.
The
technology
is
expected
to
allow
for
rapid,
continuous
or
nearly
continuous,
and
sensitive
analysis
of
most
contaminants,
including
chlorocarbons,
organophosphates,
pesticides,
and
petroleum
compounds.
The
method
will
also
allow
determination
of
molecular
structure
information,
such
as
isomer
identification.

CONCLUSION
The
Agency's
FY
1996
research
support
program
is
funding
a
significant
amount
of
research
with
the
potential
for
having
a
meaningful
positive
impact
on
the
nation's
ability
to
monitor
for
environmental
contamination.
In
FY
1997,
EPA
will
increase
its
overall
level
of
support
to
its
extramural
research
program.
While
the
FY
1997
solicitation
does
not
call
specifically
for
research
in
the
Analytical
and
Monitoring
Methods,
significant
levels
of
funding
for
measurement
research
are
contained
in
the
technical
areas
that
are
specified
in
the
FY
97
Research
Grants
Announcement.
For
example,
the
Ecosystem
Indicators
research
area
has
funding
available
for
projects
which
focus
on
the
development
of
monitoring
methods
for
sampling
and
measuring
the
ecosystem
indicators.

EPA
has
and
continues
to
provide
significant
funding
for
analytical
and
monitoring
methods
research
projects.
The
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
39
objective
of
the
research
should
be
focused
on
innovative
technologies
that
fill
existing
gaps
in
current
analytical
and
monitoring
protocols
and
which
can
have
significant
impact
on
future
monitoring
programs.

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
FIELD
DETERMINATION
OF
ORGANICS
FOR
SOIL
AND
SLUDGE
USING
SUB­
CRITICAL
WATER
EXTRACTION
COUPLED
WITH
SOLID
PHASE
EXTRACTION
Steven
B.
Hawthorne,
Kimberly
J.
Hageman,
Carol
B.
Grabanski,
and
David
J.
Miller
Energy
and
Environmental
Research
Center,
University
of
North
Dakota,
Grand
Forks.
North
Dakota,
USA
58202
The
primary
purpose
of
the
proposed
investigations
is
to
couple
well
known
extraction
methds
for
water,
solid
phase
extraction
(
SPE)
and
solid
phase
microextraction
(
SPME),
with
subcritical
water
extraction
of
soils
and
sludges
to
allow
field­
portable
water
methods
to
be
applied
to
contaminated
solids.

For
water
samples,
both
SPE
and
SPME
can
be
used
to
extract
and
concentrate
organics
in
the
field
for
subsequent
analysis
(
e.
g.,
field­
portable
GC),
but
are
not
applicable
to
extracting
organic
pollutants
from
solid
samples.
If
organic
pollulants
on
soils
and
sludges
could
be
efficiently
transferred
to
water,
both
SPE
and
SPME
could
be
very
useful
for
field
determinations
of
organic
poIlutants
from
solids.
We
have
demonstrated
that
subcritical
water
(
hot
water
maintained
as
a
liquid
by
a
few
bar
pressure)
is
an
excellent
solvent
to
quantitatively
extract
polar
and
non­
polar
organics
from
soils
and
sludges.
Subcritical
water
extractions
can
be
highly
selective;
polar
organics
extract
at
lower
temperatures
(
e.
g,
phenols
and
amines
extract
at
50
to
100
°
C),
and
non­
polar
organics
extract
at
high
temperatures
(
e.
g.,
200
to
250
°
C).
By
heating
water
under
low
pressure,
solubiIities
of
polar
organics
increase
dramatically,
and
even
non­
polar
organics
such
as
PAHs
can
increase
solubilities
by
>
106­
foId.

Initial
experiments
have
demonstrated
that
coupling
subcritical
water
extraction
with
SPE
and
SPME
can
provide
an
extremely
simple,
rapid,
and
inexpensive
method
to
determine
organic
pollutants
found
on
soils,
sediments,
and
sludges.
With
proper
internal
standards,
quantitative
results
for
PAHs,
PCBs,
and
aromatic
amines
have
been
obtained
with
total
sample
preparation
(
including
extraction
from
the
solid
and
SPE
or
SPME
sorption)
of
ca.
30
minutes.
Detection
limits
of
<
ppb
are
obtained.

During
this
three­
year
study
we
will:

1.
Investigate
and
develop
the
use
of
subcritical
water
coupled
to
SPE
(
exhaustive
extraction)
for
the
quantitanve
determination
of
polar
and
non­
polar
organics
from
soils
and
sludges.

2.
Investigate
and
develop
the
use
of
subcritical
water
extraction
coupled
with
SPME
(
equilibrium
extraction)
for
the
rapid
and
qualitative
determination
of
polar
and
non­
polar
organics
from
soils
and
sludges.

3.
Optimize
the
selectivity
of
the
extractions
based
on
water
extraction
temperature
and
on
SPE
and
SPME
sorbent
selectivity.

4.
Compare
analyzing
extracts
in
the
field
to
shipping
SPE
and
SPME
devices
to
the
lab.

5.
Demonstrate
the
best
approaches
in
the
field
and
compare
results
to
conventional
EPA
extraction
and
analysis
methods.

Based
on
initial
results,
it
is
expected
that
the
proposed
investigations
will
yield
an
extremely
simple
(
requiring
only
an
extraction
cell
and
oven),
inexpensive
and
field­
portable
approach
to
utilizing
SPE
and
SPME
with
subcritical
water
for
the
extraction
and
quantitative
determination
of
polar
and
non­
polar
organics
from
contaminated
solids
and
semi­
solids.
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
40
SOLID
PHASE
MICROEXTRACTION
COUPLED
WITH
INFRARED
SPECTROSCOPY
FOR
THE
DETERMINATION
OF
ORGANIC
POLLUTANTS
IN
WATER
Sheila
A.
Lofthos­
Merschman
and
David
C.
Tilotta
Department
of
Chemistry,
University
of
North
Dakota,
Grand
Forks,
ND
58202
The
current
standard
methods
for
determining
organic
compounds
in
water
include
liquid­
liquid
extraction,
solid
phase
extraction,
solid
phase
microextraction
(
SPME),
and
purge
and
trap
gas
chromatography.
Although
each
of
these
methods
have
certain
advantages
and
disadvantages,
many
of
them
are
time­
consuming,
labor­
intensive,
difficult
to
implement,
and
require
relatively
expensive
instrumentation.
In
addition,
most
are
not
readily
adaptable
to
field
detections.
Field
determination
of
contamination
is
attractive
because
It
eliminates
many
of
the
problems
connected
with
collection
and
transporting
samples
(
e.
g.,
representative
sampling,
contamination,
loss
of
volatiles.
storage,
etc.).

Work
in
this
laboratory
is
focusing
on
the
development
of
a
simple,
solventless
method
of
analysis
that
combines
SPME
and
infrared
(
IR)
spectroscopy.
Infrared
spectroscopy
is
a
proven
sensitive
and
selective
analytical
technique
that
can
potentially
be
used
to
detect
organic
contaminants.
However,
the
direct
detection
of
organics
in
water
by
IR
spectroscopy
yields
poor
results
due
to
the
severe
spectral
interference
of
the
water.
SPME
is
a
relatively
new
analytical
technique
used
to
selectively
partition
organic
compounds
from
an
aqueous
phase
into
a
solid
sorbent.
Because
the
partition
coefficients
for
many
environmentally­
important
compounds
are
large
(
e.
g.,
in
the
2­
2,000
range),
a
substantial
preconcentration
enhancement
is
enjoyed.

We
are
constructing,
simple,
reusable
"
dipsticks"
that
contain
a
thin
film
of
a
chromatographic
solid
phase
that
is
capable
of
partitioning
target
organics
from
water
After
the
organics
partition
into
the
solid
phase,
a
portable
FT­
IR
(
Fourier
transform
infrared)
spectrometer
is
used
to
determine
the
organics
partitioned
in
the
solid
phase.
This
solventless
preconcentration
step
minimizes
water
and
matrix
interferences
and
makes
it
possible
to
detect
organic
compounds
in
the
ppm­
ppb
range.
Additionally,
because
the
SPME/
IR
method
uses
no
solvents
or
cumbersome
equipment,
it
is
field
portable.

This
presentation
will
discuss
recent
work
in
our
lab
on
the
examination
of
various
polymer
films
for
suitability
as
SPME/
lR
sorbents.
Selection
criteria
examined
include:
optical
transparency,
equilibration
times,
rigidity,
reusability,
and
analyte
selectivity.
Additionally.
we
will
also
discuss
our
primary
work
on
the
application
of
this
method
to
determining
selected
organic
contaminants
(
e.
g.,
benzene,
ethylbenzene,
trichloroethylene,
etc.)
in
aqueous
samples.
Preliminary
results
obtained
from
sorbent
films
such
as
poly(
dimethylsiloxane)
and
paraffin­
impregnated
poly(
butylene)
show
that
the
SPME/
IR
method
is
fast
(
partitioning
times
as
fast
as
40
min),
and
reproducible
(
RSDs
of
1­
11%)
with
detection
limits
ranging
from
ca.
15
ppb
­
1
ppm.
In
addition,
we
will
show
that
the
SPME/
IR
results
are
in
good
agreement
with
those
obtained
from
standard
EPA
protocols.
We
anticipate
that
the
initial
laboratory
tests
will
be
completed
by
the
end
of
1997,
and
field
testing
of
this
new
method
in
1998.

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
REAL­
TIME
TRACE
DETECTION
OF
ELEMENTAL
MERCURY
AND
ITS
COMPOUNDS
Robert
B.
Barat
New
Jersey
lnstitute
of
Technology,
Department
of
Chemical
Engineering,
Chemistry,
and
Environmental
Science,
Newark,
NJ
07102
INTRODUCTION
Emission
of
mercury
vapor
from
combustion
and
other
processes
is
now
considered
to
be
a
major
environmental
issue
[
Von
Burg
and
Greenwood,
1991].
It
can
be
emitted
directly
to
the
air
in
its
highly
volatile
elemental
form.
In
addition,
mercury
compounds
[
e.
g.
HgCl2,
Hg(
CH3)
2]
are
quite
volatile,
and
have
been
detected
in
incinerator
effluent
and
the
biosphere.

The
present
research
addresses
the
need
to
develop
real
time
stack
monitoring
of
emissions
of
Hg
and
its
compounds.
Such
technology
will
enable
combustion
facilities
to
identify
peak
emission
events
and
take
corrective
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
41
action.
It
will
replace
the
current
uncertainty
in
emission
estimations
based
on
feed
mercury
content.
Realistic
emission
inventory
will
enable
regulatory
agencies
to
better
assess
health
risks
associated
with
future
siting
of
waste
incinerators.

To
be
applicable
to
a
wide
range
of
applications,
an
instrument
should
be
capable
of
detecting
Hg
in
the
range
of
1­
5000
µ
g/
m3,
with
an
ultimate
sensitivity
limit
on
the
order
of
0.1
µ
g/
m3
(
ca.
10
pptv).
Current
best
technology
for
mercury
detection
uses
a
collection
/
concentration
step
followed
by
off­
line
analysis
using
atomic
absorption,
atomic
fluorescence,
or
plasma
emission
spectroscopy
[
Baeyens,
1992].
None
of
these
three
detection
methods
in
their
commercially
available
configurations
are
suitable
for
real­
time
monitoring
of
elemental
mercury.

The
detection
concept
for
elemental
mercury
involves
Doppler­
shifted
resonant
atomic
fluorescence
excited
by
a
UV
laser
or
a
low
pressure
mercury
lamp.
The
Doppler­
shifted
fluorescence
will
be
separated
from
the
unshifted
background
signal
by
use
of
a
mercury
vapor
filter
precisely
matched
to
the
spectral
linewidth
of
the
source.
The
detection
concept
for
compounded
mercury
will
use
the
photofragment
fluorescence
(
PFF)
excited
by
deep
UV
light.
Examination
of
the
fluorescence
spectrum
will
permit
identification
of
the
original
mercury
compounds.
Photo­
fragment
fluorescence
has
been
successfully
applied
for
the
gas­
phase
analysis
of
HgCl2,
Hg(
CH3)
Cl,
and
HgI2
(
Poulos
and
Barat,
1997).

APPROACH
­­
ELEMENTAL
HG
Atomic
Fluorescence
Spectroscopy
(
AFS)
is
a
highly
sensitive
spectroscopic
marker
for
elemental
Hg
detection.
Current
AFS
instruments
use
a
cold
vapor
trap
for
collection/
concentration
of
the
air
sample
followed
by
purging
and
excitation
with
an
atomic
vapor
mercury
lamp
(
253.7
nm)
and
measurement
of
the
resonant
fluorescence
(
at
the
same
wavelength).
Sensitivity
is
limited
by
the
elastically
scattered
light
from
the
exciting
source.
In
the
case
of
253.7
nm
Hg
fluorescence,
this
limitation
is
problematic
because
the
fluorescence
signal
is
already
reduced
due
to
quenching
by
air.

The
technique
under
study,
shown
in
Figure
1,
will
expand
the
Hg­
contaminated
air
stream
across
a
supersonic
nozzle
into
a
high
vacuum
chamber.
Light
at
253.4
nm
will
be
directed
across
the
jet.
Atomic
Hg
fluorescence
will
be
Doppler­
shifted
by
between
1
and
3
GHz
due
to
the
jet
motion
(
Barat,
1996).
Total
collected
light,
comprised
of
the
shifted
fluorescence
and
stray
elastic
scattering,
will
be
passed
out
of
the
vacuum
chamber
and
through
a
sharp
cutoff
atomic
mercury
vapor
filter
to
attenuate
stray
elastic
scattering
while
transmitting
the
fluorescence
signal.

As
illustrated
in
Figure
2,
when
narrow
bandwidth
radiation
is
incident
upon
a
high
speed
flow
field,
elastically
scattered
background
light
is
superimposed
upon
the
Doppler­
shifted
mercury
vapor
fluorescence.
If
the
excitation
source
coincides
exactly
with
the
absorption
band
of
an
optically
thick
mercury
vapor,
and,
if
a
cell
filled
with
mercury
vapor
is
placed
in
front
of
a
detector,
then
the
elastically
scattered
background
light
will
be
strongly
attenuated
(
Miles,
1991;
Finkelstein,
et
al.,
1994).

The
expansion
reduces
the
collision
rate
of
excited
mercury
with
the
air,
so
collisional
quenching
and
collisional
broadening
are
both
substantially
reduced.
The
low
temperature
associated
with
this
expansion
further
reduces
the
fluorescence
linewidth,
enhancing
the
performance
of
the
mercury
vapor
filter.

APPROACH
­­
COMPOUNDED
HG
In
general,
mercury
compounds
absorb
light
strongly
below
250
nm
[
Gowenlock
and
Trotman,
1955],
and
these
absorption
bands
are
dissociative.
In
PFF,
a
photolyzing
UV
photon
dissociates
the
target
molecule
into
fragments,
some
of
which
are
imparted
with
excess
energy.
The
energy
might
then
be
lost
by
fluorescence:

h
n
A­
B
­­­­>
A
+
B
*
(
1a)
B*
­­­­>
B
+
h
n
'
(
1b)

where
A
and
B
can
be
atoms
or
polyatomic
species.
The
fragment
identities
and
distributions,
as
revealed
in
the
fluorescence
spectrum,
can,
in
principle,
provide
information
on
the
composition
of
the
target,
in
a
manner
analogous
to
mass
spectrometry
and
other
fragmentation
spectroscopies.
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
42
For
example,
Figure
3
shows
the
PFF
spectrum
from
193
nm
excimer
laser
excitation
of
Hg(
CH3)
Cl
vapor
in
a
static
cell
(
Poulos
and
Barat,
1997).
Two
features
are
evident:
atomic
Hg
emission
lines
at
546,
577,
and
579
nm;
and
a
broad
continuum
assigned
to
the
B­­>
X
system
of
HgCl**
excited
state.
Some
of
the
following
photochemical
processes
could
explain
these
findings:

h
n
Hg(
CH3)
Cl
­­­­­­>
HgCl**
+
CH3
(
2)
HgCl**
­­­­­­>
HgCl
+
h
n
'
(
2a)
­­­­­>
Hg
+
Cl
(
2b)

Processes
(
2a)
and
(
2b)
are
clearly
implicated,
but
the
origin
of
the
Hg
atomic
emission
lines
is
open
to
question.

The
concentration
of
the
target
compound
may
be
related
to
the
fluorescence
intensity
from
a
hot
fragment.
For
example,
excitation
laser
energy
of
2
mj/
pulse
at
a
repetition
rate
of
10
Hz
at
229
nm
was
applied
to
HgCl:
vapor
in
a
static
cell
(
Poulos
and
Barat,
1997).
Analyte
concentrations
were
varied
by
changing
the
bath
temperature
around
an
analyte
reservoir
of
the
static
optical
cell
used.
The
PFF
signal
was
monitored
at
253.7
nm,
the
strongest
fluorescence
line.
As
shown
in
Figure
4,
signal
linearity
was
observed
over
at
least
2.5
orders
of
magnitude
of
concentration.
The
concentrations
studied
overlap
the
range
of
inorganic
mercury
levels
measured
exiting
from
stacks
of
coal­
powered
utilities
and
municipal
waste
incinerators
by
standard
methods.

The
supersonic
jet
spectroscopy
are
expected
to
further
improve
sensitivity
of
PFF
detection.
Fluorescence
spectra
will
be
sharpened,
leading
to
better
discrimination
of
fragment
vibrational
structure.
Quenching
by
O2
will
be
reduced
due
to
the
low
pressure.
The
UV
source
will
penetrate
deeper
into
the
sample
because
there
is
no
optical
filtering
of
the
light
by
O2
from
the
window
to
the
jet,
and
reduced
filtering
in
the
jet.

QUALITY
ASSURANCE/
PERFORMANCE
ASSESSMENT
For
comparison
of
time­
averaged
test
concentrations,
the
mercury­
containing
sample
stream
will
be
diverted
to
a
reference
method,
such
as
optical
absorption.
Testing
of
the
research
technology
will
consist
of
obtaining
data
on
five
performance
measures:

Relative
Accuracy:
the
absolute
mean
difference
between
the
metals
concentration
determined
by
the
monitor
and
that
determined
by
the
reference
method,
plus
a
2.5
percent
uncertainty
confidence
coefficient
based
on
a
test
series.

Calibration
Drift:
the
difference
in
the
monitor
output
reading
from
the
established
reference
value
after
a
stated
period
of
operation.
The
reference
value
is
established
by
a
calibration
standard
which
has
a
concentration
nominally
80
percent
or
greater
of
the
full
scale
reading
capability
of
the
monitor.

Zero
Drift:
calibration
drift
when
the
reference
value
is
zero.

Response
Time:
the
time
interval
between
the
start
of
a
step
change
in
the
concentration
of
the
monitored
gas
stream
and
the
time
when
the
output
signal
reaches
95
percent
of
the
final
value.

Detection
Limit:
three
times
the
standard
deviation
of
nine
repeated
measurements
of
a
low­
level
(
near
blank)
sample.

IN
CONCLUSION
This
program
will
determine
the
capabilities
and
limitations
of
mercury
detection
in
real­
time
by
the
proposed
techniques.
It
is
anticipated
that
the
detection
limits
for
elemental
mercury
will
be
on
the
order
of
0.1
µ
g/
m3
(
ca.
10
pptv).
It
is
desirable
that
response
be
linear
up
to
cat
5000
µ
g/
m3.
It
is
anticipated
that
PFF
signals
will
successfully
discriminate
between
species
such
as
HgCl2,
HgO,
Hg(
CH3)
2,
and
Hg(
CH3)
Cl,
at
detection
levels
below
1
µ
g/
m3.

ACKNOWLEDGMENT
The
author
would
like
to
thank
Dr.
Arthur
T.
Poulos,
President
of
Poulos
Technical
Services,
Inc.,
for
his
significant
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
43
contributions
to
this
research.
The
author
would
also
like
to
thank
the
U.
S.
Environmental
Protection
Agency
­
National
Center
for
Environmental
Research
and
Quality
Assurance
for
its
financial
support
of
this
work,
and
grant
project
officer
Mr.
William
Stelz
for
his
administrative
guidance.

REFERENCES
Baeyens,
W.,
Trends
in
Analytical
Chemistry,
vol.
11,
245
(
1992).
Barat,
R.
B.,
"
Real­
Time
Trace
Detection
of
Elemental
Mercury
and
its
Compounds,"
research
proposal
to
U.
S.
Environmental
Protection
Administration,
NCERQA,
Analytical
and
Monitoring
Methods
(
1996).
Finkelstein,
N.,
Gambogi,
J.,
Lempert,
W.
R.,
Miles,
R.
B.,
Rine,
G.
A.,
Finch,
A.
and
Schwarz,
R.
A.,
Proceedings
of
32nd
Aerospace
Sciences
Meeting,
January,
1994,
Reno,
NV,
Paper
#
A1AA
94­
0492.
Gowenlock,
B.
G.
and
Trotman,
J.,
J.
Chem.
Society,
pt.
2,
1454
(
1955).
Miles,
R.
B.,
"
Absorption
Line
Filter
Window
and
Method
for
Velocity
Measurements
by
Light
Scattering,"
U.
S.
Patent
#
4,988,190
(
1991).
Poulos,
A.
T.
and
Barat,
R.
B.,
''
Detection
of
Mercury
Compounds
in
the
Gas
Phase
by
Laser
Photo­
Fragment
Fluorescence,"
submitted
to
Applied
Spectroscopy
(
1997).
Von
Burg,
R.
and
Greenwood,
M.
R.,
in
Metals
and
Their
Compounds
in
the
Environment,
ed.
by
E.
Merian,
VCH,
Weinheim
(
1991).

Figure
2.
Mercury
Atomic
Filter
Concept
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
44
Figure
1.
Experimental
Test
Apparatus
Figure
3.
PFF
Spectrum
of
Hg(
CH3)
Cl
Figure
4.
PFF
Intensity
(
at
254
nm)
vs.
HgCl2
Concentration
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
45
ELECTRODIALYTIC
NAOH
ELUENT
PRODUCTION
AND
GRADIENT
GENERATION
Purnendu
K.
Dasgupta
Texas
Tech
University,
Lubbock,
TX
79409­
1061
A
small
inexpensive
system
is
described
that
allows
high
performance
suppressed
anion
chromatography
on
a
capillary
scale.
A
fully
computer
controlled
stepper
motor
driven
syringe
type
dispenser,
equipped
with
a
500
mL
capacity
glass
syringe
is
capable
of
pumping
at
pressures
up
to
1000
psi
when
equipped
with
an
appropriate
inlet
check
valve.
Fused
silica
capillary
columns
~
50
cm
in
length
and
180
mm
in
i.
d.,
packed
in­
house
with
a
commercial
packing,
provide
excellent
performance,
significantly
exceeding
the
efficiencies
observed
for
the
same
packing
in
commercially
avaiIable
2
mm
bore
format.
The
system
operates
with
a
pressure
drop
of
<
800
psi
at
a
flow
rate
of
2
mL/
min.
The
system
utilizes
a
novel
electrodialytic
NaOH
eluent
generator
that
is
deployed
on
the
high
pressure
side
of
the
pump
and
thus
requires
no
special
measures
for
electrolytic
gas
removal
This
device
permits
both
isocratic
and
gradient
operation
with
excellent
eluent
purity;
the
NaOH
concentration
is
generated
linearly
with
applied
current
with
near­
Faradaic
efficiency
up
to
a
concentration
of
at
least
100
mM.

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
SENSORS
FOR
DIRECT
MONITORING
OF
ENVIRONMENTAL
POLLUTANTS
F.
Yan,
M.
Masila,
A.
Sargent,
and
Omowunmi
A.
Sadik
Department
of
Chemistry,
State
University
of
New
York
at
Binghamton,
PO
Box
6016,
Binghampton,
NY
13902­
6016
Ph
(
607)
777­
4132
Fax:
(
607)
777­
4478
The
considerable
amount
of
time
and
costs
involved
in
carrying,
out
field
sampling
and
laboratory
analysis
of
environmental
samples
has
resulted
in
the
need
to
develop
alternative
screening
methods.
Immunosensors
are
capable
of
"
decentralizing"
environmental
testing
through
on­
site
rapid
screening.
Current
research
and
development
efforts
are
directcd
towards
more
compact
cheap,
and
user­
friendly
devices.
In
our
laboratories
we
are
pursuing
the
development
of
portable,
rapid,
cost­
effective,
and
in­
situ,
multi­
layer
electro­
optical
sensors
for
continuous
detection
of
environmental
samples
in
real
time.
In
our
previous
work,
we
have
shown
that
conducting
electroactive
polymer
(
CEP)­
based
sensors
possess
significant
advantages
over
conventional
methods
of
detection
and
quantitation
of
environmental
analyses.
Also
the
uniqueness
of
CEP­
based
sensors
in
combining
the
role
of
transducers
required
for
measuring
immunological
reactions
with
that
of
antibody
(
Ab)
entrapment
matrices
translates
into
substantial
equipment
miniturization
as
well
as
huge
reduction
in
response
time.
This
promises
to
open
up
new
horizons
in
environmnental
monitoring,
medical
and
clinical
applications.
This
paper
will
discuss
the
deveIopment
of
a
direct
multi­
electrode
array
detection
and
quantitation
of
polychlorinated
biphenyls
(
PCBs),
heavy
metals,
and
atrazines,
using
CEP­
based
sensors.
Preliminary
results
obtained
with
enzyme­
based
electro­
optical
detection
methods
will
aIso
be
analyzed
and
discussed.
The
advantages
of
utilizing
these
types
of
assay
formats
to
provide
unparalleled
rapid,
sensitive,
and
cost­
effective
options
for
environmental
analsis
of
pesticides
will
be
presented.
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
46
INORGANIC
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
47
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
48
PRACTICAL
CLEAN
CHEMISTRY
TECHNIQUES
FOR
TRACE
AND
ULTRA­
TRACE
ELEMENTAL
ANALYSIS
H.
M.
"
Skip"
Kingston
and
Peter
J.
Walter
Duquesne
University,
Department
of
Chemistry
&
Biochemistry,
Pittsburgh,
PA
15282­
1503
Trace
and
ultra­
trace
analysis
is
as
much
dependent
on
the
control
of
the
analytical
blank
as
it
is
on
the
accuracy
and
precision
of
the
instrument
making
the
measurement.
Inability
to
control
contamination
that
is
external
to
the
sample,
or
those
contributions
of
the
analyte
coming
from
all
other
sources
than
the
sample,
is
frequently
the
limiting
factor
in
trace
(
parts
per
million
(
ppm)
to
parts
per
billion
(
ppb))
and
ultra­
trace
analysis
(
ppb
to
parts
per
trillion
(
ppt)).
Analytical
blank
contributions
occur
from
four
major
sources:
surroundings,
reagents,
vessels,
and
the
analyst's
technique.

The
analytical
blank
can
be
dramatically
reduced
through
the
combination
of
clean
chemistry
and
microwave
sample
preparation.
Microwave
sample
preparation
systems
minimizes
continuous
sustained
transfer
of
elements
from
the
air
to
the
sample
solution
by
providing
a
controlled,
closed,
or
restricted
sample
container
during
decomposition1
It
prevents
the
entry
of
the
laboratory
atmosphere
and
thus
prevents
the
majority
of
environmental
airborne
contamination.
This
leaves
only
the
air
within
the
vessel,
captured
during
transfer,
and
digestion/
extraction.
To
prevent
this
volume
of
air
from
contaminating
the
sample,
the
sample
must
be
transferred
in
a
clean
environment.

Portable
class
100
clean
benches
were
designed
and
produced
to
allow
for
addition
of
sample
and
reagents,
and
to
open
or
close
vessels
of
both
the
atmospheric
and
closed
vessel
design.
These
carts
can
be
constructed
for
$
1,000­$
1,600,
depending
on
whether
they
need
to
be
exhausted
or
are
just
islands
of
clean
laboratory
air.
Similar
portable
clean
systems
are
becoming
available
from
analytical
instrument
manufacturers.
A
clean
laboratory
of
dimensions
18'
x
13.5'
was
designed
and
constructed
for
approximately
$
30,000
using
readily
available
materials
and
specialized
HEPA
clean
hood
components.
This
clean
facility
is
used
for
both
microwave
vessel
sample
preparation
and
ICP­
MS
analysis.
The
ICP­
MS
is
located
in
the
clean
laboratory,
so
its
extremely
low
detection
capability
may
be
utilized.

95%
of
the
capability
of
a
cleanroom
can
be
made
available
to
standard
environmental
laboratories
for
between
$
1,500
and
$
15,000.
The
quality
of
the
analysis
can
be
improved
for
ultra­
trace
analysis
of
many
type
for
modest
cost.
Some
of
these
relatively
low
cost
systems
will
be
described
for
both
their
effectiveness
and
method
of
implementation.
These
modest
cost
methods
will
be
proposed
for
inclusion
in
chapter
3
of
SW­
846
as
guidelines
to
improve
environmental
analytical
analysis.

1.
H.
M.
Kingston,
P.
J.
Walter,
S.
J.
Chalk,
E.
Lorentzen,
and
D.
Link,
Chapter
3:
Environmental
Microwave
Sample
Preparation
Fundamentals,
Methods,
and
Applications
In
Microwave
Enhanced
Chemistry;
H.
M.
Kingston,
S.
Haswell,
Eds.;
American
Chemical
Society:
Washington,
D.
C.
(
1997).

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
FLAME
ATOMIC
ABSORPTION
FOR
THE
DETERMINATION
OF
ARSENIC
AND
SELENIUM
IN
TCLP
EXTRACTS
Zoe
Grosser
The
Perkin­
Elmer
Corporation,
50
Danbury
Road
MS­
219
Wilton,
CT
06897­
0219
grosseza@
perkin­
elmer.
com
ABSTRACT
Toxicity
Characteristic
testing
is
performed
to
assess
the
potential
of
a
material
to
leach
hazardous
constituents
after
disposal
in
a
landfill.
The
Toxicity
Characteristic
Leaching
Procedure
(
TCLP)
extract
solution
is
analyzed
for
31
organic
components
and
8
metals
and
if
the
regulated
limits
are
exceeded
the
material
is
considered
hazardous
and
must
be
treated
appropriately.
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
49
Currently
approved
EPA
Methods
for
the
determination
of
As
and
Se
include
graphite
furnace
atomic
absorption
(
GFAA)
(
7060A,
7740)
and
hydride
generation
atomic
absorption
(
7061A,
7741A).
With
the
advent
of
modern
instrumentation,
including
electrodeless
discharge
lamp
light
sources,
more
stable
electronics,
background
correction,
and
improved
sample
introduction
systems,
flame
atomic
absorption
analysis
for
low
levels
of
these
elements
becomes
viable.

Detection
limits
ten
times
below
the
MCL
are
generally
accepted
as
necessary
to
ensure
reliability
at
the
decision­
making
point.
Method
detection
limits
were
calculated
for
As
and
Se
based
on
the
EPA
procedure
published
at
40CFR,
part
136.
The
method
detection
limit
obtained
for
As
was
0.12
ppm
and
for
Se
was
0.04,
using
the
high
sensitivity
nebulizer.
The
detection
limit
for
As
would
also
be
lower
using
the
high
sensitivity
nebulizer
(
0.04),
but
meets
our
goal
of
ten
times
below
the
MCL
(
0.5
ppm
needed)
with
the
universal
nebulizer.

Calibration
is
an
important
component
of
an
analysis
and
we
will
discuss
the
choice
of
calibration
standards
and
the
resulting
calibration.
The
linear
range
will
be
documented
and
the
value
of
extending
the
calibration
curve
past
the
linear
range
will
be
discussed.

The
developed
flame
atomic
absorption
method
will
be
applied
to
the
analysis
of
As
and
Se
in
real
TCLP
samples.

INTRODUCTION
One
of
the
more
popular
tests
performed
under
the
EPA
Resource
Conservation
and
Recovery
Act
(
RCRA)
program
is
the
Toxicity
Characteristic
Leaching
Procedure
(
TCLP).
TCLP
testing
is
performed
to
assess
the
potential
of
a
material
to
leach
hazardous
constituents
after
disposal
in
a
landfill.
The
first
step
involves
extraction
of
the
potentially
hazardous
material
(
method
1311)
with
a
mildly
acidic
buffered
solution.
The
extract
is
then
analyzed
for
31
organic
compounds
and
8
metals
and
if
any
of
the
analyses
are
present
at
levels
exceeding
the
maximum
allowed,
the
material
is
considered
hazardous
and
must
be
treated
as
such.
Table
I
lists
the
maximum
contaminant
levels
(
MCLs)
permitted
in
the
extract
for
metals.

Table
I.
TCLP
Metal
Limits
5.0
Ag
1.0
Se
0.2
Hg
5.0
Pb
5.0
Cr
(
total)
1.0
Cd
100
Ba
5
As
MCL
(
mg/
L)
Element
The
methods
for
contaminant
analysis
can
come
from
many
reputable
sources
but
the
EPA
RCRA
publication
Test
Methods
for
Evaluating
Solid
Waste,
Physical/
Chemical
Methods;
SW­
846
is
a
primary
source
for
many
organizations
performing
RCRA
measurements.
The
methods
available
in
SW­
846
for
As
and
Se
determination
are
listed
in
Table
II.

Table
II.
RCRA
SW­
846
Methods
­­
6010
7741A
7740
Se
6020
6010
7061A
7060A
As
ICP­
MS
ICP­
OES
Hydride
GFAA
With
the
advent
of
modern
instrumentation
including
electrodeless
discharge
lamp
light
sources
more
stable
electronics
background
correction
and
improved
sample
introduction
systems
flame
atomic
absorption
determination
of
these
elements
becomes
viable.
Flame
atomic
absorption
is
an
inexpensive
technique
to
acquire
and
operate
and
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
50
many
laboratories
might
prefer
the
choice
if
a
method
was
readily
available.

The
goal
of
this
paper
is
to
evaluate
the
possibility
of
FLAA
for
the
determination
of
As
and
Se
in
TCLP
extracts.
The
detection
limits
will
be
determined
to
assess
the
analytical
feasibility.
The
linear
ranges
and
potential
interferences
will
be
evaluated.
Stability
over
the
course
of
an
unattended
analytical
run
and
the
recovery
of
spikes
in
real
matrices
will
be
presented.

EXPERIMENTAL
The
Perkin­
Elmer
®
AAnalyst
 
100
atomic
absorption
spectrometer
was
used
for
all
measurements.
The
analytical
parameters
are
shown
in
Table
III.
The
high
sensitivity
GemTip
 
nebulizer
was
used
exclusively
for
Se
because
the
estimated
detection
limits
indicated
that
the
additional
sensitivity
would
be
required.
The
universal
and
high
sensitivity
GemTip
 
nebulizers
were
both
evaluated
for
As.
Standards
were
prepared
to
contain
approximately
1%
HNO3
and
2%
HCI
(
GFS
Chemicals
®
,
Columbus,
OH)

Table
III.
AAnalyst
100
Instrumental
Conditions
0.1,
0.5,
1.0,
2.0
0.2,
2.0,
5.0,
10.0
Standards
(
mg/
L)
Linear
Linear
Calibration
High
Sensitivity
Universal
or
High
Sensitivity
Nebulizer
Air­
Acet.
Air­
Acet.
Flame
AA­
BG
AA­
BG
Mode
2.0
0.7
Slit
(
nm)
196.0
193.7
Wavelength
(
nm)
Se
As
Parameter
Keith
Hutchinson
of
NET
Midwest,
Inc.
provided
the
digested
TCLP
extracts.
The
synthetic
sample
was
prepared
to
approximate
the
provided
samples
by
combining
single
element
stock
solutions
to
create
a
matrix
of
1500
mg/
L
Na,
20
mg/
L
Ca,
and
70
mg/
L
Fe.

RESULTS
AND
DISCUSSION
Detection
limits
ten
times
below
the
MCL
are
generally
accepted
as
necessary
to
ensure
reliability
at
the
decision­
making
point.
This
means
that
we
would
need
a
detection
limit
for
As
of
0.5
mg/
L
and
0.1
mg/
L
for
Se
to
meet
the
goal.
Method
detection
limits
were
calculated
based
on
the
EPA
procedure
published
at
40CFR,
part
136
1.
The
detection
limits
shown
in
Table
IV
meet
the
goals
described.
The
universal
nebulizer
is
adequate
for
the
measurement
of
As
or
the
high
sensitivity
can
be
used
for
additional
margin.
The
high
sensitivity
nebulizer
is
necessary
to
meet
the
goals
for
the
determination
of
Se.

Table
IV.
Method
Detection
Limits
(
MDLs)

0.041
0.043
High
Sensitivity
­­
0.12
Universal
Se
As
Nebulizer
Linear
ranges
must
be
determined
initially
to
characterize
instrumental
performance.
It
is
generally
accepted
that
a
calibration
is
no
longer
linear
when
the
measured
concentration
based
on
a
lower
level
calibration
deviated
from
the
the
value
by
more
than
5%.
Using
a
spreadsheet
program
to
calculate
the
percentage
difference
at
increasing
concentrations
the
linear
range
was
rounded
off
to
20
mg/
L
for
Se
and
60
mglL
for
As.
Figure
1
shows
the
curves
for
Se
and
Figure
2
shows
the
analogous
curves
for
As
(
universal
nebulizer).

Modern
instruments
using
algorithms
to
reliably
model
the
curves
obtained
at
higher
concentration
can
give
accurate
and
precise
measurements
above
the
linear
range.
The
choice
of
calibration
standards
is
more
critical
and
three
standards
should
be
used
in
a
ratio
of
6:
3:
1,
with
6
representing
the
highest
concentration2.
If
performance
can
be
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
51
verified
through
the
use
of
quality
control
checks
for
calibration
verification
precision
and
recovery
this
extends
the
useful
working
range
of
the
instrument.
For
atomic
absorption
with
one
of
the
shorter
working
ranges
this
can
improve
the
range
four­
six
times
above
the
concentration
at
the
top
of
the
linear
range.

Figure
1.
Se
linear
range.

Before
measuring
real
samples
a
synthetic
sample
was
evaluated
to
determine
if
matrix
effects
were
present.
As
and
Se
were
spiked
at
a
low
level
and
at
the
MCL
into
a
synthetic
matrix
approximated
from
ICP­
MS
analysis
of
a
digested
sample.
Table
V
shows
the
results
indicating
that
the
matrix
was
not
a
problem
for
the
sample
introduction
system.

Figure
2.
Arsenic
linear
range.

Table
V.
Synthetic
Sample
Results
95.2
1.0
93.0
0.5
Se
92.2
5.0
113
0.5
As
%
Recovery
High
Spike
(
mg/
L)
%
Recovery
Low
Spike
(
mg/
L)
Element
The
analysis
of
samples
spiked
before
the
digestion
process
can
indicate
the
entire
process
is
in
control.
This
includes
the
digestion
instrument
calibration,
and
method.
Laboratory
control
samples
provided
from
NET
spiked
at
1
mg/
L
prior
to
digestion
are
shown
in
Table
VI.

Table
VI.
Predigestion
Spikes
%
Recovery
of
1
mg/
L
Spike
108
109
4­
23B
102
99.3
4­
19E
Se
As
Sample
The
percentage
recovery
falls
between
99.3
to
108%
well
within
the
EPA
limits
of
80­
120%
recovery
of
the
expected
value.

Post
digestion
spikes
were
added
to
TCLP
samples
at
a
low
level
and
high
level
to
confirm
the
method
performance
with
a
variety
of
matrices.
The
results
are
shown
in
Table
VII
and
demonstrate
acceptable
recoveries
in
all
cases.
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
52
Table
VII.
Post
Digestion
Spike
Results
93.2
1.0
98.1
5.0
333133
105
1.0
89.8
5.0
333132
98.6
0.5
95.5
1.0
333131
102
0.5
106
1.0
333130
99.0
0.5
107
1.0
3­
16M
LCS
%
Recovery
Spike
Added
(
mg/
L)
%
Recovery
Spike
Added
(
mg/
L)
Selenium
Arsenic
Sample
Productivity
is
an
important
component
of
an
environmental
analysis
and
TCLP
digests
contain
a
fairly
high
amount
of
dissolved
material.
The
stability
of
a
standard
was
monitored
during
an
autosampler
analysis
of
160
samples
at
regular
intervals.
The
5­
mg/
L
As
standard
was
recovered
within
±
5%
of
the
expected
value
during
the
more
than
2
hour
run.
The
1­
mg/
L
Se
standard
was
recovered
within
±
6%
of
the
expected
value.

CONCLUSIONS
Modern
flame
atomic
absorption
can
meet
the
analytical
requirements
of
a
TCLP­
digest
analysis.
The
method
detection
limits
are
more
than
ten
times
below
the
MCL
where
a
decision
is
made.
The
linear
range
has
been
characterized
and
provides
a
useful
range
for
the
amounts
of
As
and
Se
expected
in
a
typical
digested
sample.
The
useful
concentration
could
be
extended
with
the
use
of
nonlinear
calibration
curves.
Recoveries
in
synthetic
and
real
samples
demonstrate
freedom
from
interferences
and
the
utility
of
the
method.

Flame
atomic
absorption
is
less
time­
consuming
and
labor
intensive
than
GFAA
or
hydride
analysis.
If
used
in
the
determination
of
the
full
suite
of
TCLP
metals,
it
can
help
reduce
the
number
of
techniques
required
for
the
analysis
of
8
metals.
Consolidation
of
techniques
can
reduce
the
analysis
time
for
a
sample
and
improve
turnaround
time.

The
EPA
should
be
encouraged
to
include
a
flame
atomic
absorption
method
in
SW­
846
that
includes
these
elements.
Data
demonstrating
performance
at
low
levels
will
give
laboratories
additional
flexibility
in
choosing
the
method
best
suited
for
the
analytical
scenario
and
business
requirements.

ACKNOWLEDGMENTS
Thanks
to
Keith
Hutchinson
and
Jackie
Webster
of
NET
Midwest,
Inc.
for
the
digested
TCLP
extracts.
Thanks
to
Dr.
Ruth
Wolf,
Perkin­
Elmer,
for
the
ICP­
MS
TCLP
profile.

REFERENCES
1.
United
States
Code
of
Federal
Regulations
40
Part
136
Appendix
B
(
July
1,
1996).
2.
Barnett
W.
B.
Spectrochim.
Acta.
39B
(
6)
829
(
1984).

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
FIELD
AND
LABORATORY
ANALYSIS
OF
MERCURY
Peter
J.
Walter,
H.
M.
"
Skip"
Kingston,
Helen
Boylan,
and
Ye
Han
Duquesne
University,
Department
of
Chemistry
&
Biochemistry,
Pittsburgh,
PA
15282­
1503
Mercury
is
a
toxic
element
in
many
chemical
forms
whose
analysis
is
hampered
by
its
intrinsic
mobility
and
volatility.
There
is
a
potential
for
loss
of
mercury
during
every
stage
of
analysis
including:
collection,
storage,
sample
preparation,
and
analysis.
Collection
and
storage
of
the
sample
can
have
adverse
effects
on
the
mercury
concentration
and
its
species.
Sample
preparation
can
lose
mercury
during
drying,
digestion,
and
reactions
of
the
mercury
with
vessel
walls.
Analyses
of
mercury
are
also
hampered
by
mercury
reacting
or
sticking
to
components
of
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
53
the
analysis
system.
Two
approaches
to
solving
these
problems
are
evaluated:
analysis
of
mercury
through
direct
analysis
of
solid
materials,
requiring
little
or
no
sample
preparation;
and
alternatively
identification
of
sources
of
mercury
loss
during
every
stage
of
sample
preparation1.

Using
an
automated
mercury
analyzer
(
AMA­
254,
Milestone
S.
r.
l),
the
solid
or
liquid
sample
can
be
directly
analyzed
without
sample
preparation.
This
technique
is
capable
of
on­
line
decomposition
of
solid
samples
in
an
oxygen
furnace
while
collecting
the
mercury
as
an
amalgam2.
Alternatively,
a
liquid
sample
can
be
introduced
into
the
oxygen
furnace
with
the
mercury
collected
as
the
amalgam.
Subsequent
release
of
the
mercury
into
a
detector
leads
to
total
mercury
analysis
at
the
nano­
gram
level.
Due
to
the
instruments
rugid
and
compact
design
this
instrument
will
be
demonstrated
as
a
field
technique
as
well
as
being
compatible
for
direct
and
decomposed
samples.
This
is
one
of
the
first
field
capable
instruments
that
has
the
potential
to
produce
laboratory
quality
data
in
field
environments.

Traditional
sample
preparation
steps
are
evaluated
for
mercury
quantitation.
Drying
of
the
sample
in
an
oven,
vacuum
drying,
and
other
techniques
rely
on
different
mechanisms
for
drying
which
show
varying
degrees
of
mercury
loss.
Typically,
acid
digestion
of
a
sample
in
a
closed
vessel
retains
the
mercury.
Standard
EPA
SW­
846
Methods
3052,
3051A,
3015A,
and
3050B
will
be
evaluated
for
their
recovery
of
mercury3.
Many
analysis
techniques
require
matrix
conversion
after
digestion.
This
conversion
may
inadvertently
unstabilize
the
mercury
ions,
leading
to
elemental
loss
This
study
will
demonstrate
that
a
mercury
analyzer
is
capable
of
analyzing
solid
and
liquid
samples,
either
the
collected
sample
or
the
acid
digest
solution,
at
trace
levels
in
the
laboratory
or
in
the
field.
For
these
reasons,
we
have
proposed
EPA
standard
method
7473
for
the
analysis
of
solid
and
liquid
samples
either
in
the
laboratory
or
the
field.

1.
Walter,
P.
J.,
Kingston,
H.
M.
"
The
Fate
of
Mercury
in
Sample
Preparation",
Pittsburgh
Conference,
Atlanta,
GA,
1997,
paper
no.
1223.
2.
Salvato,
N.,
Pirola,
C.,
"
Analysis
of
mercury
traces
by
means
of
solid
sample
atomic
absorption
spectrometry",
Mikrochimica
Acta
123,
pp.
63­
71,
1996.
3.
H.
M.
Kingston,
P.
J.
Walter,
S.
J.
Chalk,
E.
Lorentzen,
and
D.
Link,
Chapter
3:
Environmental
Microwave
Sample
Preparation
Fundamentals,
Methods,
and
Applications
In
Microwave
Enhanced
Chemistry;
H.
M.
Kingston,
S.
Haswell,
Eds.;
American
Chemical
Society:
Washington,
D.
C.
(
1997).

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
LEGALLY
DEFENSIBLE
SPECIATED
MEASUREMENTS
USING
SIDMS
H.
M.
"
Skip"
Kingston,
Dengwei
Huo,
Yusheng
Lu,
and
Peter
J.
Walter
Duquesne
University,
Department
of
Chemistry
and
Biochemistry,
Pittsburgh,
PA
15282­
1503
A
new
method
for
the
accurate
determination
of
chemical
species
in
the
environment
has
been
developed:
Speciated
Isotope
Dilution
Mass
Spectrometry
(
SIDMS)
1.
This
method
utilizes
isotopically
enriched
speciated
spikes
combined
with
isotope
dilution
to
accurately
determine
and
correct
for
specie
transformations
that
occur
in
sample
processing.
The
errors
in
the
measurement
are
those
that
are
limited
by
the
ability
of
the
ratio
measurement
and
the
equilibrium
of
the
species.
It
was
specifically
developed
to
address
the
problems
of
accurately
quantifying
different
species
in
complicated
matrices.
Additionally,
it
is
a
diagnostic
tool
for
identifying
both
the
error
and
bias
inherent
in
specific
methods
of
sampling
process,
storage,
sample
preparation,
and
measurement.
SIDMS
is
applicable
to
most
non­
monoisotopic
elements
and
extends
to
various
oxidation
states,
organometallics,
and
molecular
forms
of
species.

Data
for
species
such
as
the
highly
reactive
Cr(
VI)
and
Cr(
III)
demonstrate
the
capabilities
of
this
method.
Currently
these
species
are
extracted
from
soil
using
Method
3060A
and
are
analyzed
with
Method
7196A.
Both
extraction
and
analysis
methods
are
biased
in
the
speciation
measurement
process.
SIDMS
is
capable
of
accurately
correcting
back
to
the
original
concentration
samples
in
which
>
50%
degradation
of
Cr(
VI)
to
Cr(
III)
during
sample
processing.
These
unique
capabilities
will
eventually
make
it
possible
to
establish
standard
speciation
measurement
methods
and
to
develop
standard
sampling
procedures
for
speciation.
The
method
permits
bias
correction
for
degradation
of
the
analyte
species
during
collection,
storage,
extraction
and
chemical
manipulation.
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
54
A
new
reference
method
based
on
Speciated
Isotope
Dilution
Mass
Spectrometry
will
be
proposed
to
permit
the
implementation
of
this
procedure.
The
objective
is
to
provide
a
legally
defensible
reference
method
for
measurements
that
have
high
degrees
of
uncertainty
and
error
due
to
highly
reactive
analyses
such
as
species.

1.
Kingston,
H.
M.,
Patent
Number
5,414,259,
"
Method
of
Speciated
Isotope
Dilution
Mass
Spectrometry",
U.
S.
Patent,
Filed
U.
S.
Patent
Office,
Granted
May
9,
1995.
2.
Kingston,
H.
M.;
Huo,
D.;
Chalk,
S.;
Walter,
P.
J.,
"
The
Accurate
Determination
of
Species
by
Speciated
Isotope
Dilution
Mass
Spectrometry:
Exemplified
by
the
Evaluation
of
Chromium
(
VI)
in
Soil",
Proceedings
from
the
12th
Annual
Waste
Testing
&
Quality
Assurance
Symposium,
July
23­
26,
1996,
pp
112­
119.

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
EPA
METHODS
3015A
AND
3051A:
VALIDATION
STUDIES
FOR
UPDATED
MICROWAVE
LEACH
METHODS
Dirk
D.
Link,
Peter
J.
Walter,
and
H.
M.
"
Skip"
Kingston
Duquesne
University,
Department
of
Chemistry
and
Biochemistry,
Pittsburgh,
Pennsylvania
15282­
1503
ABSTRACT
Validation
studies
for
two
recently
proposed
updated
microwave
leach
methods
(
Method
3051A
and
Method
3015A)
have
been
conducted.
These
updated
methods
allow
for
the
inclusion
of
hydrochloric
acid
in
the
leaching
mixture
for
complexation
and
stabilization
of
certain
RCRA­
regulated
elements.
Experimental
details
of
the
validation
procedure
are
discussed.
Experimental
data
demonstrates
the
effectiveness
of
inclusion
of
HCl
in
recovering
analyses
like
antimony,
silver
and
iron.
Final
validation
data
for
Method
3051A
is
presented.
Validation
for
Method
3015A
has
also
progressed,
with
enhancements
in
"
problem"
analyte
recoveries
also
being
shown.

INTRODUCTION
Recently,
updated
versions
of
two
microwave
extraction
methods
have
been
proposed
for
acceptance
by
the
EPA
to
be
included
in
the
next
update
of
SW­
846.
These
updated
methods
are
EPA
Method
3051A,
an
acid
leach
method
for
sediments,
sludges,
soils,
and
oils;
and
EPA
Method
3015A,
an
acid
leach
method
for
aqueous
samples.
EPA
microwave
leach
Method
3051
was
proposed
as
an
alternative
to
EPA
hot­
plate
leach
Method
3050
(
B).
However,
because
Method
3051
is
limited
to
the
use
of
nitric
acid
only,
it
does
not
recover
certain
"
problem"
analyses
as
completely
as
Method
3050,
which
also
includes
hydrochloric
acid
and
hydrogen
peroxide.
Some
examples
of
these
"
problem"
analyses
are
antimony,
silver,
and
high
concentrations
of
iron.
These
metals,
and
others,
are
not
stable
in
the
strong
oxidizing
environment
of
nitric
acid­
only
digestions.
The
updated
microwave
methods
provide
the
analyst
with
options
for
using
alternate
reagent
combinations
to
enhance
the
performance
and
appropriateness
of
the
methods
for
certain
analyses.
For
example,
preliminary
results
have
demonstrated
that
the
recovery
of
antimony
increases
from
0%
in
an
all­
nitric
digest
to
almost
80%
in
a
digest
using
both
nitric
acid
and
hydrochloric
acid1.
The
chemistry
of
the
updated
methods
has
been
modified
to
accurately
reproduce
the
chemistry
of
the
standard
EPA
hot­
plate
methods.
The
option
to
combine
hydrochloric
acid
with
the
nitric
acid
in
the
optimized
ratio
of
9
mL
HNO3
to
3
mL
HCl,
is
provided
to
enable
the
complexation
and
stabilization
of
elements
such
as
aluminum,
antimony,
iron,
and
silver,
when
needed.
This
acid
ratio
has
demonstrated
optimum
recoveries
for
all
26
RCRA­
regulated
elements.
It
increases
recoveries
of
certain
analyses
without
sacrificing
the
recoveries
of
other
RCRA­
regulated
metals.
Data
demonstrating
the
need
for
inclusion
of
HCl
in
the
microwave
methods
will
be
presented.
The
enhanced
performance
of
this
updated
microwave
leach
method
will
be
demonstrated
for
a
variety
of
matrices,
namely
a
sediment
sample
(
SRM
2704,
Buffalo
River
sediment),
a
sludge
sample
(
a
1­
to­
1
mixture
of
SRM
2704
and
SRM
1634c,
Trace
Elements
in
Fuel
Oil),
a
soil
sample
(
SRM
4355,
Peruvian
soil),
and
an
oil
sample
(
SRM
1084a,
Wear
Metals
in
Oil).
Final
validation
for
Method
3051A
will
be
demonstrated
by
comparing
the
recovery
data
from
these
matrices
using
the
all­
nitric
Method
3051
digest
with
the
data
using
the
nitric­
and­
hydrochloric
mixed
acid
Method
3051A
digest.

EPA
Method
3015
is
also
a
nitric
acid­
only
leach,
and
does
not
allow
for
inclusion
of
HCl
for
complexation
and
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
55
stabilization
of
certain
analyses.
In
proposed
EPA
Method
3015A,
the
inclusion
of
HCl
along
with
HNO3
in
the
leaching
acid
mixture
should
overcome
the
complexation
difficulties
demonstrated
in
nitric­
only
leaches.
Digestions
using
both
the
nitric­
only
Method
3015
and
those
using
the
nitric­
and­
hydrochloric
mixed
acid
Method
3015A
are
compared.
The
results
of
including
HCl
in
the
leaching
mixture
for
aqueous
samples
parallel
those
for
solid
samples.
Recoveries
of
problem
analyses
increase
while
recoveries
of
other
analyses
are
preserved.
In
addition,
validation
for
the
elements
boron,
mercury,
and
strontium
was
not
provided
for
Method
3015.
By
analyzing
for
these
elements
in
the
current
validation
study
of
EPA
Method
3015A,
possible
validation
for
these
three
RCRA­
regulated
elements
will
also
be
examined.

EXPERIMENTAL
The
validation
study
for
Method
3051A
consisted
of
an
initial
re­
optimization
of
the
acid
ratio
to
be
used
in
the
method.
This
was
accomplished
by
comparing
the
results
of
four
leaches
of
SRM
2704
(
Buffalo
River
sediment)
using
different
acid
combinations.
Initially,
10
mL
HNO3
only
was
used
(
10:
0).
Subsequent
acid
mixtures
consisted
of
9
mL
HNO3
+
1
mL
HCl
(
9:
1),
9
mL
HNO3
+
3
mL
HCl
(
9:
3),
and
9
mL
HNO3
+
5
mL
HCl
(
9:
5).
The
goal
of
these
acid
ratio
studies
was
to
determine
the
optimum
amount
of
HCl
for
maximum
complexation
while
still
achieving
similar
oxidizing
strength
of
the
acid
mixture.
Excessive
dilution
of
the
HNO3
with
the
HCl
would
lead
to
incomplete
recoveries
of
analyses,
thus
introducing
another
bias.
Also,
as
the
method
must
be
valid
for
all
26
RCRA­
regulated
elements,
it
is
not
appropriate
to
optimize
the
method
for
only
a
few
of
the
analyses
(
i.
e.
antimony
or
iron).
The
final
method
must
be
optimized
for
recovering
all
RCRA­
regulated
analyses.
Elemental
analysis
was
performed
either
by
Inductively­
Coupled
Plasma­
Mass
Spectrometry
(
ICP­
MS)
or
by
Flame
Atomic
Absorption
Spectrometry
(
FAAS).

Upon
re­
optimizing
the
acid
ratio,
the
remainder
of
the
validation
study
consisted
of
comparing
the
recoveries
using
the
nitric­
only
Method
3051
versus
using
the
nitric­
and­
hydrochloric
mixed
acid
3051A.
Comparisons
were
made
using
the
three
remaining
types
of
matrices
for
which
3051
is
applicable,
namely
sludge,
soil,
and
oil.
Recoveries
using
the
10:
0
versus
using
the
9:
3
digests
were
compared,
as
well
as
recoveries
from
the
hot­
plate
Method
3050.
Data
for
the
hot­
plate
method
was
taken
from
the
original
report
for
the
validation
of
Method
3050
and
3051
2.
The
results
demonstrated
by
the
9:
3
3051A
leach
method
suggest
that
it
is
a
valid
alternative
leach
method,
achieving
comparable
results
to
those
of
the
hot­
plate
leach
Method
3050.

The
validation
experiments
for
the
aqueous
leach
Method
3015A
paralleled
those
for
the
solid
sample
leach
method
3051A.
The
samples
used
were
simulated
wastewater
samples
prepared
just
prior
to
the
leach
digestion.
Simulated
wastewaters
were
used
because
of
the
lack
of
appropriate
standard
reference
materials
for
wastewater
type
matrices.
Additionally,
the
validation
study
for
the
original
aqueous
leach
method
3015
was
performed
using
a
simulated
wastewater.
The
wastewaters
were
prepared
from
the
same
SRM's
used
in
the
3051A
validation
study,
combining
approximately
0.35
grams
of
the
solid
samples
with
45
mL
double­
deionized
water
in
each
digestion
vessel
just
prior
to
acid
addition.
The
SRM's
used
were
SRM
2704
(
Buffalo
River
sediment),
SRM
4355
(
Peruvian
soil),
SRM
1084a
(
Wear
Metals
in
Oil),
and
a
1­
to­
1
mixture
of
SRM's
2704
and
1634c
(
Trace
Metals
in
Fuel
Oil).

Digestions
were
performed
for
the
wastewater
samples
using
either
5
mL
HNO3
only
(
Method
3015)
or
4
mL
HNO3
and
1
mL
HCl
(
3015A).
Elemental
analysis
was
performed
using
either
ICP­
MS
or
FAAS.
The
recoveries
of
analyses
using
each
acid
mixture
were
compared.
Once
again,
the
addition
of
hydrochloric
acid
provided
better
complexation
and
stabilization
of
problem
analyses
while
preserving
the
recoveries
of
other
analyses.
Also,
the
proposed
Method
3015A
has
demonstrated
that
it
is
effective
for
the
elements
boron,
mercury,
and
strontium.
Validation
for
these
three
RCRA­
regulated
elements
was
not
provided
in
the
original
validation
of
Method
3015.

RESULTS
AND
DISCUSSION
The
following
figures
illustrate
the
data
for
the
re­
optimization
of
the
3051A
acid
ratio.
As
discussed,
analyses
were
leached
from
SRM
2704
using
these
four
acid
ratios:
10:
0,
9:
1,
9:
3,
and
9:
5.
For
the
"
problem"
analyses
shown,
the
acid
ratio
of
9:
3
provides
optimum
complexation.
The
most
dramatic
case
is
for
antimony,
whose
recovery
increases
from
approximately
0%
in
the
nitric­
only
digest
to
greater
than
80%
in
the
9:
3
digest.
Other
recovery
increases
are
shown
for
iron
and
vanadium.
The
remaining
figures
demonstrate
that
the
9:
3
mixture
also
retains
the
overall
oxidizing
strength
of
the
acid
mixture,
hence
preserving
the
recoveries
of
the
other
analyses.
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
56
Figure
1.
Recovery
of
antimony
vs.
acid
ratio
Figure
2.
Recovery
of
iron
vs.
acid
ratio
used
used
in
leach
digest.
in
leach
digest.

Figure
3.
Recovery
of
vanadium
vs.
acid
ratio
used
in
leach
digest.

The
above
figures
illustrate
the
increased
recoveries
of
Sb,
Fe,
and
V
as
HCl
is
added
to
the
leaching
mixture.
Also
demonstrated
is
the
optimization
of
the
acid
ratio
at
9
mL
HNO3
to
3
mL
HCl.
The
analyte
recoveries
using
the
9:
5
acid
ratio
either
decrease
slightly,
or
provide
no
further
recovery
enhancement.
This
indicates
that
the
9:
5
ratio
is
not
as
appropriate
and
that
the
9:
3
ratio
offers
the
highest
level
of
complexation
without
compromising
the
overall
oxidizing
strength
of
the
acid
mixture.
The
figures
below
also
illustrate
that
the
oxidizing
strength
is
not
lowered,
as
the
recovery
of
non­
biased
analyses
remains
virtually
the
same
when
the
9:
3
ratio
is
used.

Figure
4.
Recovery
of
cobalt
vs.
acid
ratio
used
Figure
5.
Recovery
of
beryllium
vs.
acid
ratio
in
leach
digest.
used
in
leach
digest.

With
the
acid
ratio,
re­
optimized
at
9:
3
using
leach
data
from
SRM
2704,
the
remainder
of
the
validation
study
continued.
Three
remaining
matrices,
a
soil
(
SRM
4355),
an
oil
(
SRM
1084a),
and
a
simulated
sludge
(
1­
to­
1
mixture
of
SRM
2704
and
SRM
1643c),
were
digested
using
both
the
nitric­
only
Method
3051
and
the
nitric
and
hydrochloric
mixed­
acid
Method
3051A.
The
data
demonstrates
the
increased
recoveries
of
biased
analytes
with
preservation
of
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
57
other
analyte
recoveries.
The
recoveries
for
Method
3051A
are
comparable
to
those
of
Method
3050,
the
hot­
plate
digestion
that
uses
nitric
acid,
hydrochloric
acid,
and
hydrogen
peroxide.
The
following
tables
will
demonstrate
these
improved
recoveries
using
the
microwave
leach
and
their
comparability
to
the
recoveries
using
the
hot­
plate
leach.
Analyte
recoveries
are
expressed
in
µ
g/
g
unless
otherwise
indicated.

Table
1.
Comparison
of
analyte
recoveries
from
SRM
4355
(
Peruvian
Soil)
for
alternative
EPA
leach
methods
(
µ
g/
g
±
95%
confidence
interval,
unless
otherwise
noted)

129
±
26
131
±
14.2
130.6
±
5.1
126.7
±
9.3
Pb
28.9
±
2.8
17.1
±
2.37
18.7
±
0.68
14.6
±
0.68
Cr
(
1.50)**
1.03
±
0.202
1.09
±
0.27
1.11
±
0.31
Cd
4.45
±
0.19%
2.98
±
0.342%
3.28
±
0.18%
2.8
±
0.09%
Fe
(
151)**
81.4
±
17.3
123
±
7.78
65.5
±
2.12
V
14.8
±
0.76
­­­*
10.4
±
0.826
0.202
±
0.347
Sb
Total
3050
3051A
3051
Element
*
Antimony
was
not
analyzed
in
the
3050
validation
study.
**
Values
in
parenthesis
are
not
certified
concentrations
and
are
for
reference
only.

The
first
three
elements
demonstrate
the
enhanced
recoveries
upon
adding
3
mL
HCl
to
the
acid
mixture.
Especially
dramatic
is
the
enhancement
for
antimony,
showing
only
1%
recovery
for
the
nitric­
only
digest
but
rising
to
approximately
70%
upon
adding
3
mL
HCl.
The
remaining
entries
in
the
table
demonstrate
that
recoveries
of
other
analyses
are
preserved
when
the
mixed
acid
is
used.

Table
2.
Comparison
of
analyte
recoveries
from
SRM
1084a
(
Wear
Metals
in
Oil)
for
alternative
EPA
leach
methods
(
µ
g/
g
±
95%
confidence
interval)

101.1
±
1.3
100.8
±
7.8
99.8
±
6.5
Pb
101.4
±
1.5
96.4
±
7.0
92.5
±
3.1
Ag
100.3
±
1.4
93.2
±
3.4
90.3
±
2.4
Mo
100.0
±
1.9
91.8
±
3.6
92.2
±
4.2
Cu
99.7
±
1.6
92.9
±
4.6
93.6
±
2.0
Ni
99.5
±
1.7
96.6
±
4.9
92.9
±
6.3
Mg
(
104)*
101.1
±
6.6
98.0
±
4.8
Al
98.3
±
0.8
93.5
±
3.1
91.2
±
4.8
Cr
Total
3051A
3051
Element
*
Values
in
parenthesis
are
not
certified
concentrations
and
are
for
reference
only.

The
recoveries
of
all
analyses
for
the
oil
matrix
are
similar
for
both
acid
combinations.
However,
most
recoveries
for
the
mixed
acid
digest
are
slightly
higher
than
the
nitric­
only
digest.

Table
3.
Comparison
of
analyte
recoveries
from
"
simulated
sludge"
mixture
of
~
0.25
g
SRM
2704
(
Buffalo
River
Sediment)
and
~
0.25
g
SRM
1634c
(
Trace
Metals
in
Fuel
Oil)
for
alternative
EPA
leach
methods
(
µ
g/
g
±
95%
confidence
interval)

76.4
±
8.17
78.2
±
6.41
83.4
±
7.05
Pb
1.67
±
0.196
1.87
±
0.431
1.81
±
0.478
Cd
<
4.0
0.216
±
0.191
0.209
±
0.067
Ag
<
2.5
1.71
±
0.61
1.53
±
0.20
Mo
31.4
±
5.56
29.3
±
9.60
32.3
±
6.06
Ni
5.89
±
1.43
6.18
±
1.18
7.69
±
1.85
Co
42.3
±
6.04
44.7
±
2.99
49.7
±
3.73
Cr
3050
3051A
3051
Element
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
58
Again,
we
see
the,
preservation
of
recoveries
for
the
mixed
acid,
as
well
as
good
agreement
with
the
hot­
plate
3050
recoveries.
This
simulated
sludge
matrix
is
highly
heterogeneous,
which
may
account
for
the
greater
variation
seen
in
the
data
for
this
matrix.

The
data
presented
serve
as
the
final
results
of
the
validation
study
for
EPA
Method
3051A.
It
has
demonstrated
effectiveness
for
enhancing
the
recovery
of
"
problem"
analyses
from
four
different
matrix
types
while
at
the
same
time
preserving
the
recovery
of
the
remaining
analyses.
It
has
demonstrated
comparable
recoveries
with
hot­
plate
Method
3050.

Validation
for
proposed
Method
3015A
proceeded
using
similar
experimental
procedures.
Comparisons
were
made
using
an
all­
nitric
digest
(
5
mL
HNO3)
versus
using
a
nitric­
and­
hydrochloric
mixed
acid
digest
(
4
mL
HNO3
+
1
mL
HCl)
on
four
"
simulated­
wastewater"
matrices.
These
matrices
consisted
of
approximately
0.35
grams
of
solid
sample
mixed
with
45
mL
double­
deionized
water
(
per
digestion
vessel)
immediately
prior
to
acid
addition
and
digestion.
The
solid
samples
used
parallel
those
used
for
the
3051a
study,
namely
SRM
2704,
SRM
4355,
SRM
1084a,
and
the
mixture
of
SRM
2704
and
SRM
1634c.
The
mixed
acid
digest
should
complex
and
stabilize
certain
problem
analyses,
leading
to
increased
recoveries
of
these
analyses.
The
leach
recovery
data
for
the
2704­
wastewater
and
the
4355­
wastewater
are
presented
in
the
following
tables.

Table
4.
Comparison
of
analyte
recoveries
from
"
simulated
wastewater"
mixture
of
SRM
2704
(
Buffalo
River
Sediment)
and
double­
deionized
water
for
EPA
leach
methods
(
µ
g/
g
±
95%
confidence
interval)

­­­*
0.36
±
0.115
0.33
±
0.037
Ag
44.1
±
3.0
39.9
±
4.64
38.6
±
4.24
Ni
438
±
12
406
±
8.7
392
±
5.6
Zn
14.0
±
0.6
11.9
±
0.45
11.0
±
0.47
Co
­­­*
0.95
±
0.170
0.85
±
0.116
Be
­­­*
2.92
±
1.19
2.07
±
1.71
Mo
95
±
4
51.6
±
2.23
32.6
±
4.83
V
Total
3015A
3015
Element
*
The
total
concentration
of
this
analyte
is
not
certified
for
SRM
2704.

The
results
for
this
"
sediment­
wastewater"
matrix
demonstrate
the
enhanced
recovery
for
certain
analyses
(
i.
e.
vanadium,
molybdenum,
and
beryllium)
and
similar
recoveries
for
other
analyses
using
the
nitric­
and­
hydrochloric
mixed
acid
Method
3015A
digest.

Table
5.
Comparison
of
analyte
recoveries
from
"
simulated
wastewater"
mixture
of
SRM
4355
(
Peruvian
Soil)
and
double­
deionized
water
for
EPA
leach
methods
(
µ
g/
g
±
95%
confidence
interval)

129
±
26
136
±
4.7
135
±
4.6
Pb
(
1.7)*
1.09
±
0.13
0.98
±
0.07
Mo
(
13)*
10.8
±
2.40
11.1
±
1.44
Ni
368
±
8.2
418
±
16.4
415
±
16.2
Zn
14.8
±
0.76
10.4
±
0.51
10.5
±
0.42
Co
(
1.9)*
1.62
±
0.14
1.31
±
0.16
Ag
14.3
±
2.2
5.20
±
0.67
3.70
±
0.37
Sb
(
151)*
38.7
±
5.94
33.9
±
3.57
V
Total
3015A
3015
Element
*
Values
in
parenthesis
are
not
certified
concentrations
and
are
for
reference
only.

This
set
of
digestions
also
demonstrates
that
inclusion
of
HCl
enhances
the
recovery
of
certain
analyses
(
V,
Sb,
Ag)
while
preserving
the
recovery
of
other
analyses.
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
59
Two
additional
types
of
simulated
wastewater
matrices
are
currently
in
the
digestion
and
data
acquisition
process
for
the
validation
of
proposed
Method
3015A.
These
are
the
oil­
wastewater
(
SRM
1084a
and
DDI
water)
and
the
"
sludge"­
wastewater
(
SRM's
2704
and
1634c
and
DDI
water).
It
is
expected
that
similar
enhancements
in
"
problem"
analyte
recoveries,
as
well
as
preservation
of
other
analyte
recoveries
will
be
demonstrated
by
the
mixed
acid
digest.

SUMMARY
The
final
validation
data
for
proposed
EPA
Method
3051A
has
been
given.
The
data
shows
the
effectiveness
of
including
HCl
in
the
leaching
acid
mixture
for
enhancing
the
recoveries
of
"
problem"
RCRA­
regulated
metals
such
as
antimony
and
iron.
This
mixed
acid
also
shows
comparable
recoveries
for
other
non­
biased
analyses.
This
updated
microwave
leach
method
minimizes
the
reagent­
induced
biases
between
the
two
alternative
leach
methods
3050
(
hot­
plate)
and
3051
(
microwave).
This
study
also
demonstrates
the
effectiveness
of
adding
HCl
in
Method
3015A
for
the
leaching
of
aqueous
samples
(
wastewaters).
Data
for
digestions
of
two
matrix
types
using
the
mixed­
acid
leach
Method
3015A
has
demonstrated
similar
enhancements
in
recoveries.
Digestions
for
the
remaining
two
wastewater
matrix
types
are
in
progress,
with
similar
enhancements
in
the
recoveries
of
problem
analyses
expected.

A
paper
explaining
the
depth
of
the
information
in
this
validation
study
is
in
preparation.
The
discussion
will
include
chemical
equilibria
and
other
data
documenting
the
reactions
and
summarizing
these
results.

REFERENCES
1.
Link,
D.
D.;
Walter,
P.
J.;
Kingston,
H.
M.;
"
EPA
Methods
3015A
and
3051A:
Validation
Studies
for
Updated
Microwave
Leach
Methods",
Paper
1121,
Pittsburgh
Conference,
March,
1997,
Atlanta,
GA.
2.
Binstock,
D.
A.,
Yeager,
W.
M.,
Grohse,
P.
M.,
Gaskill,
A.;
"
Validation
of
a
Method
for
Determining
Elements
in
Solid
Waste
by
Microwave
Digestion",
Technical
Report
Draft,
November,
1989.

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
LONG­
TERM
STABILITY
OF
ICP
SPECTRAL
REGISTRATION
BY
MANAGEMENT
OF
THE
MODELS:
APPLICATION
TO
QUANTITATION
USING
MULTIVARIATE
ANALYSIS
Christopher
P.
Hanna,
Ph.
D.,
Senior
Staff
Scientist;
Alan
M.
Ganz,
Ph.
D.,
Director;
Clinical
Instrumentation,
Perkin­
Elmer
Corporation,
50
Danbury
Road,
Wilton,
CT
06897­
0208
ABSTRACT
A
mathematical
technique
is
described
which
provides
constancy
of
multivariate
analysis
models
in
ICP
emission
spectroscopy.
Once
an
analysis
protocol
has
been
developed,
the
profiles
of
all
active
spectral
components
are
saved
in
an
archive.
Through
the
use
of
well­
defined
spectral
standards,
this
archive
can
be
mathematically
transformed
at
any
time
to
correspond
to
any
new
conditions
for
an
instrument.
This
results
in
analytical
measurements
which
are
completely
independent
of
wavelength
accuracy
and
wavelength
calibration,
providing
not
only
exceptional
long­
term
accuracy,
but
also
complete
transferability
of
ICP
emission
methods
between
instruments.
Preliminary
analytical
results
are
shown
which
illustrate
the
utility
of
this
technique.

INTRODUCTION
ICP
emission
spectroscopy
is
frequently
applied
for
the
determination
of
metals
and
metalloids
in
a
wide
variety
of
environmental
matrices.
Mathematical
procedures
are
often
required
for
the
correction
of
background
shifts
and
spectral
overlaps,
and
are
well
represented
in
promulgated
regulations
(
particularly
the
use
of
interelement
correction
factors,
or
IECs).
While
the
IEC
technique
works
properly
in
well
defined
circumstances,
it
does
not
lend
itself
well
to
all
situations1.
Some
examples
are:
1)
IECs
require
the
careful
selection
of
background
correction
points,
which
often
requires
a
lengthy
and
somewhat
subjective
selection
process;
2)
slight
differences
in
plasma
conditions
in
the
course
of
an
analysis
can
render
the
IEC
factors
nonviable
due
to
changing
line
intensity
ratios;
3)
IEC
factors
are
not
always
linear
across
all
concentration
ranges
of
an
interferent.

Multivariate
calibration
techniques
have
been
used
in
ICP
emission
spectroscopy
for
interference
correction,
particularly
when
array­
based
solid­
state
detectors
are
used2­
4.
Multivariate
calibration
techniques
provide
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
60
exceptional
capacity
for
resolving
spectral
overlaps
in
a
straightforward
manner2,
and
provide
improved
analytical
precision
through
the
process
of
redundant
sampling3.
Numerous
suggestions
for
improved
multivariate
methods
have
been
made
during
the
most
recent
public
comment
period
for
ICP
emission
Method
6010B.
These
multivariate
methods
are
expected
to
be
given
strong
consideration
in
the
final
version
of
the
Method.

All
spectroscopic
interference
correction
techniques
require
that
the
mathematical
correction
models
accurately
reflect
current
wavelength
axis
conditions.
In
the
case
of
ICP
emission
analyses,
this
is
independent
of
whether
IECs
or
multivariate
techniques
are
used,
and
is
the
case
for
array­
based
and
non­
array
detection.
For
example,
if
the
spectral
peak
for
an
interferent
occurs
at
one
wavelength
on
day
one
and
at
a
slightly
different
wavelength
on
day
two,
the
interference
correction
models
must
be
redefined
(
or
"
rebuilt")
to
reflect
these
new
conditions.
For
this
reason,
the
true
requirements
for
wavelength
accuracy
and
stability
are
largely
unrealistic.

A
unique
aspect
of
multivariate
calibration
is
that
spectral
shape
information
is
part
of
the
correction
model.
By
contrast,
IECs
employ
wavelength
position
and
relative
intensities
but
ignore
shape.
Shape
information
provides
certain
advantages
over
IECs
in
terms
of
analytical
precision,
accuracy,
etc.,
and
have
been
presented2,
4.
Of
particular
significance
is
that
shape­
based
correction
models
can
be
mathematically
"
shifted"
(
through
mathematical
interpolation)
once
they
are
built.
Thus,
any
particular
wavelength
condition
can
be
accommodated.
This
capacity
adds
a
number
of
hitherto
unachievable
features
to
ICP
emission
determinations.
Interference
correction
models,
and
hence
analytical
accuracy,
are
independent
of
wavelength
conditions.
Furthermore,
true
transferability
of
analytical
methods,
leading
to
"
turnkey"
conditions,
is
achievable.
Ultimately,
the
potential
for
transferring
spectral
archives
between
instruments,
thus
creating
more
"
absolute"
analytical
conditions
within
ICP
emission
analyses,
will
be
within
reach.

Example
data
for
the
technique
described
above
will
be
presented.
This
example
was
selected
due
to
its
strong
illustration
of
the
necessity
for
wavelength­
model
coherence.
The
implications
for
this
technique
will
also
be
presented
and
discussed.

EXPERIMENTAL
A
Perkin­
Elmer
Optima
3000
DV
was
used
for
all
experiments.
A
cross­
flow
nebulizer
(
1
mL/
min)
in
a
Scott­
type
double­
pass
spray
chamber
was
used
for
sample
introduction,
and
plasma
conditions
were:
0.8
mL/
min
nebulizer
flow;
0.5
mL/
min
auxiliary
flow;
15.0
mL/
min
plasma
flow;
1450
W
power.
All
spectra
were
collected
with
the
"
normal"
spectral
resolution
setting.
For
collection
of
the
archival
model
spectra,
slit­
scanning
was
applied
for
the
purposes
of
giving
data
densities
high
enough
for
interpolative
model
shifting.
Conversely,
the
sample
data
to
which
the
models
were
applied
was
not
slit­
scanned,
as
the
additional
data
density
is
not
required.
All
data
were
collected
using
ICP
WinLab
software
operating
on
a
Digital
Pentium
100
MHz
personal
computer.
All
mathematical
operations
were
programmed
either
in
MatLab
(
MathWorks,
Natick,
MA)
or
in
Microsoft
Excel
(
Microsoft
Corporation,
Redmond,
WA).

Pure
component
spectra
(
molybdenum,
chromium,
vanadium)
were
collected
by
diluting
and
analyzing
appropriate
concentrations
of
PE
Pure
standard
solutions.
These
pure
component
solutions
were
then
combined
to
provide
an
appropriate
test
sample.
All
data
was
collected
at
the
spectral
region
of
the
Optima
detector
corresponding
to
270.028­
270.160
nm.
A
wavelength
monitoring
solution
was
prepared
which
provided
isolated
lines
over
representative
regions
of
the
detector
surface.
Upon
running
this
wavelength
monitoring
solution,
it
is
possible
to
calculate
current
wavelength
registration
status
for
any
and
all
regions
of
the
detector
surface.

RESULTS
AND
DISCUSSION
The
test
sample's
individual
components
(
each
scaled
to
approximate
equal
intensity)
are
shown
at
the
top
of
Figure
1.
This
combination
of
spectral
components,
particularly
at
the
levels
shown,
is
most
useful
for
this
demonstration
since
very
small
changes
in
wavelength
registration
can
lead
to
significant
analytical
errors
(>
10%).
The
reason
for
this
high
degree
of
dependence
on
wavelength
registration
is
that
the
analyte
(
vanadium)
is
positioned
on
the
shoulder
of
interferent
1
(
chromium),
which
in
turn
is
positioned
on
the
shoulder
of
interferent
2
(
molybdenum),
and
is
clearly
shown
at
the
bottom
of
Figure
1.
Furthermore,
both
interferents
are
present
at
levels
in
great
excess
(
1,000­
fold)
of
the
analyte.
This
spectral
positioning
results
in
an
amplification
of
analytical
error
brought
about
by
changes
in
wavelength
registration.
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
61
Figure
1.
Vanadium,
chromium
and
molybdenum
components;
individually
and
mixed.

Once
the
multivariate
models
were
built
with
the
pure
spectral
components,
the
mixed
sample
was
analyzed
and
100%
recovery
of
the
vanadium
analyte
was
achieved.
After
24
hours,
the
thermal
conditions
of
the
instrument
were
deliberately
changed
such
that
the
wavelength
registration
was
altered.
The
sample
mixture
was
reanalyzed,
and
there
was
a
dramatic
loss
in
analyte
recovery.
At
this
time
the
degree
of
change
in
wavelength
registration
was
measured
by
analyzing
the
wavelength
monitoring
solution.
The
pure
archival
spectra
which
make
up
the
multivariate
models
were
then
"
shifted"
through
interpolation,
to
an
extent
equal
to
that
determined
with
the
wavelength
monitoring
solution.
For
example,
if
the
wavelength
monitoring
solution
indicates
a
change
in
wavelength
registration
equal
to
x,
then
the
archival
component
spectra
are
interpolatively
"
shifted"
x.
Once
shifted,
the
multivariate
models
are
rebuilt
and
used
for
subsequent
samples.
In
the
case
of
the
sample
described
above,
the
shifting
of
the
model
spectra
resulted
in
complete
restoration
of
100%
analyte
recovery.
The
results
of
these
operations
are
shown
in
Table
1.

Table
1.
Analytical
results
for
vanadium
in
a
highly
interfering
matrix;
with
and
without
wavelength
registration
compensation
100
%
Induced
drift,
shifted
models
24
hours
20
%
Induced
drift
24
hours
100
%
­­­
0
hours
V
recovery
(%)
Condition
Time
CONCLUSIONS
The
use
of
this
technique
shows
that
it
is
possible
to
perform
ICP
emission
analyses
in
a
manner
that
is
independent
of
the
wavelength
axis.
Provided
there
is
a
means
of
determining
current
wavelength
registration
status,
it
is
possible
to
update
multivariate
calibration
models
to
the
newer
status,
thus
eliminating
wavelength
registration
changes
as
a
source
of
miscalculation.
It
is
then
possible
to
have
true
transferability
of
analytical
methods,
as
well
as
the
possible
transfer
of
spectral
component
archives.

REFERENCES
1.
G.
A.
Laing,
et
al.
in
Proceedings:
Tenth
Annual
Waste
Testing
and
Quality
Assurance
Symposium,
July
11­
15,
1994,
Arlington,
VA.
2.
J.
C.
Ivaldi
and
T.
W.
Barnard,
Spectrochimica
Acta,
1993,
48B,
1265.
3.
J.
C.
Ivaldi
and
J.
F.
Tyson,
Spectrochimica
Acta,
1995,
50B,
1207.
4.
J.
C.
Ivaldi,
et
al.
in
Proceedings:
Eleventh
Annual
Waste
Testing
and
Quality
Assurance
Symposium,
July
23­
28,
1995,
Washington,
DC.
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
62
ORGANIC
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
63
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
64
SOLID
PHASE
MICROEXTRACTION
PREPARATIVE
APPLICATIONS
IN
THE
ANALYSIS
OF
ORGANIC
COMPONENTS
IN
RADIOACTIVE
WASTES
John
E.
Young
and
Stephen
L.
Crump
Westinghouse
Savannah
River
Company,
Aiken,
South
Carolina
ABSTRACT
The
analytical
chemistry
of
radioactive
materials
is
complicated
by
numerous
challenges.
Primarily,
the
radiation
level
of
material
of
interest
may
be
sufficient
to
prohibit
direct
human
contact.
Additionally,
materials
used
in
the
analysis
may
result
in
the
production
of
mixed
waste
streams,
which
contain
radionuclides
in
addition
to
hazardous
wastes.

Solid
Phase
Microextraction
(
SPME),
allows
preparation
of
aqueous
radioactive
samples
for
analysis
without
the
introduction
of
hazardous
solvents
while
simultaneously
achieving
bulk
separation
from
the
radioactive
components
in
the
sample.

SPME
does
not
generate
conventional
extraction
solvent
residues
which
are
regulated
by
federal
and
state
agencies
as
"
hazardous"
wastes
These
solvent
wastes
typically
require
treatment
by
incinerators
operating
under
state
and
federal
operating
permits.
Utilization
of
SPME
eliminates
the
need
for
an
auxiliary
solvent,
and
the
subsequent
need
for
treatment
and
disposal
of
the
solvent
waste
stream.

Isolation
of
organic
materials
from
sample
matrices
by
traditional
extraction
techniques
has
been
extensively
studied
and
developed
to
optimize
analyte
recovery
and
removal
of
interfering
components.
Radioactive
components
in
samples
may
make
these
methods
inappropriate
or
inapplicable.
Methods
which
require
lengthy
contact
with
the
sample
may
be
prohibitive
due
to
radiation
exposure.
SPME
is
a
technique
which
reduces
the
effects
of
varying
sample
matrices
and
simultaneously
achieves
decontamination
from
the
inorganic
radioactive
components.

The
advantages
of
SPME
as
a
sample
preparation
technique
for
the
analysis
of
organic
constituents
in
radioactive
wastes
are
summarized,
with
data
presented
on
the
demonstration
of
these
advantages
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
SOLID
PHASE
EXTRACTION
APPLICATIONS
IN
THE
SAMPLING
OF
ORGANIC
COMPONENTS
IN
RADIOACTIVE
WASTES
Stephen
L.
Crump,
John
E.
Young,
David
T.
Hobbs,
and
Matthew
E.
Jamison
Westinghouse
Savannah
River
Company,
Aiken,
South
Carolina
ABSTRACT
The
recent
advancement
and
maturation
of
the
technology
used
in
solid
phase
extraction
leads
to
new
and
interesting
opportunities
in
the
analytical
chemistry
of
radioactive
sample
matrices.
Solid
phase
extraction
(
SPE)
disks
are
readily
applicable
to
the
sampling
of
surface
films
in
radioactive
materials
which
must
be
accessed
and
handled
remotely.

Remote
sampling
of
unknown
surface
films
from
liquid
radioactive
waste
storage
tanks
at
the
Savannah
River
Site
using
SPE
disks
has
been
shown
to
accomplish
sampling
objectives
while
reducing
radiation
exposure
from
the
resulting
samples.

Conventional
sampling
of
surface
films
is
limited
by
the
ability
to
isolate
the
uppermost
layer
of
the
liquid
into
the
sampling
container.
This
inability
is
complicated
by
the
hazards
associated
with
exposures
to
highly
radioactive
materials.
Sampling
of
these
materials
must
be
done
remotely
to
minimize
radiation
exposure.
Most
often,
this
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
65
remote
handling
is
performed
with
mechanical
devices
such
as
cable
activated
sample
grabbers,
robotic
instruments,
or
conventional
"
dip­
bucket"
grab
samplers.

In
order
to
selectively
sample
the
uppermost
layer
of
the
liquid,
the
sampling
device
must
be
accurately
positioned
to
avoid
collecting
lower
layers
of
the
tank
contents.
In
most
applications,
grab
sampling
is
adequate
to
capture
an
adequate
quantity
of
the
surface
film
for
analysis.
However,
in
the
case
of
HLW
storage
tank
sampling,
obtaining
an
adequate
sample
is
extremely
expensive
and
time
consuming
because
of
the
high
radiation
field.

A
properly
conditioned
SPE
disk
has
the
ability
to
pass
through
the
surface
film
of
a
liquid
while
selectively
absorbing
the
organic
species
of
interest.
Samples
obtained
by
this
method
sorb
little
of
the
aqueous
liquid
phase
and
can
be
sufficiently
rinsed
free
of
most
of
the
radioactive
contaminants
to
allow
human
contact
while
preserving
the
analysts
ability
to
characterize
the
organic
components
in
the
film.

Recent
experiences
using
SPE
disk
sampling
methods
at
the
Savannah
River
Site
will
be
presented,
summarizing
the
potential
advantages
of
the
technique.

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
A
NEW
PULSED
FLAME
PHOTOMETRIC
DETECTOR
FOR
THE
ANALYSIS
OF
PESTICIDES
Ronald
D.
Snelling
OI
Analytical,
PO
Box
9010,
College
Station,
Texas
77842­
9010
(
409)
690­
1711
Many
of
the
commonly
pesticides
contain
sulfur,
nitrogen,
or
phosphorus
heteroatoms.
In
many
of
the
pesticides
more
than
one
of
the
heteroatoms
is
present.
There
are
a
variety
of
detectors
used
to
analyzed
these
pesticides,
but
all
of
these
detectors
have
serious
limitations.
Flame
photometric
detectors
suffer
from
interferences
and
quenching.
Nitrogen
­
phosphorus
detector
have
an
unstable
baseline
and
a
short
active
element
lifetime.
Chemiluminescence
detectors
are
very
selective
and
sensitive
but
are
complex
and
difficult
to
maintain
at
peak
performance.

The
pulsed
flame
photometric
detector
addresses
these
problems.
The
emissions
from
the
flame
are
time
resolved
are
well
as
wavelength
resolved,
so
the
detector
is
very
selective.
The
increase
in
selectivity
increased
the
sensitivity
of
the
detector.
The
pulsed
flame
photometric
detector
is
less
affected
by
quenching
than
a
conventional
flame
photometric
detector.
The
detector
is
stable
over
a
long
period
of
time
and
requires
minimum
maintenance.

The
pulsed
flame
photometric
detector
has
dual
channel
data
acquisition
capability.
Chromatograms
for
two
different
elements
may
be
collected
simultaneously
from
one
injection
and
one
column.
This
can
be
a
very
powerful
tool
for
the
confirmation
of
the
identity
of
compounds
containing
more
than
one
heteroatom.

The
pulsed
flame
photometric
detector
offers
significant
advantages
in
selectivity,
sensitivity,
low
maintenance,
and
the
ability
to
acquire
two
independent
chromatograms
simultaneously.

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
EASIER
AND
FASTER
GC/
ECD
ANALYSES
OF
PESTICIDES
AND
PCB'S
Susan
M.
Brillante
Hewlett­
Packard
Co.,
2850
Centerville
Road,
Wilmington,
DE
19808­
1610
(
302)
633­
8616
The
generation
of
environmental
data
for
pesticides
and
PCB's
in
various
matrices
can
be
a
very
time­
consuming
process
for
laboratories
and
engineering
firms.
In
order
to
keep
a
GC/
ECD
system
operating
within
control
limits,
precious
analytical
time
must
be
spent
on
tasks
such
as:
recalibrations,
reinjection
of
samples,
cleaning
detectors,
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
66
reintegration
of
chromatography
peaks,
etc.
All
of
these
tasks
take
time
away
from
running
billable
samples,
and
adversely
affect
the
throughput
or
profitability
of
the
lab.

Hewlett­
Packard
has
developed
a
"
new"
micro­
ECD
that
shows
improved
performance
in
several
key
areas:
increased
linear
working
range,
increased
sensitivity,
more
robust
and
less
prone
to
contamination.
The
practical
result
of
these
features
is
that
more
GC
analysis
time
can
be
spent
analyzing
billable
samples
instead
of
cleaning
up
and
recalibrating
the
system.
In
addition,
this
detector
is
compatible
with
small
id
columns
for
fast
chromatography.

Data
will
be
presented
from
several
laboratories
to
illustrate
each
of
these
features.
In
addition,
some
practical
examples
of
how
profitability
(
sample
throughput)
can
be
increased
by
the
use
of
fast
GC
will
be
shown.

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
THE
DEVELOPMENT
OF
AN
ION
CHROMATOGRAPHY
METHOD
TO
MONITOR
ORGANIC
AND
INORGANIC
INDICATORS
OF
INTRINSIC
BIOREMEDIATION
AT
HAZARDOUS
WASTE
SITES
R.
Rediske,
Ph.
D.,
A.
Stiop,
D.
Van
Dyke,
Ph.
D.,
and
P.
Durda
Water
Resources
Institute,
Grand
Valley
State
University,
Allendale,
MI.,
49401
(
616­
895­
3047)
(
redisker@
gvsu.
edu).

ABSTRACT
Many
studies
have
demonstrated
that
the
natural
attenuation
processes
present
at
certain
landfills
and
hazardous
waste
sites
can
effectively
limit
the
migration
of
contaminant
plumes.
Indigenous
microbial
populations
present
in
aquifer
and
soil
systems
often
control
contaminant
migration
by
metabolic
or
co­
metabolic
degradation
mechanisms.
This
process
is
defined
as
intrinsic
bioremediation
and
has
been
utilized
as
a
plume
management
strategy
at
many
DOD,
NPL,
and
state
hazardous
waste
sites.
During
the
intrinsic
bioremediation
process,
contaminants
are
degraded
to
CO2
and
simple
organic
acids
such
as
formic,
acetic,
and
propionic.
Electron
acceptor
compounds
such
as
sulfate
and
nitrate
are
also
involved
and
undergo
reduction
to
sulfide
and
ammonia.
The
demonstration
and
implementation
of
intrinsic
bioremediation
depends
on
a
comprehensive
characterization
of
site
conditions
and
a
long
term
monitoring
program
that
documents
plume
containment
and
reduction.
The
analysis
of
organic
acid
intermediates
and
electron
acceptor
compounds
are
important
components
of
site
characterization
and
monitoring
programs.

This
paper
describes
the
development
and
validation
of
a
single
ion
chromatography
method
for
the
analysis
of
low
molecular
weight
organic
acids
and
electron
acceptor
compounds
in
groundwater
samples.
Method
9056
cannot
be
used
for
all
of
these
parameters
due
to
limitations
of
the
carbonate
eluent.
The
method
described
uses
a
Dionex
AS­
11
column
with
a
nonlinear
sodium
hydroxide
gradient
from
0.35
mM
to
26.5
mM.
Compounds
analyzed
by
the
method
include:
fluoride,
chloride,
formate,
acetate,
propionate,
nitrate,
sulfate,
phosphate,
pyruvate,
fumarate,
benzoate,
and
succinate.
The
method
was
validated
by
using
deionized
water,
surface
water,
and
groundwater
as
test
matrices.
Method
detection
limits
for
the
compounds
of
interest
range
from
0.01
mg/
l
to
0.05
mg/
l.
Highly
buffered
groundwaters
were
found
to
interfere
with
the
analysis
of
fluoride,
acetate,
and
formate.
This
interference
was
removed
by
the
use
of
a
modified
pretreatment
resin
cartridge
(
Dionex
OnGuard­
Ag
 
)
.
A
complete
separation
of
the
above
analyses
requires
approximately
20
min.

The
method
was
also
tested
on
groundwater
samples
from
a
petroleum
release,
a
landfill,
and
a
spill
of
chlorinated
solvents.
Precision
and
accuracy
results
for
MS/
MSD
analyses
were
70%
­
110%.
Data
for
organic
acids
and
electron
acceptors
were
important
components
in
the
demonstration
of
intrinsic
bioremediation
at
the
above
sites.
Laboratories
with
ion
chromatography
equipment
can
readily
implement
the
method.
It
provides
similar
performance
to
Method
9056
for
common
anions.

INTRODUCTION
In
recent
years,
intrinsic
bioremediation
has
become
increasingly
accepted
as
a
plume
management
strategy
for
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
67
organic
compounds
dissolved
in
groundwater.
Intrinsic
bioremediation
is
defined
as
the
sum
of
naturally
occurring
biodegradation,
dispersion,
dilution,
sorption,
and
chemical
transformation
reactions
that
effectively
reduce
the
toxicity,
mobility,
and
volume
of
a
contaminant
to
levels
that
are
protective
of
human
health
and
the
environment1.
Protocols
for
data
collection
and
analysis
are
available
for
fuels,
aromatic
hydrocarbons,
and
chlorinated
solvents2,3.
Central
to
the
evaluation
of
natural
attenuation
is
a
rigorous
monitoring
program
that
demonstrates
the
reduction
of
a
contaminant
and
the
presence
of
indicators
that
reflect
the
degradation
process.
Inorganic
anions
such
as
nitrate
and
sulfate
are
important
investigative
parameters
as
they
serve
as
electron
acceptors
in
the
anaerobic
degradation
process
of
chlorinated
and
aromatic
solvents4,5.
These
compounds
are
reduced
to
ammonia
and
sulfide
by
microorganisms
during
anaerobic
degradation.
Other
anions
such
as
fluoride
and
chloride
can
be
used
as
conservative
ions
to
determine
diffusion
and
dispersion6.
Low
molecular
weight
organic
acids
such
as
acetate,
formate,
and
propionate
are
also
important
indicators
of
the
degradation
process
as
they
represent
intermediates
in
the
metabolic
and
co­
metabolic
processes
that
convert
organic
solvents
to
CO2
and
CH4
2,3,5.
While
physical/
chemical
mechanisms
such
as
volatilization,
sorption,
and
dispersion
can
account
for
the
reduction
of
contaminants
in
groundwater,
the
presence
of
organic
acids
provides
direct
evidence
of
the
biological
degradation
process.

Inorganic
anions
can
be
analyzed
by
a
variety
of
wet
chemical
and
ion
chromatography
methods7.
Ion
chromatography
can
also
be
used
for
the
measurement
of
low
molecular
weight
organic
acids8,9,10.
The
ion
chromatography
method
listed
in
SW­
846.
Method
9025,
cannot
be
effectively
used
for
the
determination
of
organic
acids
due
to
limitations
of
the
carbonate
eluent.
This
paper
describes
the
development
of
an
ion
chromatography
method
that
can
be
used
for
the
simultaneous
analysis
of
the
inorganic
anions
and
organic
acids
commonly
measured
in
the
demonstration
of
intrinsic
bioremediation
at
hazardous
waste
sites.
The
method
includes
a
sample
pretreatment
step
that
removes
matrix
interference
problems
related
to
high
levels
of
chloride
and
alkalinity.
Data
for
precision,
accuracy,
and
method
detection
limit
are
presented
along
with
information
related
to
the
application
of
ion
chromatography
methods
to
the
demonstration
of
intrinsic
bioremediation.

EXPERIMENTAL
The
ion
chromatography
method
was
developed
on
a
Dionex
DX­
500
equipped
with
an
electrochemical
detector
operated
in
conductivity
mode.
A
complete
description
of
instrumental
conditions
is
presented
in
Table
1.
The
NaOH
gradient
described
in
the
Dionex
Technical
Bulletin11
was
modified
to
provide
an
improved
separation
of
early
and
middle
eluting
analyses.
The
gradient
program
is
described
in
Table
1.

A
chromatogram
showing
the
separation
of
the
inorganic
anions
and
organic
acids
is
shown
in
Figure
1.
While
acceptable
chromatography
was
obtained
in
samples
with
low
alkalinity,
the
concentrations
of
carbonate
and
bicarbonate
typically
found
in
natural
waters
resulted
in
the
loss
of
resolution
of
fluoride,
acetate,
formate,
and
propionate.
The
concentration
of
carbonate
species
in
natural
water
is
sufficient
to
change
the
acid/
base
chemistry
of
the
dilute
NaOH
eluent
at
the
beginning
of
the
gradient
program.
Carbonate
also
elutes
as
an
interfering
peak
in
the
middle
of
the
analysis.
A
typical
chromatogram
showing
the
effect
of
carbonate/
bicarbonate
on
the
analysis
of
organic
acids
and
inorganic
anions
is
shown
in
Figure
2.
High
levels
of
chloride
can
also
interfere
with
the
analysis
since
this
anion
is
often
present
in
the
100
mg/
l
range
while
organic
acids
are
found
at
50
µ
g/
l
concentrations.

In
order
to
eliminate
the
interference
problems
related
to
sample
alkalinity
and
chloride,
a
modified
sample
pretreatment
system
was
developed.
The
use
of
pretreatment
cartridges
for
the
removal
of
interferences
in
ion
chromatography
has
been
described
previously12.
Cartridges
are
commercially
available
for
the
removal
of
interferences
from
high
levels
of
carbonate,
chloride,
sulfate,
and
organic
compounds.
To
remove
carbonate/
bicarbonate
and
chloride
interferences,
two
cartridges
used
in
series
would
be
required.
A
silver
resin
cartridge
would
first
be
required
to
remove
chloride
interferences.
The
sample
would
then
be
treated
with
a
protonated
resin
cartridge
to
reduce
the
pH
to
4
and
be
sparged
with
helium
to
remove
CO2.
These
cartridges
are
generally
viewed
as
disposable
items
and
must
be
added
to
the
cost
of
analysis.
As
a
component
of
the
development
of
this
method,
a
Dionex
OnGuard­
Ag
 
cartridge
was
modified
by
rinsing
with
a
small
amount
of
5%
nitric
acid.
During
this
process
the
resin
becomes
partially
protonated
while
retaining
some
sites
substituted
with
silver.
The
preparation
procedure
for
the
modified
resin
is
described
in
Figure
3.
Cartridges
can
be
used
for
multiple
injections
by
rinsing
with
150
ml
of
deionized
water
at
a
rate
of
10
ml/
min.
It
is
recommended
that
the
cartridge
be
regenerated
after
10
samples
according
to
the
procedure
outlined
in
Figure
3.
We
have
noticed
no
degradation
of
cartridge
performance
or
carry
over
when
this
sequence
is
followed.
The
performance
of
the
modified
cartridge
was
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
68
compared
to
the
Dionex
OnGuard­
H
 
which
functions
as
a
protonated
resin
to
remove
carbonate
interferences.

Table
1.
Instrumental
Conditions
for
the
Analysis
of
Inorganic
Anions
and
Organic
Acids
0.2
to
10
mg/
l
0.02
to
1
mg/
l
0.02
to
1
mg/
l
0.2
to
10
mg/
l
0.02
to
1
mg/
l
0.02
to
1
mg/
l
0.02
to
1
mg/
l
0.02
to
1
mg/
l
0.05
to
5
mg/
l
0.05
to
5
mg/
l
0.02
to
1
mg/
1
0.02
to
1
mg/
l
0.02
to
1
mg/
l
Chloride
Fluoride
Nitrate
Sulfate
Phosphate
Acetate
Formate
Propionate
Benzoate
Pyruvate
Glutarate
Succinate
Fumarate
Analytes
(
concentration
range)
Dionex
DX­
500
Dionex
ED40
Electrochemical
Conductivity
Mode
AS­
11
with
AG­
11
Guard
ASRS­
1
100
µ
l
2
ml/
min
NaOH
0.35
mM
­
26.5
mM
Eluent
A
Deionized
Water
Eluent
B
100
mM
NaOH
Eluent
C
5
mM
NaOH
Time
%
A
%
B
%
C
0
min
93
0
7
2
min
93
0
7
8
min
30
0
70
20
min
7
23
70
21
min
93
0
7
Instrument
Detector
Column
Suppressor
Injection
Flow
Gradient
For
sample
analysis,
the
modified
cartridge
was
rinsed
with
10
ml
of
deionized
water.
A
10
ml
aliquot
of
the
sample
was
then
passed
through
the
cartridge.
The
first
3
mls
were
discarded
and
the
remaining
7
mls
were
collected
in
a
small
test
tube.
The
treated
sample
was
the
sparged
with
helium
at
a
rate
of
10
ml/
min
for
10
min
to
remove
CO2.
After
helium
sparging,
the
sample
was
ready
for
analysis.
A
five
point
initial
calibration
curve
was
used
for
quantitation.
Instrument
calibration
was
verified
by
the
analysis
of
a
check
standard
after
10
samples.
Analytical
standards
were
prepared
from
neat
materials
at
the
concentration
ranges
are
listed
in
Table
1.
A
100
µ
g/
l
concentration
was
used
for
MS/
MSD
analysis.
All
standards
and
samples
were
pretreated
with
the
modified
cartridge.
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
69
Figure
1.
Standard
chromatogram
of
inorganic
anions
and
organic
acids.
Since
the
cartridge
removes
chloride,
a
separate
analysis
of
an
untreated
sample
is
required
for
this
parameter.
Method
detection
limits
were
calculated
by
the
analysis
of
7
replicates
as
described
in
40CFR136.
Method
detection
limit
evaluations
were
performed
in
reagent
water,
surface
water,
and
groundwater.
Samples
from
hazardous
waste
sites
were
frozen
in
the
field
using
dry
ice
to
prevent
the
degradation
of
organic
acids.

Figure
2.
The
effect
of
high
carbonate/
bicarbonate
concentrations
on
early
eluting
inorganic
anions
and
organic
acids.

Figure
3.
Preparation
and
regeneration
sequence
for
modified
sample
pretreatment
cartridges.
Sequence
begins
with
a
Dionex
OnGuard­
AG
 
cartridge.

RESULTS
AND
DISCUSSION
The
effects
of
the
cartridge
pretreatment
on
a
groundwater
sample
spiked
with
organic
acids
is
shown
in
Figure
4.
Severe
peak
broadening
occurs
during
the
analysis
of
the
untreated
sample
due
to
the
interaction
of
the
natural
carbonate
buffering
system
with
the
NaOH
eluent.
The
Dionex
OnGuard­
H
 
cartridge
improves
the
resolution
however;
fluoride
and
acetate
are
not
adequately
separated.
The
modified
cartridge
provides
a
separation
of
acetate
and
fluoride
that
can
be
quantitated
by
tangent
skimming
algorithms.
The
mechanism
for
the
improved
performance
of
the
modified
cartridge
has
not
been
determined.
It
is
possible
that
the
addition
of
silver
to
the
resin
results
in
sites
that
can
decompose
humic
materials
that
are
common
in
groundwater
and
can
also
influence
the
resolution
of
early
eluting
compounds.

The
results
of
method
detection
limit
(
MDL)
studies
are
summarized
in
Table
2.
All
analyses
reported
received
cartridge
pretreatment.
With
the
exception
of
fluoride,
MDLs
for
reagent
water
were
comparable
to
results
obtained
in
surface
water
and
groundwater.
The
fluoride
result
was
influenced
by
background
concentrations
found
in
the
water
samples.
Based
on
the
MDLs,
reporting
limits
were
set
at
the
level
of
the
lowest
standard.
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
70
Figure
4.
The
effect
of
cartridge
pretreatment
on
the
chromatographic
resolution
of
fluoride,
acetate,
formate,
and
propionate
in
a
groundwater
sample
spiked
with
organic
acids.

Table
2.
Results
of
the
Method
Detection
Limit
Study
0.024
0.007
0.004
0.006
0.017
0.006
0.012
0.004
0.005
0.003
0.017
0.008
0.002
0.001
0.002
0.005
0.002
0.004
0.001
0.002
0.001
0.005
0.020
0.009
0.007
0.008
0.015
0.008
0.009
0.007
0.004
0.010
0.002
0.006
0.003
0.002
0.003
0.005
0.003
0.003
0.002
0.001
0.003
0.001
0.005
0.003
0.006
0.006
0.010
0.008
0.007
0.004
0.004
0.006
0.005
0.002
0.001
0.002
0.002
0.003
0.002
0.002
0.001
0.001
0.002
0.002
0.02
mg/
l
0.02
mg/
l
0.02
mg/
l
0.02
mg/
l
0.05
mg/
l
0.02
mg/
l
0.05
mg/
l
0.02
mg/
l
0.02
mg/
l
0.02
mg/
l
0.02
mg/
l
Fluoride
Acetate
Propionate
Formate
Pyruvate
Nltrate
Benzoate
Glutarate
Succinate
Fumarate
Phosphate
MDL
(
mg/
l)
S.
D.
MDL
(
mg/
l)
S.
D.
MDL
(
mg/
l)
S.
D.
Surface
Water
Groundwater
Reagent
Water
Spiked
Concentration
Parameter
S.
D.
=
Standard
Deviation
MDL
=
Method
Detection
Limit
calculated
from
7
replicates
as
described
in
40CFR136.

Samples
from
ten
hazardous
waste
sites
were
analyzed
for
inorganic
anions
and
organic
acids
using
the
ion
chromatography
method
with
cartridge
pretreatment.
These
sites
represented
a
variety
of
solvent
and
fuel
releases
and
a
municipal
solid
waste
landfill.
The
precision
and
accuracy
results
for
MS/
MSD
analyses
performed
on
groundwater
samples
from
these
sites
are
summarized
in
Table
3.
Duplicate
precision
varied
from
±
12
%
to
±
20
%
for
the
target
analyte
list.
Accuracy
varied
from
70%
to
100%
for
all
analyses
except
phosphate.
During
sample
analysis,
it
was
noted
that
high
levels
of
ferrous
iron
caused
the
cartridge
to
retain
phosphate.
Some
of
the
ferrous
iron
was
retained
by
the
cartridge,
resulting
in
active
sites
that
would
precipitate
phosphate.
Samples
with
visible
ferrous
iron
require
the
use
of
a
freshly
prepared
cartridges
for
improved
recovery
of
phosphate.
After
analysis,
the
cartridge
must
be
regenerated.

Sample
chromatograms
for
three
of
the
hazardous
waste
sites
are
included
in
Figures
5,
6,
and
7.
The
sample
represented
in
Figure
5
was
from
a
solid
waste
landfill
that
had
groundwater
plume
with
low
levels
of
chlorinated
solvents.
Formic
acid
was
detected
in
a
number
of
wells
at
the
site
using
this
method.
Gas
chromatographic
analysis
detected
dissolved
methane
in
the
same
wells.
The
oxidation
of
methane
to
formic
acid
is
the
result
of
a
microbial
process
that
produces
an
enzyme
that
degrades
chlorinated
solvents3,5.
These
results
were
used
along
with
a
groundwater
model
to
demonstrate
that
intrinsic
bioremediation
was
occurring
at
the
site
and
that
the
process
was
reducing
and
containing
the
chlorinated
solvents
in
the
groundwater.
The
sample
shown
in
Figure
6
was
from
a
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
71
release
of
1,1,1­
trichloroethane
used
for
parts
washing.
Acetate,
dichloroacetic
acid,
and
trichloroacetic
acid
were
detected
in
the
groundwater
plume.
These
compounds
are
known
microbial
metabolites
and
chemical
hydrolysis
products
of
1,1,1­
trichloroethane4.
These
analytical
results
and
groundwater
modeling
were
used
to
show
that
the
plume
was
contained
by
intrinsic
bioremediation
and
that
continued
operation
of
the
groundwater
treatment
system
at
the
site
was
not
necessary.
In
a
similar
manner,
the
presence
of
acetate
and
large
reductions
in
nitrate
and
sulfate
in
the
groundwater
plume
of
a
toluene
release
(
Figure
7)
were
used
to
show
that
indigenous
microbial
populations
were
actively
degrading
toluene
in
the
aquifer.

Table
3.
Precision
and
Accuracy
Data
for
Organic
Anions
and
Organic
Acids
88
±
12%
86
±
15%
89
±
11%
84
±
14%
88
±
10%
92
±
10%
85
±
12%
88
±
14%
89
±
15%
84
±
12%
83
±
15%
±
15
%
±
18
%
±
17
%
±
19
%
±
20
%
±
12
%
±
15
%
±
14
%
±
18
%
±
16
%
±
13
%
0.50
mg/
l
0.10
mg/
l
0.10
mg/
l
0.10
mg/
l
0.25
mg/
l
1.0
mg/
l
0.25
mg/
l
0.10
mg/
l
0.10
mg/
l
0.10
mg/
l
0.50
mg/
l
Fluoride
Acetate
Propionate
Formate
Pyruvate
Nitrate
Benzoate
Glutarate
Succinate
Fumarate
Phosphate
Accuracy
%
Recovery
Precision
%
RPD
Spiked
Concentration
Parameter
RPD
=
Relative
Percent
Difference
%
RPD
and
%
Recovery
are
based
on
the
average
of
10
MS/
MSD
pairs
SUMMARY
The
cartridge
pretreatment
step
was
an
essential
component
of
the
ion
chromatography
method
described
in
this
paper.
These
cartridges
can
be
regenerated
and
used
for
multiple
analyses,
which
reduces
the
cost
of
consumables.
The
method
yields
results
that
are
comparable
with
other
ion
chromatography
and
wet
chemical
methods
for
inorganic
anions.
It
also
provides
data
on
organic
acids
that
are
useful
indicators
of
intrinsic
bioremediation.
Laboratories
using
this
method
can
provide
a
"
value
added"
service
to
consultants
developing
intrinsic
bioremediation
demonstrations
at
hazardous
waste
sites.

Figure
5.
Sample
from
a
solid
waste
landfill.
The
presence
of
formate
was
used
to
indicate
methane
oxidizing
bacteria
which
can
degrade
chlorinated
solvents.
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
72
Figure
6.
Sample
from
a
release
of
1,1,1­
trichloroethane.
The
presence
of
acetate,
dichloroacetic
acid,
and
trichloroacetic
acid
were
used
to
demonstrate
intrinsic
biodegradation.

Figure
7.
Sample
from
a
toluene
release.
The
presence
of
benzoate
and
acetate
were
used
to
document
intrinsic
bioremediation.

ACKNOWLEDGMENTS
The
authors
would
like
to
thank
the
NASA
Space
Grant
Program
and
the
Water
Resources
Institute
for
providing
funding
for
the
development
of
this
method.
We
also
thank
Horizon
Environmental
and
Earth
Tech
for
providing
hazardous
waste
site
samples
for
analysis
by
this
method.

REFERENCES
1.
National
Research
Council.
1993.
In­
situ
bioremediation:
When
does
it
work?
Washington,
DC:
National
Academy
Press.
2.
Wiedemeier,
T.
H.,
J.
T.
Wilson,
D.
H.
Kampbell,
RN.
Miller,
and
J.
E.
Hansen.
1995.
Technical
protocol
for
implementing
intrinsic
remediation
with
long­
term
monitoring
for
natural
attenuation
of
fuel
contamination
dissolved
in
ground
water.
San
Antonio,
TX:
U.
S.
Air
Force
Center
for
Ennronmental
Excellence.
3.
Wiedemeier,
T.
H.,
M.
A.
Swanson,
D.
E.
Moutoux,
J.
T.
Wilson,
D.
H.
Kampbell,
J.
E.
Hansen,
and
P.
Haas.
1996.
Overview
of
the
Technical
Protocol
for
Natural
Attenuation
of
Chlorinated
Aliphatic
Hydrocarbons
in
Ground
Water
Under
Development
for
the
U.
S.
Air
Force
Center
for
Environmental
Excellence.
in
United
States
Environmental
Protection
Agency.
Symposium
on
Natural
Attenuation
of
Chlorinated
Organics
in
Ground
Water.
EPA/
540/
R­
96/
509.
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
73
4.
McCarty,
P.
L.
1996.
Biotic
and
Abiotic
Transformations
of
Chlorinated
Solvents
in
Ground
Water.
in
United
States
Environmental
Protection
Agency.
Symposium
on
Natural
Attenuation
of
Chlonnated
Organics
in
Ground
Water.
EPA/
540/
R­
96/
509.
5.
Wilson,
J.
1996.
Natural
Attenuation:
Site
Characterization
Attenuation
of
Petroleum
Hydrocarbons
and
Solvents
in
Ground
Water.
in
United
States
Environmental
Protection
Agency.
Bioremediation
of
Hazardous
Waste
Sites:
Practical
Approaches
to
Implementation.
EPA/
625/
K­
96/
001.
6.
Bishop,
D.
F.
1996.
Natural
Attenuation
of
Landfills.
in
United
States
Environmental
Protection
Agency.
Bioremediation
of
Hazardous
Waste
Sites:
Practical
Approaches
to
Implementation.
EPA/
625/
K­
96/
001.
7.
USEPA.
1994.
Test
Methods
for
Evaluating
Solid
Waste.
Physical/
Chemical
Methods.
SW­
846.
3rd.
Edition.
8.
Steinmann,
P.,
and
W.
Shotyk.
1995.
Ion
chromatography
of
organic­
rich
natural
waters
from
peatlands.
III.
Improvements
for
measuring
anions
and
cations.
J.
Chromatogr.
A.
706:
281­
286.
9.
Ammann,
A.,
and
T.
Ruttimann.
1995.
Simultaneous
determination
of
small
organic
and
inorganic
anions
in
environmental
water
samples
by
ion­
exchange
chromatography.
J.
Chromatogr.
A.
706:
259­
269.
10.
Chen,
J.
1996.
Determination
of
organic
acids
in
industrial
streams
by
ion
chromatography
after
solid­
phase
extraction.
J.
Chromatogr.
A.
739:
273­
280.
11.
Dionex.
1995.
Installation
Instructions
and
Troubleshooting
Guide
for
the
Ionpac
As11
Analytical
Column.
Dionex
Document
No.
034791.
12.
Henderson,
I.,
R.
Saari­
Nordhaus,
and
J.
Anderson,
Jr.
1991.
Sample
preparation
for
ion
chromatography
by
solid­
phase
extraction.
J.
Chromatogr.
546:
61­
71.

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
DETERMINATION
OF
CHLORINATED
HYDROCARBON
CONCENTRATIONS
IN
SOIL
USING
A
TOTAL
ORGANIC
HALOGEN
METHOD
T.
B.
Lynn,
Director
of
Research,
J.
C.
Kneece,
Intern,
B.
J.
Meyer,
Research
Assistant,
and
A
Lynn,
Director
of
Analytical
Services
Dexsil
Corporation,
Hamden,
Connecticut
ABSTRACT
Total
organic
halogen
screening
has
been
used
extensively
to
quantify
chlorinated
organic
compounds
in
soil
and
is
the
basis
for
a
new
EPA
method
­
SW­
846
Draft
Method
9078
''
Screening
Test
Method
for
Polychlorinated
Biphenyls
(
PCB)
in
Soil".
This
method
uses
an
organic
solvent
to
extract
the
chlorinated
organics
from
the
soil
and
a
Florisil
column
to
remove
any
inorganic
chloride
from
the
extract.
The
extracted
chlorinated
organics
are
then
reacted
with
metallic
sodium
and
the
resulting
chloride
ions
are
quantified
using
a
chloride
specific
electrode.
Using
a
commercially
available
field
test
kit
(
the
L­
2000
PCB/
Chloride
Analyzer
 
)
,
the
ability
of
this
technology
to
measure
concentrations
of
chlorinated
pesticides
and
chlorinated
solvents
in
soil
was
determined.
The
compounds
investigated
were:
DDT,
pentachlorophenol
(
PCP),
toxaphene,
chlordane,
trichloroethylene,
tetrachloroethylene.
The
L2000
response
was
found
to
be
linear
over
the
range
0­
100
ppm
for
all
analyses
and
the
method
detection
limits
for
these
analyses
ranged
from
a
low
of
2.7
ppm
for
Chlordane
to
a
high
of
4.8
ppm
for
Trichloroethylene.
The
average
extraction
efficiency
varied
from
39%
for
PCP
to
greater
than
90%
for
the
chlorinated
solvents.

INTRODUCTION
The
procedure
for
total
organic
chloride
analysis
was
originally
developed
for
use
on
PCB
contaminated
soils
and
the
L2000
has
been
used
extensively
since
1990
for
this
purpose.
There
is
a
fairly
large
body
of
data
amassed
demonstrating
the
effectiveness
of
the
L2000
at
quantifying
PCB
in
soil1­
7.
The
underlying
principals,
however,
are
equally
applicable
to
other
chlorinated
organic
compounds
such
as
chlorinated
solvents
and
chlorinated
pesticides,
most
of
which
are
regulated
in
some
way.

The
L2000
has,
in
fact,
been
used
to
measure
other
chlorinated
compounds
in
soil.
In
the
majority
of
these
cases
the
end
user
has
undertaken
to
validate
the
feasibility
of
the
technology
for
their
particular
use.
This
validation
information
is
usually
site
specific
and
not
available
to
the
general
public.
With
the
growing
interest
in
the
remediation
of
other
chlorinated
compounds
in
soil
and
the
increase
in
information
requests
for
L2000
chlorinated
organics
applications,
we
have
undertaken
a
validation
program
for
these
applications
of
the
L2000.
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
74
The
first
and
most
important
step
in
a
total
organic
halogen
analysis,
or
any
chemical
analysis,
is
the
extraction
of
the
chlorinated
compounds,
quantitatively,
from
the
soil
matrix.
Performing
this
step
in
the
field,
quickly
and
reproducibly,
on
the
broad
range
of
soil
matrices
typically
encountered
is
not
a
simple
task.
The
solvent
system
must
be
designed
to
handle
everything
from
wet
clay
to
bone
dry
organic
material.
Unlike
other
field
analytical
methods,
the
organic
chlorine
is
converted
to
inorganic
chloride
in
a
non­
aqueous
solvent.
(
The
chloride
ions
are
then
extracted
for
quantification
using
a
chloride
specific
electrode.)
The
solvent
can,
therefore,
be
easily
tailored
and
optimized
for
a
particular
application.

The
standard
L2000
procedure
uses
a
proprietary
organic
solvent
that
is
polar
enough
to
penetrate
a
wet
clay
matrix
to
solvate
the
PCB,
but
is
itself
not
soluble
in
water.
Water
is
added
to
the
system
to
help
partition
the
inorganic
chloride
into
the
water
layer
and
away
from
the
solvent
layer.
A
Florisil
column
is
used
to
remove
any
residual
water
and
inorganic
chloride
from
the
extract.

This
solvent/
clean­
up
procedure
has
been
shown
to
be
effective
at
extracting
PCB
from
most
types
of
soils1­
7.
In
some
types
of
heavy
clay
soils
with
high
water
content,
the
extraction
efficiency
may
be
lowered
and
some
of
the
more
polar
chlorinated
organic
compounds
are
removed
by
the
Florisil
column.
Dexsil
has
developed
an
improved
alternative
two­
step
extraction
procedure
that
has
been
shown
to
efficiently
extract
PCB
from
wet
clay
soils
and
can
also
be
used
on
polar
compounds
such
as
PCP.
8
This
system
uses
both
a
polar
and
a
non­
polar
organic
solvent
combination
and
an
aqueous/
organic
solvent
partition
step.
An
optional
Florisil
clean­
up
step
can
be
added
if
the
analyte
is
not
one
of
the
polar
chlorinated
organics
such
as
PCP
or
if
PCP
is
considered
an
interfering
compound.

In
this
study
all
of
the
non­
polar
compounds
were
analyzed
using
the
standard
solvent
system
and
the
alternative
system
was
used
to
analyze
the
PCP
contaminated
soils.
PCP
was
analyzed
in
this
study
using
the
alternative
solvent
system
to
demonstrate
the
flexibility
of
the
L2000
solvent
system.

Following
the
solvent
extraction
and
clean­
up
(
if
necessary),
the
extract
is
reacted
with
metallic
sodium
in
the
presence
of
a
catalyst.
This
removes
the
covalently
bonded
chlorine
from
the
organic
backbone
producing
chloride
ions.
The
chloride
is
then
extracted
into
an
aqueous
buffer
and
then
quantified
using
a
chloride
specific
electrode.
The
user
can
select
a
standard
conversion
factor
for
one
of
the
typical
PCB
Aroclors
to
quantify
the
chloride
ions
as
"
equivalent
Aroclor".
The
actual
chlorine
content
of
the
original
sample
can
be
also
be
displayed.
Using
the
chlorine
content
of
the
specific
analyte,
the
equivalent
concentration
of
the
specific
contaminant
can
be
determined.
Because
the
response
of
the
instrument
follows
the
standard
Nernst
equation
and
the
quantified
result
is
the
chlorine
content
of
the
sample,
any
chlorinated
organic
compound
can
be
quantified
knowing
only
the
percent
chlorine
in
the
compound.
If
the
contaminant
is
unknown
at
the
time
of
measurement,
the
results
can
later
be
converted
using
a
simple
linear
transform,
once
the
contaminant
has
been
identified.

This
study
is
the
first
in
a
series
documenting
the
performance
of
the
L2000
in
new
applications.
Starting
with
the
fundamental
information
required
to
determine
if
the
L2000
technology
is
suitable
for
a
particular
application
we
have
limited
the
scope
of
this
investigation
to
determining:
the
method
MDL,
the
range
of
linearity,
and
the
extraction
efficiency
for
a
few
of
the
most
commonly
encountered,
regulated,
chlorinated
compounds.
We
have
used
laboratory
spiked
soils
to
simplify
the
experimental
considerations.

EXPERIMENTAL
Preparation
of
Spiked
Soil
Samples
To
ensure
a
consistent
soil
matrix
throughout
the
spiking
experiments,
a
large
batch
of
composite
soil
was
prepared
prior
to
beginning.
To
simulate
soils
found
in
uncontrolled
waste
contaminated
environments,
the
soil
composite
was
prepared
by
mixing
two
types
of
clay
and
one
type
of
sand.
Each
of
the
three
soils
were
sifted
through
an
0.850
µ
m
sieve,
and
then
combined
in
a
1:
1:
2
ratio
to
form
the
composite.

The
method
of
spiking
depended
on
the
particular
analyte
characteristics.
For
the
non­
polar,
semi­
volatile
compounds,
DDT,
toxaphene,
and
chlordane,
a
1%
stock
solution
in
chlorine
free
mineral
oil
was
prepared.
PCP,
being
more
polar,
is
not
soluble
in
mineral
oil;
therefore,
methanol
was
used
to
make
up
the
1%
stock
solution.
The
soils
were
spiked
at
100
ppm
by
adding
5
grams
of
the
1%
stock
solutions
to
500
gram
aliquots
in
aluminum
pans.
The
spiked
soil
aliquots
were
then
slurried
with
hexane
(
or,
in
the
case
of
PCP,
methanol)
and
allowed
to
evaporate
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
75
overnight
in
a
hood
space.
The
soils
were
then
transferred
to
16
oz
glass
jars
and
tumbled
for
one
hour.
The
jars
were
then
stored
at
room
temperature
for
later
use.

For
each
experiment,
soils
were
prepared
with
the
desired
contaminant
concentration
by
mixing
together
the
correct
proportions
of
the
100
ppm
spiked
soil
and
clean
composite
soil.
The
mixture
was
then
tumbled
for
an
hour
prior
to
use.

Spiking
soils
with
volatile
solvents,
uniformly
and
reproducibly,
presented
a
formidable
challenge.
In
previous
work,
all
attempts
to
produce
a
quantity
of
soil,
uniformly
spiked,
without
loss
of
the
analyte
proved
to
be
ineffective.
Therefore,
for
the
volatile
solvents,
trichloroethylene
and
tetrachloroethylene,
each
10
gram
soil
sample
was
spiked,
using
a
microliter­
syringe,
just
prior
to
analysis.

Method
Detection
Limit
Determination
The
L2000
method
detection
limits
for
each
of
the
chlorinated
compounds
were
determined
from
replicate
analysis
using
the
method
prescribed
by
the
EPA9.
An
estimate
of
each
of
the
detection
limits
was
made
using
the
concentration
equivalent
of
three
times
the
standard
deviation
of
replicate
measurements
of
the
analyses
in
the
composite
soils.
Soil
was
then
spiked
at
the
estimated
detection
limit.
The
spiking
concentration
for
each
of
the
chlorinated
compounds
are
listed
in
Table
1
below:

Table
1.
MDL
Soil
Spiking
Concentrations
20.5
ppm
85.5
24
ppm
Tetrachloroethylene
13
ppm
81.0
16
ppm
Trichloroethylene
3.5
ppm
69.2
5
ppm
Chlordane
3.4
ppm
~
68
5
ppm
Toxaphene
20
ppm
66.6
30
ppm
PCP
2.5
ppm
50.0
5
ppm
DDT
Soil
Chlorine
Content
Percent
Chlorine
Spiked
Level
Analyte
Each
of
the
spiked
soils
were
analyzed
seven
times
using
the
standard
extraction
method,
or
the
alternative
solvent
method
in
the
case
of
the
PCP
contaminated
soils.
Seven
matrix
blanks
were
also
analyzed
using
each
method.
The
average
blank
measurements
were
subtracted
from
the
respective
sample
measurements.
The
MDL
was
then
computed
using
the
following
formula:

MDL
=
t(
n­
1*
1­
a
=
0.99)*
S
where:
t
=
the
students
t
value
S
=
the
standard
deviation
of
the
replicate
analyses
The
student's
t
value
for
6
degrees
of
freedom
at
a
99%
confidence
interval
used
was
3.143.
The
mean
recovery
for
each
analyte
was
calculated
by
dividing
the
measured
concentration
by
the
theoretical
concentration
of
analyte.

Response
Curve
Determination
In
addition
to
the
stock
soil
spiked
at
100
ppm,
standards
were
prepared
in
the
composite
soil
at
2,
5,
10,
20,
and
50
ppm
of
each
of
the
chlorinated
solvents
and
pesticides.
Standards
were
analyzed
on
the
L2000
using
the
standard
extraction
method,
except
for
PCP
which
was
analyzed
using
the
alternative
extraction
method.
A
reagent
blank
was
run
with
each
analyte.
These
data
were
then
compared
to
analysis
by
gas
chromatography.
The
extraction
for
the
DDT,
PCP,
chlordane
and
toxaphene
samples
to
be
measured
by
gas
chromatography
at
the
following
concentrations:
2,
5,
10,
20,
50,
and
100
ppm
was
performed
by
adding
three
10
mL
aliquots
of
1:
1
acetone­
hexane
solvent
to
5
gram
aliquots
of
each
of
the
spiked
semivolatile
soils
while
rinsing
each
soil
with
each
addition.
The
solvent
was
then
removed
from
the
soil
and
run
through
a
polypropylene
filter
into
a
25
mL
volumetric
flask.
The
volume
was
filled
to
the
mark
with
excess
1:
1
acetone­
hexane.
The
solvent
was
then
transferred
to
another
25
mL
glass
test
tube
and
capped
with
a
teflon
cap,
then
centrifuged
to
remove
remaining
soil
particles,
and
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
76
prepared
for
gas
chromatography
analysis.
The
extraction
method
used
for
the
soils
spiked
with
volatile
analyses,
trichloroethylene
and
tetrachloroethylene,
utilized
10
mL
methanol
mixed
with
5
grams
of
soil
at
each
of
the
concentrations
2,
5,
10,
20,
50
and
100
ppm.
The
methanol
was
then
removed
from
the
soil
and
the
samples
were
then
prepared
for
gas
chromatography.
The
results
were
then
analyzed
and
compared
to
the
results
obtained
from
the
L­
2000
analysis.

RESULTS
AND
DISCUSSION
Method
Detection
Limits
The
MDLs
calculated
from
the
replicate
analysis
of
spiked
soils
were
within
the
recommended
range
for
all
analyses.
(
See
Table
2
below).
The
MDLs
calculated
for
the
non­
polar
compounds
using
the
standard
analysis
method
ranged
from
a
low
of
2.7
ppm
for
Chlordane
to
a
high
of
4.8
ppm
for
Trichloroethylene
and
4.4
ppm
for
Tetrachloroethylene.
The
semi­
volatile
MDLs
being
all
lower
than
the
MDLs
for
the
volatile
compounds.
A
contributing
factor
to
the
higher
MDLs
for
the
two
volatile
compounds
was
the
difficulty
in
preparing
the
spiked
soils.
This
was
not
unexpected,
given
the
difficulty
of
working
with
volatile
compounds
in
the
field.

The
MDL
of
8.7
ppm
calculated
for
the
analysis
of
PCP
was
higher
than
expected.
This
may
have
been
due
to
low
extraction
efficiency
of
the
new
solvent
system
on
polar
compounds.
A
low
extraction
efficiency
indicates
that
the
combination
solvent
was
not
able
to
penetrate
the
soil
matrix
to
completely
solvate
the
more
polar
PCP.
In
this
type
of
a
situation
the
analyte
recovery
is
very
sensitive
to
the
exact
handling
of
each
sample
replicate.
Small
changes
in
the
shaking
of
the
extraction
tube
or
the
length
of
extraction
will
have
a
larger
effect
on
the
recovery
than
is
acceptable.

While
the
new
solvent
system
facilitated
the
analysis
of
polar
compounds,
this
sensitivity
to
extraction
conditions
is
not
a
desirable
characteristic.
It
produces
variable
results
in
the
field
and
it
indicates
that
the
extraction
efficiency
will
vary
excessively
with
soil
matrix.
A
second
generation
two­
step
alternative
solvent
system
has
been
developed8
and
will
be
the
subject
of
the
next
phase
of
this
project.

Table
2.
MDL
Results
4.4
ppm*
1.3
ppm
110%
23
ppm
Tetrachloroethylene
4.8
ppm*
1.54
ppm
102%
21
ppm
Trichloroethylene
2.7
ppm
0.85
ppm
57%
5
ppm
Chlordane
2.8
ppm
0.91
ppm
37%
5
ppm
Toxaphene
8.7
ppm*
2.8
ppm
56%
30
ppm
PCP
3.6
ppm
1.15
ppm
54%
5
ppm
DDT
Calculated
MDL
Replicate
Standard
Deviation
Mean
Recovery
Spiked
Level
Analyte
*
Determined
using
the
alternative
extraction
method.
*
Determined
using
a
direct
spiking
method.

Response
Linearity
For
each
of
the
analyses
investigated,
the
response
of
the
L2000
using
either
solvent
system
was
found
to
be
linear
over
the
range
of
concentrations
studied.
The
resulting
R2
was
greater
than
0.96
for
all
analyses.
(
See
Figures
1­
6).
This
indicates
that
the
extraction
efficiency
is
consistent
over
this
analyte
range.
The
results
from
the
L2000
can,
therefore,
be
corrected
using
the
known
recovery.
There
is
no
indication
from
this
data
that
the
range
of
linearity
is
limited
to
100
ppm.

Extraction
Efficiency
Data
on
the
extraction
efficiency
of
both
solvent
systems
were
obtained
from
the
MDL
determinations
at
a
single
point
and
from
the
response
curve
determination.
The
single
point
and
the
response
curve
determination
of
average
extraction
efficiency
correlated
well
over
the
range
0­
100
ppm.
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
77
Table
3.
Extraction
Efficiency
98%
110%
Tetrachloroethylene
89%
102%
Trichloroethylene
70%
57%
Chlordane
65%
37%
Toxaphene
39%
56%
PCP
52%
54%
DDT
Average
Recovery
(
from
slope)
MDL
Mean
Recovery
Analyte
Summary
In
this
study
it
has
been
shown
that
the
L2000
can
be
used
effectively
to
analyze
soil
for
chlorinated
volatiles
and
semi­
volatiles.
The
method
detection
limits
were
shown
to
be
in
the
low
ppm
range.
This
should
be
adequate
for
most
contaminated
sites.
The
response
has
been
shown
to
be
linear
over
the
range
of
concentrations
studied
and
a
good
correlation
with
lab
methods
demonstrated.
A
new
solvent
system
suitable
for
polar
organic
compounds
was
shown
to
be
promising.
The
effectiveness
of
a
second
generation
two­
step
solvent
system,
demonstrated
in
a
separate
study,
will
be
the
subject
of
phase
two
of
this
project.
Furthermore,
the
list
of
suitable
chlorinated
compounds
will
be
expanded
and
the
effectiveness
of
the
new
alternative
solvent
system
on
other
chlorinated
compounds
will
be
investigated.
The
analyte
concentration
range
will
also
be
extended
to
2000
ppm,
the
upper
limit
on
the
L2000'
s
quantification
range.

REFERENCES
1.
"
Case
Study
of
a
New
Field
Screening
Tool
for
Delineating
Soil
PCB
Contamination",
Mark
B.
Williams,
PE
and
John
S.
Flickinger­
Dames
&
Moore
(
Madison,
WI),
Joseph
E.
Shefchek,
CHMM­
Wisconsin
Power
&
Light
(
Madison,
WI),
E.
Jonathan
Jackson,
CHMM­
Haliburton
NUS
Environmental
Corp.
(
Aiken,
SC);
Proceedings:
1991
EPRI
PCB
Seminar.
2.
"
PCB
Determination,
Simple/
Low
Cost
or
Complex/
Expensive,
Which
Method
is
the
Most
Reliable
in
the
Field?",
S.
Finch,
pp.
45­
1
to
45­
4;
Proceedings:
1991
EPRI
PCB
Seminar
3.
"
Application
of
a
New
PCB
Field
Analysis
Technique
for
Site
Assessment",
Roger
D.
Griffin­
Griffin
Environmental;
Proceedings
of
Hazmacon
'
92
March­
1992.
4.
"
A
Comparison
of
Popular
Field
Screening
Methods
for
PCB
Contamination
in
Soil",
Alvia
Gaskill­
Environmental
Reference
Materials,
1993
EPRI
PCB
Seminar.
5.
"
Comparison
of
the
Response
of
PCB
Test
Methods
to
Different
PCB
Aroclors",
Stephen
Finch­
Dexsil
Corporation
Proceedings
of
"
The
Tenth
Annual
Waste
Testing
and
Quality
Assurance
Symposium
July
11­
15,
1994
Arlington,
VA.

Figure
1.
Response
Curve
for
DDT
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
78
6.
"
Effect
of
Transformer
Oil
&
Petroleum
Hydrocarbons
as
Interferences
in
Field
Screening
for
PCB
Contamination
of
Soil",
Alvia
Gaskill;
Proceedings
of
"
The
Tenth
Annual
Waste
Testing
and
Quality
Assurance
Symposium",
July
11­
15,
1994
Arlington,
VA.
7.
"
Accurate
On­
Site
Analysis
of
PCBs
in
Soil,
A
Low
Cost
Approach",
Deborah
Lavigne­
Dexsil
Corporation.
8.
"
Improved
Extraction
Efficency
of
Polychlorinated
Biphenyls
from
Contaminated
Soil
Using
a
total
Halogen
Screening
Method,
W.
S.
Schutt­
Young,
A.
C.
Lynn,
T.
B.
Lynn,
M.
J.
Krumenacher,
Manuscript
Submitted
9.
Definition
and
Procedure
for
the
Determination
of
the
Method
Detection
Limit,
Environmental
Protection
Agency
Publication,
40
CFR
Ch.
1
(
7­
1­
89
Edition),
Appendix
B,
part
136,
525­
527
Figure
2.
Response
Curve
for
PCP
Figure
3.
Response
Curve
for
Toxaphene
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
79
Figure
4.
Response
Curve
for
Chlordane
Figure
5.
Response
Curve
for
Trichloroethylene
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
80
Figure
6.
Response
Curve
for
Tetrachloroethylene
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
A
FIELD
USEABLE
METHOD
FOR
TOXICITY
SCREENING
OF
WASTE
STREAMS
GENERATED
DURING
DESTRUCTION
OF
CHEMICAL
WARFARE
AGENTS
Theresa
R.
Connell
and
Kevin
M.
Morrissey
SciTech
Services,
Inc.,
1311
Continental
Drive,
Suite
G,
Abingdon,
Maryland
21009
Tel:
(
410)
671­
7104/
Fax:
(
410)
676­
2304
E­
mail:
opsmgr@
scitechinc.
com
H.
DuPont
Durst
US
Army,
CBDCOM,
Edgewood
Research,
Development,
and
Engineering
Center,
Edgewood,
Maryland
21009
Tel:
(
410)
671­
5270/
Fax:
(
410)
671­
2081
E­
mail:
hddurst@
cbdcom.
apgea.
army.
mil
A
program
to
chemically
neutralize
chemical
warfare
agents
(
CWA)
was
carried
out,
and
involved
the
neutralization
of
bis(
2­
chloroethyl)
sulfide
(
HD)
with
monoethanolamine:
water
solutions.
This
program
supports
an
effort
to
develop
and
deploy
field
transportable
systems
which
can
destroy
CWA
recovered
from
small
burial
sites.
To
ensure
operator
safety
and
completeness
of
destruction,
a
fieldable
second
tier
screening
method
was
required
to
estimate
toxicity
levels
of
the
resulting
reaction
masses.
This
second
tier
screening
is
in
addition
to
conventional
GC/
MSD
methods
for
quantitation
of
residual
CWA.
In
this
case,
the
toxicity
characteristic
of
the
HD
was
its
vesiccant,
or
alkylating
capacity.

Alkylating
capacity
is
defined
as
those
compounds
which
react
with
4­(
4­
nitrobenzyl)
pyridine
(
4­
NBP),
and
yield
a
product
which
absorbs
in
the
range
of
536
to
546
nm.
Alkylating
capacity
is
expressed
as
µ
moles
of
alkylating
sites
per
liter
of
sample,
or
µ
M.
Quantitation
is
accomplished
using
a
UV/
Vis
spectrophotometer,
with
external
standard
calibration.
The
method
was
determined
to
have
a
linear
range
of
98.7
to
987
µ
M
of
alkylating
capacity,
with
the
response
beginning
to
exhibit
non­
linear
behavior
after
approximately
1000
µ
M
of
alkylating
capacity.

The
use
of
4­
NBP
for
the
detection
of
alkylating
agents
is
well
documented,
and
has
been
extensively
reviewed.
The
method
presented
here
has
been
optimized
for
a
particular
sample
matrix,
and
consists
of
reaction
of
4­
NBP
with
acidified
(
pH
4
to
5)
sample
to
form
a
dye
precursor.
The
solution
is
then
made
basic
with
K2CO3
to
generate
a
final
product
with
an
absorption
maxima
at
approximately
540
nm.
The
reaction
product
is
then
extracted
into
2M
isopropylamine
in
toluene,
and
alkylating
capacity
is
quantitatively
estimated
using
standard
methods
of
colorimetric
analysis.
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
81
The
analytical
parameters
were
optimized
with
respect
to
acid
type,
base
type,
extraction
solvent,
reagent
amounts,
and
heating
time.
Parameter
selection
was
based
on
that
value
which
gave
a
maximum
absorbance
and
a
minimum
relative
standard
deviation.
Examples
of
optimization
experiments
will
be
presented.
In
addition,
results
from
experiments
evaluating
spiking
procedure,
response
of
method
to
other
alkylating
agents,
and
stability
of
reaction
product
will
also
be
presented.

The
method
was
validated
by
spiking
and
analyzing
seven
replicates
at
each
of
four
spike
levels.
All
four
spike
levels
were
performed
in
simulated
matrix,
and
one
spike
level
was
also
performed
in
an
actual
sample.
There
was
not
a
significant
difference
in
percent
recoveries
when
the
spike
was
performed
in
simulated
matrix
or
an
actual
sample.
The
average
percent
recovery
at
537
µ
M
alkylating
capacity
was
102%
(
n=
14,
CV
=
11.7%).
The
method
limits
of
detection
and
quantitation
will
be
discussed.

The
method
was
used
to
evaluate
three
independent
reactor
runs,
and
gave
consistent
patterns
for
each
kinetic
series
within
a
reactor
run.
In
addition,
a
sample
was
monitored
for
approximately
90
hours
after
reactor
drain
to
follow
the
continued
breakdown
of
HD.
Parallel
analyses
using
the
standard
GC/
MSD
methods
were
also
conducted.
Results
from
both
sets
of
analyses
will
be
presented
and
compared.

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
DETERMINATION
OF
OIL
CONTENT
OF
CONTAMINATED
SOILS
AND
SLUDGES
Abdul
Majid,
and
Bryan
D.
Sparks
Institute
for
Chemical
Process
and
Environmental
Technology,
National
Research
Council
of
Canada,
Ottawa,
Ontario
K1A
0R9
ABSTRACT
This
paper
reports
potential
applications
of
liquid
phase
agglomeration
techniques
in
the
removal
of
oil
from
contaminated
soil
and
sediment
samples
for
subsequent
quantitative
measurements.
The
use
of
liquid
phase
agglomeration
techniques
greatly
improves
the
efficiency
of
solids­
solvent
separation.
As
a
result,
better
contaminant
recoveries
were
achieved
in
a
shorter
period
of
time
compared
with
the
conventional
Soxhlet
­
Dean
and
Stark
solvent
extraction
method.
Quantitative
determination
of
the
contaminant
was
carried
out
using
gas
chromatographic
and
spectrophotometric
methods.
Two
case
studies
are
discussed.
The
first
deals
with
the
quantitative
determination
of
diesel
fuel
from
a
contaminated
diesel
invert
drilling
mud
sample.
The
second
evaluates
the
extraction
and
subsequent
quantification
of
heavy
oil
from
a
contaminated
soil
sample
The
objective
of
this
study
was
to
develop
a
quick
and
efficient
procedure
for
the
extraction
and
subsequent
quantification
of
total
petroleum
hydrocarbons
in
contaminated
wastes.
The
proposed
method
can
be
adapted
to
the
extraction
and
subsequent
quantification
of
a
variety
of
hydrocarbon
pollutants
from
contaminated
soils
and
sediments.

INTRODUCTION
Both
the
design
and
planning
of
a
soil
reclamation
program
and
the
evaluation
of
disposal,
reprocessing
or
reclamation
options
for
sludges,
oily
wastes
and
tailings
depend
upon
a
reliable
and
accurate
means
of
evaluating
the
total
oil
content
of
polluted
soils.
A
precise
estimate
of
the
amount
of
hydrocarbons
is
essential
to
determine
the
extent
of
contamination
and
the
success
of
the
reclamation
program.
Treatment
technologies
and
site
remediation
progress
lean
heavily
on
analytical
techniques
that
are
accurate,
reproducible
and
of
real
time
value.
Current
analytical
methods
are
inadequate
because
of
the
following
concerns:

w
The
recovery
of
oil
from
contaminated
samples
is
influenced
by
the
sample
preparation
method
and
the
oil
extraction
methods.
w
The
solvent
extraction
methods,
used
to
remove
oil
from
contaminated
samples
for
subsequent
measurements,
have
limitations
with
respect
to
the
water
content
of
the
sample
and
the
particle
size
distribution
of
the
matrix.
Usually
these
methods
are
less
efficient
for
fine
textured
materials.
Also,
water
is
known
to
reduce
the
extraction
efficiency
of
organic
solvents
because
of
the
formation
of
oil­
in­
water
emulsions1­
2.
w
Loss
of
volatile
organics,
during
solvent
removal
for
sample
concentration,
results
in
significant
bias.
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
82
w
Spectroscopic
determination
of
contaminants
has
the
disadvantage
that
degradation
of
pollutants
in
the
soil
results
in
altered
spectral
characteristics.
Absorbance
at
certain
wavelengths
may
increase
with
time
even
though
the
total
amount
of
oil
in
the
soil
would
be
decreasing.
Also,
in
clay
soils
where
oil
components
may
be
absorbed,
spectral
properties
of
the
extracted
oil
may
also
change1.
w
A
wide
variety
of
laboratory
techniques
used
to
measure
TPH
provide
data
of
varying
quality.

In
our
previous
work
we
demonstrated
the
potential
applications
of
proton
NMR
in
the
quantitative
determination
of
bitumen
and
solvents
extracted
from
oil
sands
and
fine
tailings3­
4.
The
NMR
method
can
be
used
to
estimate
the
amount
of
any
residual
solvent
in
the
oil.
The
retention
of
low
boiling
components,
normally
lost
during
solvent
removal,
produces
a
more
representative
sample
of
the
oil
for
calibration
purposes.

Liquid
phase
agglomeration
techniques,
in
combination
with
solvent
extraction
have
been
successfully
used
for
the
removal
of
hydrocarbon
contaminants
from
fine
textured,
organic
contaminated
soils5.
Agglomeration
of
fines
with
the
coarse
particles
is
achieved
by
the
addition
of
water
to
a
vigorously
mixed
slurry
of
soil
in
the
selected
solvent.
Water
acts
as
a
solids
bridging
liquid
and
dense
soil
agglomerates
are
formed
under
appropriate
conditions.
The
bridging
liquid
remains
in
the
agglomerate
pores
where
interfacial
tension
provides
the
forces
holding
the
aggregates
together.
Judicious
selection
of
agglomerate
size,
by
controlling
water
content,
greatly
improves
the
efficiency
of
solids­
solvent
separation.

The
objective
of
this
study
was
to
explore
the
use
of
liquid
phase
agglomeration
techniques
in
the
development
of
a
more
efficient
solvent
extraction
procedure.
Two
hydrocarbon
contaminants,
one
associated
with
a
fine
textured
soil
and
the
other
with
a
coarser
matrix,
were
selected
for
subsequent
quantitative
measurements.

EXPERIMENTAL
Samples.
One
of
the
samples
used
for
this
study
was
a
used
Diesel
Invert
Drilling
Mud
(
DIDM)
from
Alberta.
This
sample
contained
over
70
w/
w%
of
solids
<
41
µ
m
diameter.
The
diesel
content
of
this
sample
was
estimated
by
extraction
with
toluene
using
the
Dean
and
Stark
Soxhlet
method6.
The
extract
was
characterized
and
subsequently
quantitated
by
GC.
The
GC
chromatogram
of
the
extract
was
a
good
match
with
the
chromatogram
for
a
sample
of
diesel
obtained
from
a
local
gas
station
in
the
summer
of
1995.
This
commercial
diesel
sample
was
used
for
preparing
calibration
standards.

The
second
sample
used
in
this
study
was
a
highly
saline
soil
sample
provided
courtesy
of
Newalta
Corporation,
Calgary,
Alberta.
It
was
a
mixture
of
tank
bottoms,
frac
sand
and
spill
material
containing
a
range
of
organic
and
inorganic
contaminants.
The
organic
contaminant
in
this
sample
was
a
high
boiling
heavy
oil.

Proton
NMR
measurements.
Proton
NMR
measurements
were
performed
on
a
Brucker
AM­
400
NMR
spectrometer
(
400
MHz);
500
µ
L
of
solution
in
a
5
mm
outer
diameter
tube
was
used
in
each
case.
A
repetition
time
of
2
seconds
was
selected.
Each
spectrum
was
the
Fourier
transform
of
1000
free
induction
decay
curves.
Once
adjusted
all
parameters
were
kept
constant
for
subsequent
measurements.

GC
determination
of
diesel.
The
extract
was
analyzed
employing
a
Varian
Model
3300
GC
equipped
with
an
FID
detector
and
a
temperature
gradient
program.
A
DB­
5,
megabore,
capillary
column,
30
meters
in
length
with
a
0.53
mm
internal
diameter
and
a
1.5
micron
film
thickness
(
J
&
W
Scientific)
was
used.
The
initial
operating
temperature
was
50
degrees
Celsius
with
an
initial
hold­
time
of
5
minutes.
The
temperature
was
then
increased
at
the
rate
of
10
degrees
per
minute
to
a
final
temperature
of
300
degrees
Celsius
where
it
was
held
for
5
minutes.
Calibration
curves
were
prepared
using
the
purchased
sample
of
#
2
Diesel
Fuel.

Soxhlet
Dean
and
Stark
Method.
The
extraction
of
heavy
oil
from
the
Newalta
sample
was
carried
out
for
20
hours
using
the
Soxhlet­
Dean
and
Stark
method6.
Both
toluene
and
methylene
chloride
were
used
separately
for
extraction.

Spectrophotometric
estimation
of
heavy
oil.
The
quantitation
of
the
oil
component
was
carried
out
using
a
spectrophotometric
method7
based
on
the
linear
relationship
between
the
absorbance
at
530
nm
and
the
concentration
of
oil
in
solution.
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
83
For
calibration
purpose
oil
samples
were
obtained
from
a
methylene
chloride
extract
of
the
soil.
Non­
filterable
solids
were
removed
from
the
oil
extract
by
centrifugation.
The
solvent
was
removed
at
40
°
C
in
a
Brinkmann
rotary
evaporator
under
reduced
pressure.
The
amount
of
residual
solvent
in
the
oil
was
quantitatively
measured
using
proton
NMR4.
A
correction
for
solvent
content
was
applied
to
the
amount
of
oil
used
in
the
preparation
of
standard
solutions.
For
spectrophotometric
measurements,
absorbances
at
530
nm
were
determined
for
toluene
solutions
of
the
oil
for
concentrations
ranging
from
0.01
­
0.4
w/
w%.
Plots
of
the
percent
oil
vs.
absorbance
produced
a
straight
line
passing
through
the
origin.

Liquid
Phase
Agglomeration
Procedure.
In
a
typical
test
a
polypropylene
Waring
Blendor
Jar
(
500mL),
equipped
with
Teflon
washers
and
a
plastic
cover,
was
accurately
weighed.
To
the
jar
were
added:
contaminated
solids
(
20g),
and
solvent
(
50mL).
The
contents
were
agitated
at
high
shear
for
1
minute.
The
solution
was
carefully
drained
into
another
500
mL
polypropylene
jar.
About
1w/
w%
of
an
additive
(
sodium
meta
phosphate,
calcium
hydroxide
or
scrubber
sludge),
additional
water
and
fresh
solvent
(
50
mL)
were
added
to
the
original
Blendor
jar
still
containing
the
extracted
solids
from
the
primary
extraction.
The
contents
were
agitated­
at
high
shear
for
1
minute
followed
by
5­
10
minutes
at
low
shear
until
the
slurry
became
clear
as
discrete
agglomerates
were
formed.
The
supernatant
solution
from
the
second
treatment
was
drained
into
the
polypropylene
jar
containing
solution
from
the
primary
extraction.
The
procedure
was
repeated
three
more
times
and
all
extracts
were
combined.
The
extracted
solids
were
surface
washed
and
the
washings
were
added
to
the
solution
in
the
polypropylene
bottle.
The
contents
of
the
bottle
were
centrifuged
and
the
clear
solution
was
transferred
to
a
500
mL
glass
measuring
flask.
The
residue
in
the
polypropylene
bottle
was
agitated
with
fresh
solvent
and
then
centrifuged.
The
solution
was
combined
with
the
solution
in
the
measuring
flask.
The
procedure
was
repeated
until
the
washings
became
colorless.
The
solution
in
the
measuring
flask
was
made
up
to
the
mark.
The
amount
of
oil
was
estimated
in
this
solution
using
the
spectrophotometric
method.

RESULTS
AND
DISCUSSION
The
Amount
of
Extractable
Oil
from
Contaminated
Solids
The
extraction
of
diesel
from
a
used
diesel
invert
mud
sample
and
a
heavy
oil
from
Newalta
contaminated
soil
sample
were
carried
out
using
both
the
conventional
Dean
&
Stark
Soxhlet
and
Solvent
Extraction
Soil
Agglomeration
(
SESA)
methods.
Subsequent
quantification
was
carried
out
using
GC/
FID
for
the
diesel
and
spectrophotometric
method
for
the
heavy
oil.
The
results
are
summarized
in
the
Table.
It
is
obvious
from
the
results
that
not
only
better
recoveries
of
the
contaminants
were
obtained
using
the
SESA
method
but
the
total
turn
around
time
was
also
much
reduced.

13.38
±
0.49
(
3)
4.99
±
0.22
(
3)
2
SESA**
12.9
±
0.2
(
5)
4.84
±
0.1
(
5)
21
Dean
&
Stark
Soxhlet
Method
NEWALTA
DIDM***
Amount
of
oil
extracted
(
g/
100g
of
wet
solids)*
Extraction
plus
analysis
turn
around
time
(
Hrs)
Method
*
Figures
in
parenthesis
represent
number
of
tests
carried
out.
**
Solvent
Extraction
Soil
Agglomeration
process.
***
Diesel
Invert
Drilling
Mud
sample
CONCLUSION
Liquid
phase
agglomeration
has
potential
for
use
as
a
quick
and
efficient
analytical
procedure
for
the
extraction
and
subsequent
quantification
of
a
variety
of
petroleum
contaminants
from
contaminated
wastes.

REFERENCES
1.
McGill,
W.
B.,
and
Rowell,
M.
J.,
"
Determination
of
Oil
content
of
oil­
contaminated
Soil'"
The
Science
of
the
Total
Environment,
14,
245­
253
(
1980).
2.
Doldson,
S.
G.,
Miller,
G.
C.,
and
Miller,
W.
W.,
"
Extraction
of
Gasoline
Constituents
from
Soil,",
Journal
Association
of
Official
Analytical
Chemists,
73,
No.
2,
306­
311
(
1990),
3.
Majid,
A.
and
Woods,
J.,
"
Rapid
determination
of
Bitumen,
Varsol
and
other
Solvents
using
Proton
NMR",
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
84
Preprints
ACS
Div.
Fuel
Chem.
28,
188­
195
(
1983).
4.
Majid,
A.
and
Sparks,
B.
D.,
"
Total
Analysis
of
Mineral
Wastes
Containing
Bitumen,
Solvent,
Water
and
Solids",
AOSTRA
J
Res.
1,
21­
29
(
1984).
5.
Meadus,
F.
W.,
Sparks,
B.
D.
and
Majid,
A.,
"
Solvent
Extraction
Using
a
Soil
Agglomeration
Approach",
Emerging
Technologies
in,
"
Hazardous
Waste
Management
Vl",
Editors:
D.
William
Tedder
and
Frederick
G.
Pohland,
American
Chemical
Society,
Chapter
11,
pages
161­
176
(
1996)
6.
Syncrude
analytical
methods
for
oil
sand
and
bitumen
processing,
J.
T.
Bulmer
and
J.
Star
Editors.
Alberta
oil
Sands
Information
Centre,
Edmonton,
AB,
pp.
46­
51
(
1979).
7.
Patel,
M.
S.,
"
Rapid
and
Convenient
Laboratory
Method
for
Extraction
and
Subsequent
Spectrophotometric
Determination
of
Bitumen
Content
of
Bituminious
Sands";
Anal.
Chem.
46,
794
(
1974).

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
IMPROVED
EXTRACTION
EFFICIENCY
OF
POLYCHLORINATED
BIPHENYLS
FROM
CONTAMINATED
SOIL
USING
A
TOTAL
HALOGEN
SCREENING
METHOD
W.
S.
Schutt­
Young,
Ph.
D.,
Application
Specialist;
A.
C.
Lynn,
Director
of
Laboratory
Services;
T.
B.
Lynn,
Ph.
D.,
Director
of
Research;
B.
J.
Meyer,
Research
Assistant
Dexsil
Corporation,
One
Hamden
Park
Drive,
Hamden,
Connecticut
06517.
M.
J.
Krumenacher,
P.
G.,
CPG,
CGWP,
CHMM,
Hydrogeologist/
Senior
Project
Manager
GZA
GeoEnvironmental,
Inc.,
N4140
Duplainville
Road,
Pewaukee,
Wisconsin
53072.

ABSTRACT
Polychlorinated
biphenyls
(
PCBs)
are
strictly
regulated
on
the
state
and
federal
levels.
Responsible
parties
must
determine
the
concentration
and
extent
of
contamination
to
make
appropriate
decisions
regarding
remediation
of
PCB
contaminated
soils.
Gas
chromatography
(
GC)
analysis
has
been
traditionally
used
to
delineate
PCB
contamination
in
soil.
On­
site,
field
screening
techniques
have
been
developed
within
the
last
decade
to
reduce
the
number
of
samples
requiring
laboratory
confirmation.
All
field
screening
methods
require
the
use
of
organic
solvents
to
extract
the
contamination
from
soil.
Soils
sampled
for
analysis
frequently
contain
water
which,
with
some
solvents,
may
result
in
poor
extraction
efficiencies
and
can
subsequently
produce
false
negatives.
In
this
study,
GZA
GeoEnvironmental,
Inc.
(
GZA)
utilized
a
total
organic
halogen
(
TOX)
screening
kit
to
characterize
complex,
multi­
component
PCB
contaminated
soil
at
an
industrial
property
located
in
the
midwest.
Furthermore,
the
results
demonstrated
the
effectiveness
of
two
novel
solvent
systems
to
extract
PCBs
from
wet
clay
soil
for
reliable
quantification
with
a
portable
field
analyzer.
Preliminary
GC
analyses
indicated
that
the
soil
was
contaminated
with
Aroclor
1248
as
well
as
diesel
range
organics.
The
excavated
soil
consisted
of
wet,
red
clay.
Two
types
of
field
test
methods
are
available
for
PCB
screening,
both
of
which
may
exhibit
lower
extraction
efficiencies
for
wet
clay
soils.
One
method
relies
on
immuno­
assay
chemistry,
and
the
second
method
involves
chemical
dehalogenation
of
the
PCBs
followed
by
analysis
with
a
colorimetric
reaction
or
chloride­
specific
electrode.
Immuno­
assay
PCB
kits
suffer
severe
negative
interferences
in
the
presence
of
hydrocarbon
co­
contaminants
and
are
not
suitable
for
this
site.
Dexsil
Corporation's
L2000
®
PCB
Analyzer
(
U.
S.
EPA
SW­
846
Draft
Method
9078,
"
Screening
Test
Method
for
Polychlorinated
Biphenyls
in
Soil")
is
not
adversely
affected
by
the
presence
of
the
aforementioned
co­
contaminating
fuels
and
oils
and
was
therefore
selected
to
quantify
the
PCBs.
Aroclor
1248
concentrations
were
determined
in
over
71
wet
samples
by
extracting
the
contaminated
wet
soil
samples
with
a
new,
two­
step,
ozone­
friendly
solvent
system
followed
by
analysis
using
the
L2000.
An
additional,
modified
aqueous­
organic
solvent
extraction
method
with
a
Florisil
cleanup
column
was
also
used
to
determine
the
PCB
content
of
42
of
the
same
wet
samples.
The
second
method
was
designed
to
quantify
PCBs
in
the
presence
of
some
non­
PCB
halogenated
solvents.
A
portion
of
each
wet
soil
sample
was
then
dried
and
reanalyzed.
Ten
percent
of
the
nondetected
samples
along
with
all
samples
testing
positive
(>
2
ppm)
for
PCBs
were
analyzed
using
U.
S.
EPA
Method
8080
(
gas
chromatography
with
Soxhlet
extraction)
to
establish
the
extraction
efficiencies
of
each
new
two­
step
solvent
system.
The
L2000
results
of
wet
soil
samples
were
adjusted
for
water
content
and
compared
to
the
values
generated
by
Method
8080
(
GC).
Both
solvent
systems
demonstrated
the
ability
to
efficiently
extract
(>
73%)
PCBs
from
wet
and
dry
lacustrine
red
clay
soil.
Data
generated
by
L2000
analysis
of
PCBs
extracted
into
both
new
solvent
systems
exhibit
excellent
correlation
to
the
data
produced
by
the
more
sophisticated
laboratory
(
GC)
technique.
This
information
lends
confidence
in
PCB
field
screening
data
for
field
engineers.
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
85
INTRODUCTION
Under
the
authority
of
the
Toxic
Substances
Control
Act
(
TSCA)
of
1976,
the
U.
S.
Environmental
Protection
Agency
(
EPA)
regulates
the
use,
storage,
and
disposal
of
polychlorinated
biphenyls
(
PCBs)
1.
These
rules
were
imposed
because
of
the
health­
related
concerns
and
potential
environmental
impacts
associated
with
PCBs.
In
July
1979
the
U.
S.
EPA
banned
the
manufacture
of
PCBs
and
most
uses
in
which
the
PCBs
were
not
contained
within
a
closed
system
(
e.
g.,
transformers
and
capacitors).
Latitude
was
given
to
certain
industries
whose
PCB
usage,
such
as
electrical
applications,
were
not
considered
to
pose
unreasonable
risk
to
the
environment
or
human
health.
In
July
1985
the
U.
S.
EPA
developed
regulations
for
phasing
out
the
use
of
PCBs
in
all
enclosed
systems.

The
U.
S.
EPA
promulgated
a
PCB
Spill
Cleanup
Policy
in
April
1987
which
is
codified
in
40
CFR
Part
761.
The
PCB
Spill
Cleanup
Policy
requires
notification
for
PCB
spills
into
sensitive
areas
and
for
all
spills
greater
than
10
pounds.
The
PCB
Spill
Policy
also
establishes
cleanup
concentrations
for
soil
and
solid
surfaces.
Clean
up
of
the
affected
media
is
required,
whether
it
is
soil,
groundwater,
surface
water,
infrastructure,
equipment,
or
inventory.

Delineation
of
the
extent
of
contamination
in
soil
is
a
critical
step
in
any
remediation
project.
Accurate
delineation
is
more
critical
in
a
PCB
remediation
project
due
to
the
limited
and
expensive
disposal
alternatives.
PCB
disposal
generally
costs
an
order
of
magnitude
more
than
disposal
of
non­
PCB
containing
soil.
After
the
specific
type
Aroclor,
concentration,
and
extent
of
PCB
contamination
are
identified
in
the
soil,
an
appropriate
cleanup
and
disposal
plan
can
be
designed.
Confirmation
of
the
specific
PCB
Aroclor
and
concentration
are
accomplished
using
analytical
testing
procedures
performed
in
accordance
with
Test
Methods
for
the
Evaluation
of
Solid
Waste,
Physical/
Chemical
Methods,
U.
S.
EPA
Publication
SW­
846.
Traditionally,
U.
S.
EPA
SW­
846
Method
8080,
gas
chromatography
(
GC)
2
for
the
detection
of
organochlorine
pesticides
and
PCBs
is
used
in
conjunction
with
U.
S.
EPA
SW­
846
Method
3540,
soxhlet
extraction3
to
determine
the
PCB
concentrations
in
soil.

While
laboratory
analyses
are
required
to
definitively
confirm
the
type
and
concentration
of
PCBs,
the
methods
are
both
time
consuming
(
requiring
2
to
3
days
for
off­
site
analyses)
and
expensive
(
ca.
$
100
per
analyses).
Within
the
last
decade,
several
field
screening
techniques
have
been
developed
to
preliminarily
define
the
extent
of
PCB
contamination
in
the
field
and
reduce
the
number
of
samples
requiring
laboratory
analytical
testing.
Field
screening
techniques
provide
field
engineers
with
on­
site,
real­
time
information
necessary
to
make
field
decisions
regarding
the
investigation
or
remediation
process.

Application
of
an
appropriate
field
screening
method
can
improve
the
quality
of
a
site
investigation,
lessen
the
time
required
to
characterize
a
site,
decrease
the
project
cost,
and
increase
the
efficiency
of
a
remediation
process.
A
reliable
field
screening
technique
is
selective
to
the
analyte
of
concern,
insensitive
to
co­
constituents,
efficient,
easy
to
use,
and
relatively
inexpensive.
The
field
method
must
also
provide
good
correlation
with
analytical
laboratory
data.
Inaccurate
quantification
of
PCB
concentrations
by
an
ineffective
field
screening
technique
could
cause
responsible
parties
to
make
inappropriate
and
expensive
decisions
regarding
crucial
remediation
steps.

Accurate
quantification
of
a
soil
contaminate
begins
with
an
efficient
extraction
of
that
chemical
from
the
soil
media.
This
report
accounts
the
careful
selection
and
utilization
of
a
total
organic
halogen
(
TOX)
field
screening
kit
by
GZA
GeoEnvironmental,
Inc.
(
GZA)
to
aid
in
the
characterization
of
a
complex,
multi­
component
PCB
contaminated
site.
Furthermore,
the
results
will
demonstrate
the
effectiveness
of
two
novel
solvent
systems
to
extract
PCBs
from
clay
soil
for
reliable
quantification
with
a
portable
field
analyzer.

BACKGROUND
GZA
was
contracted
to
delineate
the
extent
of
PCB
containing
soil
at
an
industrial
property
in
the
midwest.
Due
to
the
expense
of
GC
analytical
testing,
GZA
chose
to
use
a
PCB
field
screening
method
in
addition
to
confirmational
GC
analysis.
A
properly
chosen
field
screening
method
would
provide
quick,
on­
site,
real­
time
data
at
a
reduced
cost.

Two
types
of
field
test
methods
are
currently
available
for
PCB
screening.
One
method
relies
on
immuno­
assay
chemistry.
The
immuno­
assay
chemistry
method
is
based
on
enzyme­
linked
immunosorbent
assays
(
ELISA)
in
which
a
competitive
reaction
between
PCBs
and
a
PCB
conjugate
is
used
to
determine
the
PCB
concentration
in
soil
samples.
The
second
method
involves
chemical
dehalogenation
of
the
PCBs
followed
by
analysis
with
a
colorimetric
reaction
or
chloride­
specific
electrode.
CoIorimetric
analysis
yields
an
estimate
of
the
PCB
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
86
concentration
greater
or
less
than
a
fixed
concentration.
The
chloride­
specific
electrode
quantifies
PCBs
at
concentrations
ranging
from
2
to
2,000
parts
per
million
(
ppm).

GZA
chose
to
determine
the
PCB
Aroclor
1248
concentrations
in
soil
using
the
chemical
dehalogenation
method
followed
by
analysis
with
Dexsil
Corporation's
L2000
®
PCB/
Chloride
Analyzer.
The
L2000
utilizes
a
chloride
specific
electrode
to
quantify
the
PCBs
in
accordance
with
its
associated
U.
S.
EPA
SW­
846
Draft
Method
9078
"
Screening
Test
Method
for
Polychlorinated
Biphenyls
in
Soil"
4.
GZA
selected
the
L2000
for
two
main
reasons.
First,
immuno­
assay
PCB
field
screening
kits
are
subject
to
severe
negative
interferences
in
the
presence
of
hydrocarbon
co­
contaminants
as
documented
by
Gaskill,
1993.5
The
study
completed
by
Gaskill
also
indicated
that
immuno­
assay
systems
may
be
susceptible
to
other
non­
specific
interferences
present
in
organic
solutions.
Dexsil's
L2000
is
not
adversely
affected
by
the
presence
of
hydrocarbon
co­
contaminating
fuels
and
oils
and
was
therefore
selected
to
quantify
the
PCBs.

The
second
reason
GZA
chose
to
use
the
L2000
is
because
Dexsil
volunteered
to
research
and
develop
a
unique
solvent
system
to
extract
PCBs
from
the
wet,
red
lacustrine
clay
characteristic
of
the
area.
Extracting
organic
contaminates
from
complex
clay
(
wet
and
dry)
matrices
poses
a
notoriously
difficult
task.
PCBs
are
especially
difficult
to
extract
from
wet
clays
as
previously
documented
by
Gauger,
et
aI,
1995.6
Based
in
part
on
research
and
testing
completed
on
soil
samples
collected
by
GZA
during
this
study,
Dexsil
has
developed
two
novel
solvent
systems
specifically
designed
to
achieve
high
PCB
extraction
efficiencies
from
clay
soil.

Prior
to
this
study,
the
L2000
was
demonstrated
to
be
an
effective
PCB
field
screening
tool
in
a
variety
of
soils
including
sand
and
several
different
clay
soil
types.
The
lacustrine
red
clay
soil
present
in
east
central
Wisconsin
posed
a
new
challenge
for
the
L2000
which
required
modification
to
the
extraction
procedure
and
the
need
for
this
study.

One
extraction
method
utilizes
a
novel
aqueous­
organic
solvent
system.
The
two­
step
procedure
results
in
a
high
PCB
extraction
efficiency
from
complex
soil
matrices
such
as
wet
clays.
For
cases
in
which
co­
contaminating,
non­
PCB,
halogenated
organic
compounds
are
present
or
suspected,
a
second
extraction
method
was
developed.
This
method
utilizes
a
similar
aqueous­
organic
solvent
followed
by
a
Florisil
cleanup
column.
The
FlorisiI
cleanup
column
removes
common
non­
PCB
organic
chloride
compounds
such
as
pentachlorophenol
from
the
solvent
extract.
PCBs
are
not
lost
on
the
Florisil
cleanup
column
during
the
filtration
process.
Inorganic
salts
associated
with
sampled
soil
cause
no
interference
with
PCB
quantification
by
the
L2000
after
extraction
with
either
solvent
system.
5
The
soil
samples
collected
by
GZA
were
extracted
as
received
using
the
novel
solvent
systems
and
analyzed
with
Dexsil's
L2000
to
determine
the
PCB
Aroclor
1248
concentration.
The
data
generated
using
the
new
solvent
systems
and
L2000
showed
an
excellent
correlation
when
compared
with
the
results
from
split
samples
analyzed
in
accordance
with
U.
S.
EPA
SW­
846
Method
8080
and
Method
3540.

SAMPLING
and
ANALYSIS
For
this
study,
GZA
utilized
a
Geoprobe
to
bore
approximately
45
soil
borings
at
the
industrial
property.
The
objective
of
the
soil
boring
program
was
to
evaluate
PCB
contaminated
areas
and
define
the
extent
of
PCBs
in
the
soil.
Soil
samples
were
collected
continuously
at
2
foot
intervals
through
the
end
of
boring
at
average
depths
of
12
feet
below
ground
surface.
A
Site
Plan
showing
the
layout
of
the
soil
borings
is
presented
as
Figure
No.
1.
Precautions
were
taken
during
sampling
to
prevent
cross
contamination
by
cleaning
the
Geoprobe
®
sampling
tools
with
a
Citrisolv
detergent
solution.
The
samples
were
kept
on
ice
and
delivered
to
Dexsil
within
48
hours
of
collection
under
chain­
of­
custody
documentation.
Dexsil
homogenized
then
analyzed
the
samples
using
the
L2000
in
accordance
with
U.
S.
EPA
SW­
846
Draft
Method
9078
and
GC
in
accordance
with
Method
8080
following
soxhlet
extraction
in
accordance
with
U.
S.
EPA
Method
3540.

PCB
Aroclor
1248
concentrations
were
determined
in
the
soil
samples
using
the
L2000
with
both
extraction
methods,
the
aqueous­
organic
solvent
system
and
the
aqueous­
organic
solvent
system
followed
by
filtering
through
a
Florisil
cleanup
column.
The
solvents
do
not
contain
chlorofluorocarbons
and
are
therefore
ozone­
friendly.
After
the
chlorinated
organic
compounds
were
extracted
from
the
soil
samples
into
the
solvent
using
the
aqueous­
organic
system,
or
modified
solvent
followed
by
the
Florisil
cleanup,
the
compounds
were
reacted
with
organo­
metallic
sodium
to
strip
away
the
chloride.
The
resulting
chloride
ions
were
then
quantified
using
the
chloride
specific
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
87
electrode
in
the
L2000.
The
L2000
converts
the
chloride
content
to
the
equivalent
amount
of
PCB
Aroclor
(
specified
by
the
user)
and
displays
the
value
on
a
digital
readout.
Samples
can
also
be
quantified
with
the
total
chloride
setting.
When
the
chloride
setting
is
used,
the
chloride
reading
is
converted
to
a
specific
PCB
Aroclor
concentration
using
the
chloride
content
in
the
identified
Aroclor.

A
portion
of
each
soil
sample
was
weighed
into
a
tray
and
dried
overnight.
The
calculated
water
content
was
used
to
adjust
the
L2000
results
for
the
wet
soil
to
the
PCB
concentration
in
an
equivalent
amount
of
dry
soil.
The
clumps
of
dried
soil
samples
were
broken
up
then
extracted
with
both
solvent
systems
and
re­
analyzed
using
the
L2000.
Based
on
the
results
of
the
soil
samples
analyzed
using
the
L2000,
10
percent
of
the
samples
where
PCBs
were
not
detected
and
all
samples
testing
positive
(>
2
ppm)
were
analyzed
using
U.
S.
EPA
SW­
846
Methods
8080/
3540
(
GC
with
Soxhlet
extraction).

RESULTS
and
DISCUSSION
As
summarized
in
Table
1,
the
average
water
content
of
71
clay
samples
collected
from
the
soil
borings
was
13%.
Also
summarized
in
Table
1
are
the
PCB
Aroclor
1248
concentrations
detected
in
71
wet
soil
samples
after
extraction
using
the
aqueous­
organic
solvent
extraction
method
and
L2000
analysis,
the
fraction
of
solids
in
the
soil,
calculated
concentrations
of
PCB
Aroclor
1248
in
an
equivalent
amount
of
dry
soil,
and
the
PCB
Aroclor
1248
concentration
determined
by
GC
after
Soxhlet
extraction.

Summarized
in
Table
2
are
the
PCB
Aroclor
1248
concentrations
of
16
dry
soil
samples
determined
using
GC
following
Soxhlet
extraction
and
the
L2000
following
extraction
with
the
aqueous­
organic
system.

Table
3
summarizes
the
PCB
Aroclor
1248
concentrations
detected
in
42
wet
and
dry
soil
samples
after
extraction
using
the
aqueous­
organic
solvent
extraction
method
with
Florisil
cleanup
column
and
L2000
analysis.
Included
on
the
Table
2
are
the
fraction
of
solids
in
the
soil,
calculated
concentration
of
PCB
Aroclor
1248
in
an
equivalent
amount
of
dry
soil,
and
the
PCB
Aroclor
1248
concentration
determined
by
GC.

The
PCB
concentration
detected
in
each
wet
and
dry
soil
sample
as
quantified
by
L2000
(
with
and
without
the
Florisil
cleanup
column)
was
plotted
against
data
obtained
using
U.
S.
EPA
SW­
846
Method
8080.
Standard
residual
analyses
were
done
on
the
data
to
identify
outliers.
Linear
regression
analysis
yielded
a
value
for
the
slope,
intercept,
and
regression
coefficient
(
R2).
The
slope
of
each
plot
is
an
indication
of
the
extraction
efficiency
of
the
solvent
system
in
wet
and
dry
clay;
and
a
non­
zero
intercept
would
indicate
any
systematic
bias.
The
correlation
between
the
L2000
and
GC
data
was
represented
by
the
R2
value.
Data
plots
for
soil
samples
collected
during
this
study
are
shown
in
Graphs
1
through
4.

Graph
1
is
a
plot
of
PCB
Aroclor
1248
concentrations
of
the
71
wet
soil
analyses
determined
using
the
L2000
following
extraction
with
the
aqueous­
organic
solvent
system
versus
the
PCB
Aroclor
1248
concentration
quantified
by
Soxhlet
extraction
and
GC.
Four
outliers
(
GP104­
S2,
GP106­
S4,
GP130­
S4,
and
GP140­
S5)
were
excluded
from
the
plot
based
on
results
from
standard
residual
analysis.
Despite
the
rather
high
water
content,
the
aqueous­
organic
solvent
system
(
without
the
Florisil
cleanup
column)
yielded
a
78%
PCB
extraction
efficiency.
The
correlation
between
the
GC
and
L2000
field
methods
was
0.93.

Regression
analysis
of
the
data
shown
in
Graph
1
using
a
95%
confidence
level
illustrates
the
intercept
of
the
plot
is
statistically
equivalent
to
zero.
The
slope,
however,
is
not
statistically
equal
to
one.
Since
solvent
extraction
systems
generally
can
not
achieve
the
same
efficiency
as
the
laboratory
soxhlet
extraction
method,
the
"
non­
one"
slope
was
not
surprising.
The
high
correlation
with
the
reference
method
indicates
that
the
extraction
efficiency
is
repeatable.
The
L2000
result
can,
therefore,
be
corrected
using
the
known
extraction
efficiency.
Furthermore,
if
an
environmental
professional
establishes
that
the
extraction
efficiency
of
PCBs
from
their
particular
soil
is
significantly
different
from
the
value
presented
here,
the
L2000
results
can
be
corrected
using
a
site
specific
correction.

Graph
2
is
a
plot
of
the
16
dry
soil
analyses
summarized
in
Table
2.
Plotted
in
Graph
2
are
the
Aroclor
1248
concentrations
determined
using
the
L2000
following
extraction
using
the
aqueous­
organic
solvent
system;
versus
the
Aroclor
1248
concentration
quantified
by
GC
following
soxhlet
extraction.
One
outlier
(
GP103­
S4)
was
excluded
from
the
plot
based
on
results
from
standard
residual
analysis.
The
aqueous­
organic
solvent
system
(
without
the
Florisil
cleanup
column)
yielded
a
77%
PCB
extraction
efficiency.
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
88
Regression
analysis
of
the
data
shown
in
Graph
2
results
in
an
intercept
that
is
statistically
equivalent
to
zero
at
a
95%
confidence
level,
indicating
insignificant
bias.
As
with
the
wet
soil,
and
for
the
same
reasons,
the
slope
is
not
statistically
equal
to
one.
The
regression
coefficient
of
0.97
demonstrates
an
excellent
correlation
between
the
data
generated
using
the
L2000
and
GC.

Graphs
3
and
4
are
plots
of
the
42
wet
and
dry
soil
analyses
summarized
in
Table
3.
Plotted
in
Graph
3,
for
the
wet
soil
samples,
and
Graph
4,
for
the
dry
soil
samples,
are
the
Aroclor
1248
concentrations
determined
using
the
L2000
following
extraction
using
the
aqueous­
organic
solvent
system
and
Florisil
cleanup
column
versus
the
Aroclor
1248
concentration
quantified
by
GC
following
soxhlet
extraction.
There
were
no
outliers
in
the
resulting
data
set.
The
aqueous­
organic
solvent
system
with
the
Florisil
cleanup
column
yielded
a
73%
PCB
extraction
efficiency.

Regression
analysis
of
the
data
shown
in
Graphs
3
and
4
using
a
95%
confidence
level
illustrates
that
the
intercept
of
the
plots
are
statistically
equivalent
to
zero.
As
with
Graphs
1
and
2,
and
for
the
same
reasons,
the
slopes
are
not
statistically
equal
to
one.
The
regression
value
of
0.88
for
the
wet
soil
samples
and
0.91
for
the
dry
soil
samples
demonstrates
a
good
correlation
between
the
data
generated
using
the
L2000
and
GC.

CONCLUSION
Although
the
PCBs
present
in
the
soil
at
the
die
casting
facility
were
associated
with
diesel
range
organics,
the
PID
used
to
identify
hydrocarbon
contaminated
soil
during
excavation
did
not
detect
the
co­
contaminating
PCBs.
Further
laboratory
analysis,
established
the
presence
of
Aroclor
1248.
Recognizing
that
the
PCBs,
in
addition
to
the
diesel
range
organics,
would
need
to
be
delineated
in
the
excavated
soil,
GZA
researched
the
feasibility
of
utilizing
different
PCB
field
screening
techniques.
Few
field
screening
methods
were
found
to
have
a
reputation
of
being
capable
of
quantifying
PCBs
in
real­
time
in
complex,
co­
contaminated
matrices
such
as
the
red
clay
soil
found
at
the
die
casting
plant.

Prior
to
this
study,
the
L2000
was
demonstrated
to
be
an
effective
PCB
field
screening
tool
in
a
variety
of
soil
types
including
sand
and
several
different
clay
soils.
This
particular
study
illustrates
the
use
of
two
novel
solvent
systems
to
extract
PCBs
from
complex
clay
matrices.
Both
solvent
systems
demonstrated
highly
efficient
(>
73%)
extraction
efficiencies
of
PCBs
from
lacustrine
red
clay
soil
when
compared
with
the
soxhlet
extraction
method.
The
extraction
efficiencies
reported
using
the
new
solvent
systems
are
greater
than
the
extraction
percentages
reportedly
achievable
by
other
field
screening
methods.

The
L2000,
in
previous
studies,
has
also
been
shown
to
reliably
quantify
PCBs
in
the
presence
of
co­
contaminating
hydrocarbons.
With
the
new
modified
solvent
system
and
Florisil
column,
the
user
can
efficiently
extract
and
reliably
quantify
PCB
Aroclors
in
the
presence
of
some
non­
PCB
halogenated
solvents.

The
information
gained
by
this
study
indicates
that
the
L2000
can
be
used
as
a
valuable
PCB
field
screening
tool
in
a
variety
of
difficult
matrices
and
situations.
This
should
heighten
environmental
engineers
confidence
in
field
data.

REFERENCES
1.
Woodyard,
J.
P.
and
King,
J.
J.,
1992,
PCB
Management
Handbook,
Executive
Enterprises
Publications
Co.,
lnc.,
New
York,
New
York.
2.
Environmental
Protection
Agency,
Office
of
Solid
Waste.
EPA
Method
8080,
Organochlorine
Pesticides
and
Polychlorinated
Biphenyls
by
Gas
Chromatography,
September
1994.
Washington,
D.
C.
3.
Environmental
Protection
Agency,
Office
of
Solid
Waste.
EPA
Method
3540,
Soxhlet
Extraction,
September
1994.
Washington,
D.
C.
4.
Environmental
Protection
Agency,
Office
of
Solid
Waste.
EPA
Draft
Method
9078,
Screening
Test
Method
for
Polychlorinated
Biphenyls
in
Soil,
Scheduled
for
3rd
Update
of
SW­
846.
Washington,
D.
C.
5.
Gaskill
Jr.,
A.,
1993,
Effect
of
Transformer
Oil,
Petroleum
Hydrocarbons
and
Inorganic
Salts
as
Interferences
in
Field
Screening
for
PCB
Contamination
in
Soil
6.
Gauger,
G.
A.;
Smith,
G.
C.;
and
Sullivan,
J.
M.,
1995,
Evaluation
of
a
Portable
test
Kit
for
Testing
PCB
Contaminated
Soils
and
Oils
in
the
Field.
EPRI
Polychlorinated
Biphenyl
(
PCB)
Seminar
proceedings.
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
89
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
90
Graph
1.
L2000*
versus
GC
Method
8080
Quantification
of
Aroclor
1248
in
Wet
Soils
(*
These
samples
were
extracted
with
the
aqueous­
organic
solvent)

Graph
2.
L2000*
versus
GC
Method
8080
Quantification
of
Aroclor
1248
in
Dry
Soils
(*
These
samples
were
extracted
with
the
aqueous­
organic
solvent)
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
91
Graph
3.
L2000*
versus
GC
Method
8080
Quantification
of
Aroclor
1248
in
Wet
Soils
(*
These
samples
were
extracted
with
the
modified
solvent
and
filtered
through
a
florisil
column)

Graph
4.
L2000*
versus
GC
Method
8080
Quantification
of
Aroclor
1248
in
Dry
Soils
(*
These
samples
were
extracted
with
the
modified
solvent
and
filtered
through
a
florisil
column)
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
92
Table
1.
Quantification
of
Aroclor
1248
in
Wet
Soils
by
USEPA
Methods
8080/
3540
compared
to
Extraction
by
an
Aqueous­
Organic
System
and
L2000
Analysis
0.11
0.03
1113.00
0.67
26.10
2.10
3.11
891.88
4.44
5.55
0.88
0.87
0.87
0.87
0.87
1.84
2.71
771.75
3.85
4.81
GP127­
S4
GP128­
S6
GP129­
S4
GP129­
S6
GP130­
S2
0.07
0.40
11.00
396.00
159.00
2.86
2.71
6.37
271.99
165.31
0.86
0.87
0.88
0.88
0.85
2.45
2.36
5.60
238.00
140.35
GP121­
S2
GP123­
S2
GP124­
S2
GP124­
S4
GP125­
S2
128.00
40.00
11.00
180.00
100.00
37.79
167.77
11.55
93.15
14.57
0.88
0.88
0.89
0.88
0.87
33.08
147.61
10.33
81.81
12.69
GP118­
S4
GP118­
S5
GP119­
S2
GP119­
S4
GP120­
S5
3.00
451.00
344.00
307.00
63.00
2.34
419.71
261.75
158.63
71.21
0.86
0.88
0.89
0.87
0.88
2.01
369.25
232.75
138.16
62.74
GP117­
S2
GP117­
S3
GP117­
S4
GP117­
S5
GP118­
S3
4.10
22.00
1.30
0.17
9.20
9.66
27.39
16.31
22.97
2.35
0.93
0.86
0.96
0.88
0.85
9.01
23.54
15.66
20.13
2.01
GP113­
S1
GP113­
S2
GP114­
S1
GP114­
S2
GP115­
S2
518.00
292.00
163.00
288.00
0.31
271.75
177.76
157.48
168.42
1.32
0.88
0.88
0.86
0.85
0.86
238.88
156.54
136.15
142.36
1.14
GP107­
S2
GP107­
S3
GP107­
S4
GP108­
S2
GP110­
S2
466.00
1.20
334.00
734.00
429.00
449.71
2.21
172.87
106.23
299.10
0.87
0.87
0.86
0.88
0.87
389.38
1.93
148.75
93.98
259.00
GP105­
S4
GP105­
S6
GP106­
S3
GP106­
S4
GP106­
S5
3540.00
1096.00
613.00
0.41
114.00
1360.13
811.02
707.24
2.09
81.74
0.88
0.88
0.87
0.88
0.87
1190.88
712.25
615.13
1.84
70.79
GP104­
S2
GP104­
S3
GP104­
S4
GP104­
S6
GP105­
S3
0.19
1450.00
1320.00
682.00
252.00
1.21
1224.50
1102.79
501.35
159.31
0.87
0.87
0.88
0.87
0.87
1.05
1065.75
970.38
437.50
137.81
GP101­
S2
GP103­
S2
GP103­
S3
GP103­
S4
GP103­
S5
Soxhlet
GC
8080
1248
(
ppm)
as
1248
Dry
Wght
(
Calc)
L2000
(
ppm)
Fraction
Solid
as
1248
Wet
Wght
L2000
(
ppm)
Sample
ID
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
93
38.00
374.00
646.00
97.00
19.20
3.40
23.69
116.56
386.57
62.44
19.15
3.46
0.87
0.87
0.88
0.87
0.83
0.94
20.56
102.03
339.50
54.43
15.93
3.24
GP143­
S2
GP143­
S3
GP143­
S4
GP144­
S4
SQ110896­
2
SQ110896­
3
27.00
9.10
3.10
20.00
12.00
40.58
9.43
6.81
14.52
4.60
0.88
0.88
0.86
0.87
0.88
35.70
8.31
5.86
12.60
4.03
GP141­
S1
GP141­
S3
GP141­
S4
GP141­
S5
GP142­
S4
445.00
12.00
116.00
760.00
70.00
163.51
5.30
324.76
1096.68
60.03
0.87
0.87
0.86
0.87
0.88
142.45
4.64
280.88
954.63
52.68
GP139­
S6
GP140­
S3
GP140­
S4
GP140­
S5
GP140­
S6
2.30
45.00
119.00
0.55
42.00
5.86
62.37
236.29
1.41
53.52
0.85
0.87
0.89
0.87
0.87
4.99
54.25
210.88
1.23
46.73
GP137­
S2
GP138­
S3
GP138­
S5
GP139­
S2
GP139­
S5
542.00
2.40
0.06
0.72
1.30
843.94
3.07
0.81
3.81
3.33
0.87
0.85
0.87
0.87
0.92
735.88
2.63
0.70
3.33
3.06
GP130­
S4
GP131­
S1
GP132­
S4
GP134­
S2
GP136­
S2
Table
2.
L2000
Analysis
(
after
aqueous­
organic
extraction)
of
Aroclor
1248
in
Dry
Soil
compared
to
GC
8080
Analysis
22
0.17
9.2
3
344
40
16.7125
1.6625
7.7875
4.2
307.125
43.4
GP113­
S2
GP114­
S2
GP115­
S2
GP117­
S2
GP117­
S4
GP118­
S5
1.2
734
292
288
0.31
0.9625
678.125
253.75
139.65
0.175
GP105­
S6
GP106­
S4
GP107­
S3
GP108­
S2
GP110­
S2
0.19
682
1096
0.41
114
1.575
880.25
776.125
0.6125
32.55
GP101­
S2
GP103­
S4
GP104­
S3
GP104­
S6
GP105­
S3
Aroclor
1248
Soxhlet
w/
GC
8080
(
ppm)
Aroclor
1248
in
Dry
Soil
L2000
(
ppm)
Sample
ID
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
94
Table
3.
L2000
Analysis
(
after
modified
extraction/
florisil
filter)
of
Aroclor
1248
compared
to
GC
Method
8080
Analysis
in
Wet
and
Dry
Soils
760.0
70.0
38.0
374.0
646.0
97.0
436.0
26.6
138.6
269.5
847.0
128.5
419.8
87.0
56.3
125.1
433.6
88.8
0.87
0.87
0.86
0.88
0.86
0.87
290.5
60.2
44.0
100.1
296.6
70.3
GP140­
S5
GP140­
S6
GP143­
S2
GP143­
S3
GP143­
S4
GP144­
S4
45.0
42.0
445.0
116.0
101.7
0.8
0.5
47.4
463.0
54.9
94.1
0.4
0.3
14.2
335.8
53.0
0.84
0.94
0.86
0.87
0.88
0.86
71.8
0.4
0.3
11.2
268.6
36.2
GP138­
S3
GP139­
S1
GP139­
S4
GP139­
S5
GP139­
S6
GP140­
S4
26.1
542.0
1.3
2.3
1.7
371.5
1.4
0.4
12.5
3.4
1.9
501.6
0.3
0.3
2.5
2.9
0.88
0.87
0.87
0.85
0.93
0.86
1.5
396.7
0.3
0.3
2.1
2.3
GP130­
S2
GP130­
S4
GP131­
S4
GP133­
S4
GP136­
S2
GP137­
S2
11.0
159.0
1113.0
0.7
9.4
69.3
0.3
0.5
962.5
0.1
6.8
7.3
18.7
3.3
1008.3
2.5
0.85
0.87
0.87
0.87
0.87
0.84
5.3
5.8
14.8
2.6
797.5
1.9
GP124­
S2
GP125­
S2
GP127­
S2
GP128­
S2
GP129­
S4
GP129­
S6
344.0
307.0
128.0
40.0
180.0
100.0
273.4
263.7
75.4
9.1
148.2
9.2
387.7
111.6
64.9
18.5
141.8
27.0
0.87
0.87
0.89
0.88
0.87
0.71
306.6
88.3
52.5
13.0
112.2
17.4
GP117­
S4
GP117­
S5
GP118­
S4
GP118­
S5
GP119­
S4
GP120­
S5
518.0
292.0
163.0
288.0
22.0
451.0
637.2
316.7
75.6
48.1
3.4
370.6
440.8
255.9
67.3
89.9
4.0
440.3
0.88
0.88
0.86
0.83
0.85
0.86
352.6
204.8
52.6
67.8
3.1
344.2
GP107­
S2
GP107­
S3
GP107­
S4
GP108­
S2
GP113­
S2
GP117­
S3
1450.0
1320.0
252.0
3540.0
1096.0
429.0
1058.8
1029.9
157.9
>
2000
586.2
143.4
520.9
912.1
63.5
8601.8
681.9
34.4
0.86
0.87
0.87
0.86
0.87
0.87
407.2
721.4
50.2
5884.4
539.4
27.2
GP103­
S2
GP103­
S3
GP103­
S5
GP104­
S2
GP104­
S4
GP106­
S5
Aroclor
1248
Soxhlet
w/
GC
8080
(
ppm)
Aroclor
1248
in
Dry
Soil
L2000
(
ppm)
Aroclor
1248
Dry
Weight
(
Calc)
L2000
(
ppm)
Fraction
Solid
Aroclor
1248
Wet
Weight
L2000
(
ppm)
Sample
ID
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
95
SPME
PREPARATIVE
APPLICATIONS
IN
ANALYSIS
OF
ORGANICS
IN
RADIOACTIVE
WASTE
J.
Young
ABSTRACT
NOT
AVAILABLE
AT
TIME
OF
PUBLICATION
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
CONGENER
SPECIFIC
PCB
GC
ANALYSIS:
A
FUNDAMENTAL
APPROACH
Dennis
R.
Gere
Hewlett­
Packard,
Little
Falls
Division,
2850
Centerville
Rd.,
Wilmington
DE
19808
Phone:
(
302)
633­
8162
Fax:
(
302)
633­
8908
Internet:
Dennis_
Gere@
HP.
COM
The
US
EPA
846
method
which
describes
Congener
Specific
Quantitation
of
PCBs
by
GC
capillary
columns
is
a
significant
improvement
for
the
quantitation
of
technical
mixtures
and
environmental
samples.
However,
there
are
some
refinements
needed.
These
refinements
reflect
new
developments
in
GC
stationary
phase
and
a
more
intense
effort
by
leading
laboratories
in
the
United
States
and
in
the
European
Union
countries.

We
describe
separation
for
key
PCBs
separated
on
a
5%
phenyl
dimethylsiloxane
and
a
50%
phenyl
dimethylsiloxane
column.
Data
will
be
shown
to
illustrate
the
resolution
of
several
pairs
of
PCB
congeners,
focusing
upon
the
7
indicator
PCBs
call
out
by
the
BCR
protocol
of
the
European
Union
and
neighboring
congeners.

Table
1.
PCB
Retention
Indices
HP­
5
and
HP­
50+

2,929
3,333
CB
209
2,485
2,898
CB
180
2,341
2,763
CB
138
2,338
2,748
CB
163
2,280
2,652
CB
153
2,223
2,630
CB
149
2,233
2,617
CB
118
2,110
2,474
CB
101
1,960
2,284
CB
52
1,892
2,168
CB
31
1,892
2,207
CB
28
HP­
5
HP­
50+
Indicator
Congeners
Table
1
shows
data
given
for
standard
retention
indices
for
11
selected
probe
PCBs
that
allow
a
standard
rating
of
the
polarity
of
the
stationary
phases.
This
should
be
considered
in
order
to
compare
other
columns
and
other
protocols
to
standard
"
ground­
state"
of
conditions
for
the
separation
and
identification
of
PCBs
in
technical
mixtures
and
environmental
samples.

We
show
in
Table
2
the
retention
indices
on
the
two
different
columns
for
standard
probe
compounds
useful
for
a
generic
measure­
of­
goodness
rating
fundamentally
important
before
the
consideration
of
how
columns
separate
(
or
do
not
separate)
the
PCBs.
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
96
Table
2.
Column
Checkout
Test
Mix
Retention
Indices
1,782
1,801
1,476
1,460
Acenaphtylene
1,581
1,600
1,476
1,472
1­
Dodecanol
1,329
1,420
1,324
1,320
Methyl
Decanoate
1,338
1,333
1,257
1,241
1­
Decylamine
1,421
1,424
1,206
1,184
4­
Chlorophenol
Temp
Prog
RI
Isothermal
RI
Temp
Prog
RI
Isothermal
RI
HP­
50+
HP­
50+
HP­
5
HP­
5
Contrary
to
conventional
opinion,
the
5%
phenyl
phase
column
has
a
selectivity
for
closely
eluting
PCB
pairs
allowing
resolution
of
PCB
congener
pair
128
and
163.
It
was
previously
thought
that
more
polarity
was
needed
to
obtain
this
resolution.
This
may
be
a
result
of
a
difference
in
commercial
stationary
phases
of
the
same
generic
type,
or
the
result
of
renewed
optimization
or
some
combination
of
both
efforts.
We
will
discuss
the
advantages
and
disadvantages
of
each
of
these
columns
as
well
as
discussing
the
tradeoff
of
the
film
thickness,
column
length,
use
of
hydrogen
as
the
carrier
and
time
considerations.

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
FULL
EVALUATION
OF
A
MICROWAVE­
ASSISTED
PROCESS
(
MAP
 
)
METHOD
FOR
THE
EXTRACTION
OF
CONTAMINANTS
UNDER
CLOSED­
VESSELS
CONDITIONS
Barry
Lesnik1,
J.
R.
Jocelyn
Paré2,
Jacqueline
M.
R.
Bélanger2,
Rodney
D.
Turpin3,
Raj
Singhvi
3,
Chung
Chiu3,
and
Richard
Turle4
1U.
S.
Environmental
Protection
Agency,
Washington,
DC
00000,
USA
2Environment
Canada,
Microwave­
Assisted
Processes
Division,
Environmental
Technology
Centre,
Ottawa,
ON,
Canada
K1A
0H3
3U.
S.
Environmental
Protection
Agency,
Emergencies
Response
Branch,
Edison,
NJ
08837­
3679,
USA
4Environment
Canada,
Analysis
and
Methods
Division,
Environmental
Technology
Centre,
Ottawa,
OH,
Canada
K1A
0H3
The
Microwave­
Assisted
Process
(
MAP
 
)
1­
4
has
been
the
subject
of
enhanced
interest
from
the
environmental
sector
lately.
We
report
on
inter­
laboratory
"
round­
robin"
work
that
was
just
completed
with
the
view
to
obtain
promulgation
of
an
associated
method.
Other
data
being
presented
are
the
result
of
previous
method
development,
validation,
and
certification
work
aiming
at
the
extraction
and
subsequent
GC/
MS
determination
of
mixtures
of
representative
toxic
substances.
Target
analytes
under
study
during
the
course
of
this
one­
year
validation
process
include
PAHs,
PCBs,
base­
neutrals,
chlorinated
and
organo­
phosphorus
pesticides,
phenols
and
substituted
phenols,
and
phenoxy
acid
herbicides.
Spike
matrices
consisted
of
soils,
marine
sediments,
harbour
sediments,
a
creosote
contaminated
soil
certified
with
PAHs
and
PCBs,
sand
and
air
filter
media.
Recoveries
from
these
matrices
were
acceptable
(>
80%);
precision
was
generally
in
the
10%
(
RSD)
range.
The
potential
problem
of
degradation
of
thermally
labile
pesticides
was
addressed.
Spiked
samples
before
and
after
the
high
temperature/
high
pressure
extraction
process
did
not
result
in
additional
decomposition
products5.

1.
J.
R.
J.
Paré,
M.
Sigouin,
and
J.
Lapointe,
U.
S.
Patent
5,002,784,
March
1991
(
various
international
counterparts).
2.
J.
R.
J.
Paré,
U.
S.
Patent
5,338,557,
August
1994
(
various
international
counterparts);
U.
S.
Patent
5,458,897,
October
1995
(
various
international
counterparts).
3.
J.
R.
J.
Paré,
U.
S.
Patent
5,377,426,
January
1995
(
various
international
counterparts);
U.
S.
Patent
5,519,947,
May
1996
(
various
international
counterparts).
4.
J.
R.
J.
Paré,
J.
M.
R.
Bélanger,
and
S.
S.
Stafford,
Trends
Anal.
Chem.
13,
176
(
1994).
5.
K.
Li,
J.
M.
R.
Bélanger,
M.
Llompart,
R.
D.
Turpin,
R.
Singhvi
and
J.
R.
J.
Paré,
Spectrosc.
Int.
J.
13,
1
(
1996).
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
97
OVERVIEW
OF
RCRA
ORGANIC
METHODS
PROGRAM
B.
Lesnik
ABSTRACT
NOT
AVAILABLE
AT
TIME
OF
PUBLICATION
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
SUMMARY
OF
STABILITY
STUDIES
FOR
VOLATILE
ORGANICS
IN
ENVIRONMENTAL
SOIL
SAMPLES
O.
R.
West1,
C.
K.
Bayne,
S.
R.
Cline,
R.
L.
Siegrist
and
D.
W.
Bottrell2
1Oak
Ridge
National
Laboratory,
P.
O.
Box
2008,
Bldg.
6011,
MS­
6036;
Oak
Ridge,
Tennessee
37831­
6036
Tel:
423­
576­
0505;
FAX:
423­
576­
8543
2U.
S.
Department
of
Energy,
Germantown,
MD
20874­
1290
ABSTRACT
Characterizing
the
distribution
of
volatile
organic
compounds
(
VOCs)
in
soil
is
a
crucial
activity
in
remediating
sites
contaminated
by
chlorinated
solvents
and/
or
petroleum
hydrocarbons.
This
typically
involves
the
extraction,
speciation
and
quantification
of
VOCs
from
discrete
soil
samples
which
are
assumed
to
contain
VOC
levels
that
closely
reflect
in
situ
values.
However,
previous
studies
have
shown
that
analyte
concentrations
can
change
during
sample
collection,
pre­
analytical
holding,
and
sample
preparation
prior
to
analysis.
These
changes
usually
result
in
analyses
that
underestimate
the
actual
levels
of
contaminants
present
at
a
site.
Accurate
quantification
of
VOCs
in
soil
samples
can
only
be
achieved
if
protocols
for
sample
collection,
preservation
and
preparation
for
analyses
can
reduce
and
manage
uncertainty
of
the
total
process.
This
includes
the
potential
overwhelming
influence
of
heterogeneity
and
other
sources
of
uncertainty.

This
report
summarizes
the
work
at
Oak
Ridge
National
Laboratory
since
1993
on
the
stability
of
VOCs
in
soil
samples.
The
work
described
here
focuses
on
the
development
of
preservation
methods
that
can
improve
the
stability
of
VOCs
in
soil
samples
during
pre­
analytical
holding.
Alternative
sample
preservation
methods
were
developed
and
evaluated
for
adoption
as
a
standard
protocol
in
environmental
sampling.
The
latter
would
allow
not
only
improved
VOC
measurements
but
also
extended
holding
times
for
VOC
soil
samples
beyond
the
currently
prescribed
maximum
of
14
days.

The
final
goal
of
this
work
is
to
define
a
standard
preservation
procedure
to
maintain
the
stability
of
VOCs
in
environmental
soil
samples.
Work
at
ORNL
and
elsewhere
has
shown
acidified
water
as
a
promising
preservative
for
soil
samples
with
low
concentration
levels.
ORNL's
analyses
of
containerization
suggest
that
a
40­
ml
VOA
vial
with
a
thick
Teflon­
lined
septum
cap
minimizes
volatilization.
Findings
and
conclusions
from
the
work
in
the
extension
of
pre­
analytical
holding
times
for
soils,
as
well
as
waters,
can
contribute
to
making
quality
assurance
materials
that
may
provide
the
confidence
in
the
practicality
of
field
analytical
methods.

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
ESTIMATING
THE
TOTAL
CONCENTRATION
OF
VOLATILE
ORGANIC
COMPOUNDS
IN
SOIL
SAMPLES
Alan
D.
Hewitt,
Nicole
J.
E.
Lukash
Geological
Sciences
Division,
U.
S.
Army
Cold
Regions
Research
and
Engineering
Laboratory,
72
Lyme
Road,
Hanover,
N.
H.
03755­
1290
(
603)
646­
4388
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
98
ABSTRACT
This
manuscript
describes
an
on­
site
method
of
estimating
the
total
concentration
of
volatile
organic
compounds
(
VOCs)
in
soil,
relative
to
a
site­
specific
0.2­
mg/
kg
working
standard.
The
purpose
of
this
decision
tool
is
to
allow
on­
site
sampling
activities
to
incorporate
the
appropriate
soil
sample
collection
and
handling
protocols
necessary
for
high­
and
low­
level
gas
chromatography/
mass
spectrometry
analysis.
Combining
rapid
on­
site
analysis
with
sampling
procedures
that
limit
substrate
dissaggregation
and
exposure
improves
efforts
to
achieve
site­
representative
estimates
for
vadose
zone
contamination.

INTRODUCTION
Gas
chromatography/
mass
spectrometry
(
GC/
MS)
Methods
8260
and
8240
are
frequently
used
by
laboratories
to
identify
and
quantify
volatile
organic
compounds
(
VOCs)
in
soil
samples.
One
of
the
challenges
of
coupling
a
sample
collection
and
handling
protocol
with
GC/
MS
analysis
is
that
it
has
limited
range
of
detection
(
two
to
three
orders
of
magnitude).
Of
particular
concern
is
that
high
analyte
concentrations
may
degrade
the
performance
of
the
MS
detector.
To
cope
with
this
limitation,
samples
thought
to
be
contaminated
with
VOCs
at
levels
above
0.2
mg/
kg
are
prepared
by
extraction
with
methanol
(
MeOH)
 
this
is
known
as
the
high­
level
method.
In
contrast,
samples
thought
to
have
concentrations
below
0.2
mg
VOC/
kg
are
analyzed
directly
(
that
is,
soil
is
transferred
directly
to
the
analysis
vessel)
 
this
is
referred
to
as
the
low­
level
method.

Also
of
concern
is
that
VOCs
in
soil
samples
fail
to
maintain
their
concentration
integrity
if
they
are
not
collected
and
handled
with
limited
disruption
and
exposure.
To
address
this
issue,
two
in­
vial
sample
collection
and
analysis
methods
(
Methods
5035
and
5021)
have
been
proposed
for
the
third
update
of
the
Test
Methods
for
Evaluating
Solid
Waste,
SW­
8461.
In­
vial
methods
are
most
effective
at
maintaining
VOC
integrity
when
samples
are
transferred
directly
to
a
prepared
analysis
vessel
during
the
field
sampling
activity2.

The
successful
combination
of
in­
vial
sample
collection
and
handling
and
GC/
MS
analysis
requires
that
either
multiple
samples
be
taken
for
laboratory
analysis
or
a
rapid
field
screening
method
be
used.
In
the
former
case,
samples
would
have
to
be
collected
in
an
appropriate
fashion
for
both
high­
and
low­
level
methods
of
GC/
MS
analysis,
and
at
least
one
additional
sample
for
a
laboratory
screening
analysis
needs
to
be
obtained3.
In
the
latter
case,
the
use
of
an
on­
site
screening
method
would
establish
which
sample
handling
procedure
should
be
used
prior
to
the
collection
of
a
sample
for
GC/
MS
analysis.

In
this
study
we
investigated
the
use
of
a
total
VOC
analyzer
to
establish
which
procedure
should
be
used
for
sample
collection.
Information
concerning
the
theory
of
total
soil
vapor
analysis
can
be
found
elsewhere4,5.
The
on­
site
analysis
procedure
developed
here
uses
a
hand­
held
total
VOC
analyzer,
takes
less
than
5
seconds
for
an
analysis,
and
requires
only
10
or
20
g
of
contaminated
soil
for
many
chlorinated
and
gasoline­
range
organic
compounds.
Collection
and
preparation
of
a
soil
sample
for
on­
site
analysis
takes
less
than
1
minute.
Proper
preparation
and
analysis
of
working
standards
are
important
for
calibrating
the
results
relative
to
both
sample
matrix
and
the
existing
meteorological
conditions.
Furthermore,
depending
on
the
objectives
of
the
field
investigation,
this
procedure
could
be
used
to
estimate
a
range
of
concentrations
or
test
whether
concentrations
are
above
or
below
0.2
mg
VOC/
kg.

EXPERIMENTAL
Equipment
and
Materials
A
Microtip
HL­
2000
(
Photovac,
Inc.)
equipped
with
photo
ionization
detector
(
PID)
was
the
total
VOC
analyzer
used.
This
instrument
was
modified
by
replacing
the
inlet
tube
with
a
3­
cm­
long
Teflon
tube
(
3.17
mm
o.
d.).
Clear,
44­
mL
VOA
vials
were
selected
as
the
vessels
for
the
working
standards
and
for
the
analysis
of
on­
site
soil
vapor
samples.
These
vials
were
modified
by
punching
a
5­
mm
hole
in
the
center
of
the
Teflon­
lined
silicone
septum
to
allow
air
to
easily
pass
around
the
PID
inlet
tube
once
it
is
inserted
through
the
hole.
To
temporarily
cover
the
vials,
3­
x
3­
cm
squares
of
light­
gauge
aluminum
foil
were
pressed
over
the
mouth
of
the
glass
vial,
then
secured
in
place
with
the
septum
and
screw
cap
(
Figure
1).
The
collection
and
transfer
of
soil
to
the
modified
VOA
vials
was
accomplished
with
a
10­
mL
disposable
plastic
syringe,
with
the
tip,
rubber
plunger
cap
and
holding
post
removed3.
Reagent­
grade
trichloroethene
(
TCE),
fresh
unleaded
gasoline,
and
purge­
and­
trap
grade
MeOH
were
used
to
prepare
the
stock
standards.
A
10­
µ
L
syringe
(
Hamilton)
was
used
to
transfer
volumes
of
the
stock
standard
when
preparing
the
working
standards.
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
99
Soil
Vapor
Analysis
In
preparation
for
a
soil
vapor
analysis,
the
instrument
was
initially
calibrated
using
zero
grade
air
and
standard
gas
(
100
ppm
isopropylene)
cylinders.
To
perform
an
analysis,
10
to
20
g
of
soil,
obtained
in
5­
g
increments
(
5
g
soil
3
mL)
with
a
10
mL
syringe,
l
were
transferred
to
the
VOA
vial
after
the
cap,
septum
and
foil
liner
were
removed.
Special
care
was
taken
when
collecting
and
transferring
the
soil
subsample
to
minimize
disaggregation.
Filling
the
syringe
with
more
than
3
mL
of
soil
is
not
recommended
because
larger
amounts
are
often
difficult
to
remove
and
are
easily
disaggregated.
Once
the
appropriate
volume
of
soil
was
obtained
and
the
foil
liner,
septum
and
cap
tightly
secured,
the
VOA
vial
was
hand
shaken
for
5­
10
seconds
to
disperse
the
soil
grains.
After
the
foil
liner
was
visually
checked
for
adhering
clumps
of
soil
(
they
were
knocked
off
if
present),
the
sample
was
analyzed
by
forcing
the
inlet
tube
through
the
foil
liner.
The
highest
reading
displayed
by
the
digital
meter
within
a
couple
of
seconds
of
the
foil
liner
being
punctured
was
the
value
recorded.
The
total
amount
of
time
between
exposing
a
fresh
soil
surface
and
completing
this
analysis
was
less
than
1
minute.

Figure
1.
Modified
VOA
vials
for
rapid
total
VOC
soil
analysis.

Working
Standards
Separate
stock
standards
of
0.53
mg
TCE/
mL
and
1.1
mg
gasoline/
mL
were
prepared
by
transferring
approximately
0.010
and
0.040
mL
of
these
constituents,
respectively,
into
25
mL
of
MeOH.
Working
standards
were
prepared
by
transferring
0.004­
mL
aliquots
with
a
microliter
syringe
to
a
clean
surface
on
the
inside
of
the
VOA
vial.
For
the
site­
specific
working
standards,
the
VOA
vials
contained
10
or
20
g
of
uncontaminated
soil
so
as
to
achieve
the
desired
VOC
concentration
of
0.2
mg
VOC/
kg
concentration.
Once
capped,
the
vials
were
hand
shaken
and
allowed
to
sit
for
at
least
2
hours
prior
to
analysis.
The
working
standards
were
analyzed
using
the
same
procedure
as
described
for
soil
samples.

Field
Samples
Experiments
were
performed
to
determine
the
response
of
the
Microtip
HL­
2000
to
working
standards
with
and
without
soil
present.
In
addition
to
evaluating
the
working
standards,
a
field
trial
was
performed
on
a
site
where
TCE
contamination
has
been
present
for
25
years.
For
the
field
trial,
the
on­
site
rapid
total
soil
vapor
measurements
were
compared
with
collocated
grab
sample
concentrations.
Discrete
soil
samples
were
collected
and
handled
using
an
in­
vial
procedure
that
was
compatible
with
an
equilibrium
headspace
gas
chromatography
(
HS/
GC)
analysis
method6.
The
sampling
locations
for
this
study
were
flat
20­
x
20­
cm
surfaces
dug
with
a
spade
to
depths
of
10­
30
cm.
At
each
location
a
site­
specific
TCE
working
standard
was
analyzed
just
before
the
hole
was
dug.
Immediately
after
a
fresh
soil
surface
was
exposed,
two
10­
mL
syringes
were
used
to
collect
the
soil
for
the
total
VOC
vapor
analysis.
Once
the
soil
vapor
analysis
was
completed,
a
single
syringe
was
used
to
collect
a
collocated
sample
for
HS/
GC
analysis.
The
grab
samples
taken
for
HS
analysis
were
transferred
with
a
modified
syringe
directly
to
44­
mL
VOA
vials
containing
20
mL
of
Type
1
water.
All
three
collocated
soil
subsamples
were
obtained
within
5
cm
of
one
another,
and
on­
site
analysis
and
sample
collection
was
completed
within
2
minutes
of
exposing
a
fresh
surface.

RESULTS
Table
1
shows
the
response
of
the
Microtip
HL­
2000
to
working
standards
when
the
VOA
vials
were
empty
or
contained
10
or
20
g
of
soil.
These
working
standards
were
allowed
to
sit
for
2
to
4
hours,
or
1
to
2
days
prior
to
analysis.
The
results
for
the
working
standards,
even
with
soil
present,
were
often
more
than
ten
times
higher
than
the
laboratory
background
readings
of
0.5­
1.0
ppmv.
As
anticipated,
the
presence
of
the
soil
matrix
had
a
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
100
pronounced
effect
on
the
results
for
the
working
standards.
The
near­
surface
soil
used
for
these
site­
specific
working
standards
had
a
moisture
content
of
15
±
5%
and
an
organic
carbon
content
of
1
±
0.5%.
For
working
standards
held
for
more
than
several
hours,
the
results
were
lower
because
of
vaporization
losses
(
Table
1),
since
aluminum
foil
fails
to
form
a
hermetic
seal
with
rigid
surfaces2.

Table
1.
Microtip
readings
of
TCE
and
gasoline
standards*
in
ppmv
with
and
without
soil
present.
Laboratory
background
Microtip
readings
ranged
between
0.5
to
1.0
ppmv.

17.5
20.4
20.8
21.4
8.9
8.0
9.3
6.9
18.7
19.0
19.5
17.8
18.7
17.4
16.7
17.9
18.3
9.8*
10.0*
4.3**
6.9**
10.8
11.1
11.6
11.1
11.0
12.1
11.3
11.0
11.5
no
soil
20
g
soil**
no
soil
10
g
soil**
Gasoline
ppmv
TCE
ppmv
*
0.004
mL
of
stock
standards
(
0.53
mg
TCE/
mL
or
1.1
mg
gasoline/
mL).
**
Weight
of
soil
 
thus,
concentrations
were
approximately
0.2
mg
total
VOC/
kg.
*
Held
for
24
hr
prior
to
analysis.
**
Held
for
48
hr
prior
to
analysis.

Table
2.
Results
of
rapid
total
VOC
soil
vapor
and
collocated
grab
sample
analyses.
In
addition,
values
obtained
for
the
measurement
of
the
site­
specific
working
standards
and
background
are
included.

42
­­
­­
40
34
­­
­­
27
25
­­
­­
20
53
57
50
39+
13*
0.765
0.0064
0.142
1.62
6.75
0.0369
0.0518
1.28
1.65
0.0239
0.0422
1.02
2.25
3.50
11.9
32
1.8
7.9
64
230
1.8
1.9
34
42
0.5
0.2
20
120
200
590
Aqueous
HS/
GC
(
mg
TCE/
kg)
Soil
vapor
PID
(
ppmv)
ratio
*
Field
trial
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
101
0.6
0.3
0.0
0.0
0.0
8.2
8.0
8.3
6.0
7.4
Background
Working
Standard
Response
(
ppmv)
of
site­
specific
TCE
working
standards
and
background
*
Ratio
of
soil
vapor
to
grab
sample
concentration,
for
locations
where
TCE
was
>
0.2
mg/
kg.
*
Mean
and
standard
deviation.

Table
2
shows
the
results
obtained
during
the
field
trial,
including
the
values
obtained
for
the
site­
specific
TCE
working
standards.
The
relationship
between
the
total
VOC
soil
vapor
and
collocated
grab
sample
analyses
was
both
linear
and
significant,
with
a
correlation
(
r2)
of
0.965
(
Fig.
2).
Indeed,
a
fairly
constant
ratio
existed
between
these
two
analyses
for
the
locations
where
TCE
concentrations
were
>
0.2
mg/
kg
(
Table
2).
This
relationship
is
encouraging,
since
it
indicates
that
concentrations
over
a
range
of
at
least
0.2
to
10
mg
VOC/
kg
could
be
estimated
using
the
rapid
soil
vapor
measurement
technique,
provided
that
an
adequate
number
of
confirmation
samples
were
taken.
Over
the
concentration
range
shown
in
Figure
2,
the
response
of
the
PID
appears
linear;
however,
at
higher
concentrations,
it
is
anticipated
that
this
relationship
would
become
nonlinear.
It
is
probable
that
the
linear
range
could
be
extended
to
higher
concentrations
if
a
field
instrument
equipped
with
a
flame
ionization
detector
were
used.

Figure
2.
Relationship
between
rapid
total
VOC
soil
vapor
analysis
and
grab
samples.

DISCUSSION
Collection
of
soil
for
VOC
analysis
should
always
be
the
first
operation
performed
after
the
surface
to
be
sampled
has
been
exposed
to
the
atmosphere.
If
a
freshly
exposed
soil
surface
is
not
rapidly
sampled,
analyses
existing
in
a
vapor
phase
diffuse
away
from
an
unsaturated
porous
matrix,
thereby
disturbing
the
equilibrium
that
existed
among
the
vapor,
liquid
and
sorbed
phases.
Following
the
depletion
of
the
vapor
phase,
there
are
nearly
instantaneous
shifts
in
the
equilibria
between
the
sorbed
and
aqueous
phases7.
An
example
of
how
quickly
this
process
can
occur
was
observed
by
tracking
the
TCE
concentrations
in
a
soil
subsample
collected
from
the
middle
(
1.2
cm
below
the
surface)
of
a
3.6­
cm­
i.
d.
x
5.1­
cm­
long
split­
spoon
core
liner.
Grab
samples
taken
from
this
subsurface
location
were
shown
to
have
lost
more
than
90%
of
this
analyte
when
the
core
liner
was
left
uncovered
in
a
plastic
bag
for
40
minutes
prior
to
sampling2.
In
general,
as
the
surface
area
of
exposure
increases,
the
length
of
time
before
significant
losses
occur
decreases,
even
when
precautions
are
taken
to
limit
disruption
of
the
native
soil
structure.
Soil
texture
also
has
been
shown
to
be
a
factor.
For
instance,
sandy
soils
tend
to
lose
VOC
more
rapidly
than
cohesive
silts
and
clays3.

In
addition
to
exposure
concerns,
soil
samples
must
be
transferred
directly
to
vessels
with
hermetic
seals;
these
vessels
must
either
contain
a
solvent
or
permit
an
analysis
to
be
performed
without
them
being
opened.
Proposed
Methods
5035
and
5021
both
recommend
the
use
of
VOA
vials
with
Teflon­
faced
silicone
septa
for
in­
vial
sample
handling
and
analysis
when
VOC
concentrations
are
below
0.2
mg/
kg.
In
general,
methods
capable
of
establishing
lower
VOC
concentrations
are
practical
when
attempting
to
establish
the
full
spatial
extent
of
contamination.
However,
when
the
primary
objective
of
a
site
investigation
is
to
locate
and
remediate
source
regions
where
residual
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
102
product
often
exists,
methods
that
maintain
the
integrity
of
soil
samples
contaminated
with
high
VOC
concentrations
(>
0.2
mg/
kg)
are
of
paramount
concern.
For
concentrations
above
0.2
mg/
kg,
however,
neither
method
in
the
initial
draft
provided
guidance
on
how
to
handle
samples
for
analysis
without
incurring
large
volatilization
losses.
Losses
of
VOCs
from
soil
samples
with
high
levels
of
contamination
can
be
limited
to
the
same
extent
as
the
low
level
samples
by
transferring
them
directly
to
vessels
containing
MeOH8,
9.

To
avoid
placing
samples
with
VOC
levels
less
than
0.2
mg/
kg
in
MeOH,
this
study
has
demonstrated
a
simple
method
of
preparing
site­
specific
working
standards
and
handling
field
samples
so
that
a
rapid
on­
site
analysis
can
be
performed.
Tables
1
and
2
show
that
the
performance
of
working
standards
was
fairly
reproducible.
Trends
in
our
initial
findings
suggest
that
working
standards
should
be
prepared
within
8
hours
(
or
less)
of
their
use
and
should
be
stored
in
a
cool
location
and
out
of
direct
sunlight10.
The
shelf
life
of
a
stock
standard
is
on
the
order
of
months,
if
held
in
a
hermetically
sealed
vessel.

Although
this
study
only
considered
field
samples
contaminated
with
TCE,
other
compounds
are
expected
to
behave
similarly
because
the
response
of
a
PID
varies
by
less
than
a
factor
of
1.4
for
common
chlorinated
and
aromatic
hydrocarbons
(
i.
e.,
TCE,
tetrachloroethene,
benzene,
and
toluene).
It
is
possible
that
this
on­
site
rapid
analysis
method
would
also
work
for
fuels
heavier
than
gasoline,
although
the
lower
quantity
of
highly
volatile
constituents
would
likely
require
larger
quantities
(
50
g)
of
soil.
We
present
this
on­
site
systematic
and
rapid
approach
to
soil
vapor
analysis
so
l
that
informed
decisions
about
how
to
handle
and
prepare
samples
for
laboratory
VOC
analysis
can
be
made.

SUMMARY
Failure
to
maintain
site­
representative
VOC
concentrations
results
in
false
negative
levels,
and
leads
to
inadequate
remediation
and
possibly
even
premature
closure.
To
address
these
shortfalls,
we
recommend
estimating
the
total
VOC
concentration
at
a
sampling
location
prior
to
collecting
samples
for
laboratory
quantitation.
The
focus
of
this
method
is
to
inform
site
investigators
when
the
use
of
MeOH
is
justified
for
on­
site
sample
preparation.
Likewise,
it
also
indicates
when
a
low­
level
in­
vial
sample
handling
and
analysis
procedure
(
Method
5035
or
Method
5021)
is
appropriate.
Additional
information
concerning
this
on­
site
analysis
decision
tool
and
how
to
incorporated
it
into
field
sampling
plans
is
available
elsewhere10.

ACKNOWLEDGMENTS
Funding
for
this
work
was
provided
by
the
U.
S.
Army
Environmental
Center,
Martin
H.
Stutz,
Project
Monitor,
and
from
the
U.
S.
Army
Environmental
Quality
Technology
Research
Program
Work
Unit
AF25­
CT­
005.
The
authors
thank
Marianne
Walsh
and
Philip
Thorne
for
critical
review
of
the
text.

This
publication
reflects
the
views
of
the
author
and
does
not
suggest
or
reflect
policy,
practices,
programs,
or
doctrine
of
the
U.
S.
Army
or
of
the
Government
of
the
United
States.

REFERENCES
1.
U.
S.
Environmental
Protection
Agency
(
1986)
Test
Methods
for
Evaluating
Solid
Waste.
Vol.
1B.
SW­
846.
2.
Hewitt,
A.
D.,
N.
J.
E.
Lukash
(
1996)
Sampling
for
in­
vial
analysis
of
volatile
organic
compounds
in
soil.
Am.
Environ.
Lab.
Aug.
3.
Hewitt,
A.
D.,
T.
F.
Jenkins,
C.
L.
Grant
(
1995)
Collection,
handling
and
storage:
Keys
to
improved
data
quality
for
volatile
organic
compounds
in
soil.
Am.
Environ.
Lab.
Jan­
Feb.
4.
Fitzgerald,
J.
(
1989)
"
Onsite
analytical
screening
of
gasoline
contaminated
media
using
a
jar
headspace
procedure,"
in
Petroleum
Contaminated
Soil,
Chelsea,
Michigan:
Lewis
Publishers.
pp.
119­
136.
5.
Robbins,
G.
A.,
B.
G.
Deyo,
M.
R.
Temple,
J.
D.
Stuart,
M.
J.
Lacy
(
1990)
Soil­
gas
surveying
for
subsurface
gasoline
contamination
using
total
organic
vapor
detection
instruments.
Part
I.
Theory
and
laboratory
experimentation.
Ground
Water
Monitoring
Review,
3:
122­
131.
6.
Hewitt,
A.
D.,
P.
H.
Miyares,
D.
C.
Leggett,
T.
F.
Jenkins
(
1992)
Comparison
of
analytical
methods
for
determination
of
volatile
organic
compounds.
Envir.
Sci.
Technol.,
26:
1932­
1938.
7.
Conant,
B.
H.,
R.
W.
Gillham,
C.
A.
Mendoza
(
1996)
Vapor
transport
of
trichloroethylene
in
the
unsaturated
zone:
Field
and
numerical
modeling
investigations.
Water
Resources
Research,
32:
9­
22.
8.
Urban,
M.
J.,
J.
S.
Smith,
E.
K.
Schultz,
R.
K.
Dickinson
(
1989)
"
Volatile
organic
analysis
for
a
soil,
sediment
or
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
103
waste
sample."
in
Proceedings
of
the
5th
Annual
Waste
Testing
&
Quality
Assurance
Symp.,
U.
S.
Environmental
Protection
Agency,
Washington,
DC,
pp.
II­
87­
II­
101.
9.
Hewitt
A.
D.
(
1994)
Comparison
of
methods
for
sampling
vadose
zone
soils
for
the
determination
of
trichloroethylene.
J.
Assoc.
Off.
Anal.
Chem.,
77:
458­
463.
10.
Hewitt,
A.
D.,
N.
J.
E.
Lukash
(
in
press)
Estimating
the
total
concentration
of
volatile
organic
compounds
in
soil:
A
decision
marker
for
sample
handling,
U.
S.
Army
Cold
Regions
Research
and
Engineering
Laboratory,
Special
Report,
Hanover,
New
Hampshire.

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
AUTOMATED
SMALL
VOLUME
EXTRACTION
OF
SEMIVOLATILES
FOLLOWED
BY
LARGE
VOLUME
GC/
MS
INJECTION
Fred
Feyerherm
Hewlett­
Packard
Co.,
2000
W
Loop
South,
Houston,
TX.
77027
(
713)
439­
5416
Rick
McMillin,
Diane
Gregg,
Mike
Daggett
U.
S.
Environmental
Protection
Agency,
Region
6
Lab,
10625
Fallstone
Rd.,
Houston,
TX.
77099
(
713)
983­
2111
Historically,
semivolatile
extractions
have
been
time
consuming,
labor
intensive,
and
use
large
amounts
of
solvents.
The
EPA
has
attempted
to
reduce
some
of
these
problems
by
employing
solid
phase
extractions
in
some
cases
which
use
less
solvent
than
other
techniques.
Our
objective
is
to
take
this
improvement
further
by
using
an
automated
solid
phase
and
liquid/
liquid
extractor
in
conjunction
with
a
large
volume
injector
to
reduce
sample
and
solvent
size.

This
study
will
consist
of
extracting
water
samples
spiked
with
method
8270
analyses.
Various
sample
sizes
will
be
investigated
in
combination
with
various
injection
volumes.
Precision,
recovery,
and
MDL
data
will
be
presented.
The
final
goal
will
be
to
achieve
acceptable
8270
method
performance.

Extractions
will
be
performed
using
an
automated
robotic
extractor.
Analysis
will
be
by
full
scan
quadrupole
GC/
MS
using
a
large
volume
injector.
Procedure
will
follow
method
8270
except
for
sample
and
injection
volumes.

Current
work
shows
1ml
extractions
followed
by
100ul
injections
into
a
GC/
MS
system
are
possible
to
achieve
reasonable
detection
limits.

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
A
COMPARISON
OF
ASE
WITH
SOXHLET,
SFE
AND
SONICATION
FOR
THE
EXTRACTION
OF
EXPLOSIVES
FROM
CONTAMINATED
SOILS
Bruce
E.
Richter
and
John
L.
Ezell
Dionex
Corporation,
SLCTC,
1515
W.
2200
S,
Suite
A,
Salt
Lake
City,
UT
84119
801­
972­
9292,
801­
972­
9291
(
fax)
Frank
Hoefler
Dionex
GmbH,
Am
Woertzgarten
10,
D­
65510
Idstein,
Germany
The
ongoing
privatization
of
former
military
bases
and
former
munitions
manufacturing
facilities
has
increased
the
need
to
deliver
fast
results
for
the
determination
of
explosives
in
soils.
Generally,
the
chromatographic
methods
are
straight
forward
and
fast.
However,
there
are
several
ways
from
which
to
choose
to
extract
the
analyses
of
interest
from
the
soil
samples.
Accelerated
solvent
extraction
(
ASE)
is
a
relatively
new
extraction
technique.
It
has
been
applied
to
several
compound
classes
ard
been
shown
to
be
equivalent
to
standard
extraction
procedures
for
the
extraction
of
semivolatiles,
polychlorinated
biphenyls,
organochlorine
and
organophosphorus
pesticides,
phenoxy
herbicides,
hydrocarbon
contaminants,
and
polychlorinated
dioxins
and
furans
from
solid
and
semisolid
samples.
The
purpose
of
this
study
was
to
compare
ASE
to
standard
extraction
procedures
for
the
recovery
of
explosive
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
104
compounds
from
contaminated
soils.

In
this
study,
Soxhlet,
sonication,
supercritical
fluid
extraction
(
SFE)
and
ASE
were
used
for
the
extraction
of
explosive
compounds
from
soils.
Soxhlet
extraction,
while
generating
accurate
and
reproducible
results,
is
time­
consuming
(
12­
24
hours)
and
uses
large
amounts
of
organic
solvents
(
300­
500
mL).
Sonication
extraction
reduces
the
time
required,
but
the
accuracy
of
the
results
is
questionable.
SFE
offers
a
dramatic
savings
in
solvent
consumption,
but
hardware
and
method
development
considerations
can
eliminate
potential
time
savings.
ASE
was
developed
to
offer
extraction
laboratories
the
ruggedness
of
Soxhlet
extraction
in
a
more
rapid
and
efficient
manner.

Spiked
and
incurred
soil
samples
obtained
from
munitions
plants
in
Germany
were
extracted
using
Soxhlet,
sonication,
SFE
and
ASE.
The
extracts
were
analyzed
using
HPLC
and
GC.
The
data
from
this
study
demonstrate
that
ASE
achieves
results
equivalent
to
Soxhlet
and
superior
to
sonication
and
SFE,
in
terms
of
accuracy
and
precision,
for
the
extraction
of
explosive
compounds
from
contaminated
soils.
Results
generated
from
these
techniques
will
be
compared
and
discussed.

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
A
COMPARISON
OF
MICROWAVE
EXTRACTION
SOLVENT
SYSTEMS:
NON­
POLAR
VS.
NON­
POLAR/
POLAR,
GENERAL
DIFFERENCES
FROM
SOIL
SAMPLES
Peter
J.
Walter,
George
P.
Lusnak,
and
H.
M.
"
Skip''
Kingston
Duquesne
University,
Department
of
Chemistry
&
Biochemistry,
Pittsburgh,
PA
15282­
1503
Microwave
assisted
organic
extraction
of
environmentally
important
organic
moieties
was
introduced
several
years
ago
by
Ganzler
and
modified
by
Lopez­
Avila1,2.
Their
work
demonstrated
that
microwave
extraction
is
a
viable
alternative
to
conventional
Soxhlet
extractions.
The
preferred
solvent
of
choice
would
be
a
pure
hydrocarbon
solvent,
but
these
solvents
were
very
poor
microwave
absorbers.
As
a
result,
the
solvent
systems
were
altered
to
incorporate
at
least
10%
polar
solvent
mixture
to
absorb
microwave
radiation.
However,
the
polar
solvent
extracts
many
interfering
polar
compounds.
These
problems
with
microwave
assisted
organic
extraction
can
be
overcome
through
the
use
of
secondary
microwave
absorbing
materials
like
Weflon3.

Weflon
is
a
chemically
inert
Teflon
impregnated
with
silicon
carbide,
a
very
strong
microwave
absorbing
material.
Weflon
can
be
added
externally
to
microwave
vessels
through
the
use
of
a
sleeve
for
the
microwave
vessel
or
internally
to
the
microwave
vessel
as
a
disk
or
other
physical
designs.
During
microwave
heating,
the
microwaves
are
partially
or
totally
absorbed
by
the
Weflon,
the
Weflon
material
heats
up
and
transfers
the
heat
to
the
non­
microwave
absorbing
non­
polar
solvent.
Despite
the
lagtime
in
transferring
the
heat
from
the
Weflon
to
the
solvent,
the
solvent
is
rapidly,
reproducibly,
and
controllably
heated.
Pure
dry
hexane
is
only
heated
to
~
70
°
C
in
15
minutes
whereas
with
the
addition
of
Weflon
the
identical
experiment
reached
an
extraction
temperature
of
115
°
C
in
~
1
minute.
Similarly,
the
use
of
a
non­
polar/
polar
solvent
combination
of
90/
10
toluene/
methanol
heated
in
~
7
minutes
whereas
with
the
addition
of
Weflon
the
identical
experiment
reached
the
same
temperature
in
~
2
minutes.

The
compounds
extracted
from
non­
polar
and
non­
polar/
polar
extraction
solvent
systems
are
quite
different.
A
non­
polar/
polar
solvent
system
extracts
many
polar
compounds
in
addition
to
the
desired
environmentally
regulated
nonpolar
compounds.
Many
of
the
polar
compounds
co­
elute
during
GC­
MS
analysis
resulting
in
complex
or
unquantitatable
peaks.
Omitting
the
polar
solvent
eliminates
interfering
polar
compounds
from
the
extraction
enhancing
analysis
of
the
remaining
components.
Using
an
external
microwave
absorber
enables
the
extraction
to
be
redesigned,
allowing
the
extraction
to
be
based
solely
on
the
chemistry
of
the
compounds
of
interest
rather
then
the
microwave
absorption
capabilities
of
the
solvent.
This
approach
permits
a
chemically
optimized
and
efficient
approach
to
microwave
extraction.

1.
Ganzler,
K.;
and
Salgo,
A.,
"
Microwave­
Extraction
­
A
New
Method
Superseding
Traditional
Soxhlet
Extraction",
Z.
Lepensm.
Unters.
Forseh.
184,
274­
276
(
1987).
2.
Lopez­
Avila,
V.;
Young,
R.;
and
Beckert,
W.
F.,
"
Microwave­
assisted
extraction
of
organic
compounds
from
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
105
standard
reference
soils
and
sediments",
Analytical
Chemistry
1994,
66,
pp
1097­
1106.
3.
"
New
Product
Announcement;
Weflon";
Milestone
Corporation,
Sorisole,
Italy.

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
ACCIDENTAL
CHEMISTRY
Mark
L.
Bruce,
Raymond
M.
Risden,
Kathleen
L.
Richards
Quanterra
®
,
4101
Shuffel
Dr.
NW,
North
Canton,
OH
44720
(
brucem@
quanterra.
com)
Roseann
Ruyechan
Quanterra
®
,
450
William
Pitt
Way,
Pittsburgh,
PA
15238
Paul
C.
Winkler
Quanterra
®
,
4955
Yarrow
St.,
Arvada,
Colorado,
80002
ABSTRACT
There
are
several
procedures
which
have
been
incorporated
into
EPA
sample
preparation
methods
which
can
cause
unintended
chemical
reactions
with
target
analyses.
Knowing
the
potential
for
these
reactions
can
lead
to
improved
methods
which
avoid
the
"
accidental
chemistry"
or
aid
proper
data
interpretation.
Many
analytical
tests
rely
on
colorimetric
or
derivatization
steps
as
integral
parts
of
the
analysis
process
but
generally
organic
sample
preparation
and
analysis
attempts
quantitate
and
identify
analyses
without
changing
their
chemical
structure
until
the
actual
detector
is
reached.
On
some
occasions
interfering
compounds
are
produced
during
sample
preparation.
Three
instances
have
been
found
where
acetone,
used
in
sample
preparation,
reacted
with
either
target
analyses
or
other
compounds
present
and
degraded
the
analytical
results.
The
methods
and
analyses
affected
are:
3520­
8080/
81/
82
organochlorine
pesticides
and
PCBs,
8150/
51
dinoseb
and
3540/
50
­
8270
primary
amines.
Some
of
these
problems
have
only
come
to
light
over
time
as
analyte
lists
have
expanded
or
quality
requirements
have
increased.

ORGANOCHLORINE
PESTICIDES
AND
PCBS
Unknown
chromatographic
interferents
were
coeluting
with
some
organochlorine
pesticides
(
OCP)
and
PCBs
when
water
samples
and
blanks
were
prepared
using
continuous
liquid­
liquid
extractors
(
Method
3520).
These
interfering
peaks
were
not
present
when
the
extraction
was
performed
by
separatory
funnel
(
Method
3510).
Typical
chromatograms
are
shown
in
Figures
1
and
2.

Early
attempts
to
identify
the
source(
s)
of
the
problem
examined
glassware
cleaning,
condenser
temperature,
acetone
spike
solvent,
contamination
from
lab
air
and
reagent
water.
These
various
possibilities
were
tested
one
at
a
time.
The
most
likely
sources
were
examined
first.
The
other
possibilities
were
tested
over
the
months
that
followed.
Some
experiments
affected
contamination
levels
but
the
specific
process
which
produced
the
contaminants
was
not
understood
nor
did
the
analysts
know
how
to
"
make
the
problem
stay
away''.

After
several
one­
at­
a­
time
attempts
to
isolate
the
source(
s)
of
the
contaminants
a
set
of
factorial
design
experiments
was
used
to
systematically
investigate
the
8
most
likely
sources.
The
sources
or
variables
are
called
factors
in
statistical
jargon.
Each
factor
was
varied
between
two
levels
or
options.
The
factors
selected
were
surrogate
spike
solvent,
reagent
water
source,
exposure
to
laboratory
air,
glassware
cleaning,
extraction
method,
dichloromethane
source,
condenser
temperature
and
exposure
to
laboratory
light.
This
set
of
experiments
narrowed
down
the
list
of
probable
sources
(
factors).
The
four
most
likely
factors
were
selected
and
tested
most
extensively
in
another
round
of
factorial
design
experiments.
The
results
from
this
study
lead
to
probable
sources
and
a
plausible
mechanism.
Several
possible
ways
to
reduce
or
eliminate
the
problem
were
deduced.

The
first
factorial
design
experiment
set
examined
the
8
factors
listed
in
Table
1.
Each
factor
was
tested
at
two
"
levels".
A
16
run
subset
of
the
256
possible
combinations
was
selected
to
provide
the
best
information
about
the
effect
of
each
factor
and
possible
two
factor
interactions.
The
factorial
design,
calculations
and
data
interpretation
were
carried
out
with
the
aid
of
Design­
Ease
software
from
Stat­
Ease
of
Minneapolis,
Minnesota.
Table
1
shows
each
factor
and
the
two
levels
investigated.
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
106
Figure
1.
OCP/
PCB
chromatogram
of
continuous
liquid­
liquid
extract
(
acetone)

Figure
2.
OCP/
PCB
chromatogram
of
separatory
funnel
extract
(
acetone)

The
normal
probability
plot
of
the
standardized
effect
calculated
from
the
design
shows
a
group
of
factors
which
have
an
effect
larger
than
the
noise
(
or
random
variability)
calculated
from
the
other
factors.
These
factors
and
factor
interactions
were
extraction
method,
DCM
source,
spike
solvent
&
reagent
water
interaction,
spike
solvent,
spike
solvent
&
extraction
method
interaction,
spike
solvent
&
DCM
source
interaction
and
reagent
water.
Because
this
was
a
1/
16th
fraction
of
all
possible
experimental
combinations,
all
factor
interaction
effects
were
aliased
with
other
effects.
For
example,
the
calculated
effect
assigned
to
spike
solvent
&
reagent
water
also
included
the
effects
from
lab
air
&
condenser
temperature,
cleaning
&
light
exposure
and
extraction
method
&
DCM
source.
Also,
higher
order
interactions
would
be
included.
Thus,
care
must
be
exercised
when
interpreting
the
results
to
avoid
misassigning
a
particular
effect
to
an
individual
factor
or
interaction.
Previous
chemistry
knowledge
allows
some
potential
interactions
to
be
labeled
as
very
unlikely.
For
example,
potential
interactions
involving
reagent
water
source
and
CLLE
condenser
temperature
were
unlikely
because
they
do
not
come
in
contact
with
each
other.

This
factorial
design
did
not
completely
define
all
problem
factors
but
eliminated
some
so
focused
effort
could
be
directed
to
the
most
likely
factors.
The
effects
of
spike
solvent,
DCM
source,
extraction
method,
reagent
water
source
and
light
exposure
appeared
strong
enough
to
warrant
further
investigation.

A
four
factor
design
(
16
extractions)
with
spike
solvent
(
methanol
/
acetone),
DCM
source
[
both
sources
were
residue
analysis
grade
this
time]
(
X/
Y'),
reagent
water
(
commercial
HPLC/
DI),
light
exposure
(
foil
wrap/
standard
light
exposure)
was
carried
out
using
continuous
liquid­
liquid
extraction
exclusively.
This
was
to
test
the
most
probable
off
the
shelf
fixes
and
determine
whether
or
not
light
exposure
and
reagent
water
were
part
of
the
problem.
Since
this
was
a
full
factorial
design,
all
main
factors
and
interactions
would
be
free
of
aliases,
so
no
confusion
about
attributing
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
107
effects
to
particular
factors
was
anticipated.
Table
2
shows
each
factor
and
the
two
levels
investigated.

Table
1.
Factors
and
Levels
from
First
Factorial
Design
Set
wrap
extraction
glassware
in
aluminum
foil
(
foil
wrap)
standard
exposure
to
light
(
std
light)
H)
exposure
to
fluorescent
light
(
light
exposure)
chiller
cooled
condenser
(
0
°
C)
tap
water
cooled
condenser
(
10
°
C)
G)
CLLE
condenser
temperature
(
cond
temp
°
C)
brand
X
dichloromethane,
residue
analysis
grade
(
X)
brand
Y
dichloromethane
HPLC
grade
(
Y)
F)
contamination
in
dichloromethane
(
DCM)
separatory
funnel
(
sep
fun)
continuous
liquid­
liquid
extraction
(
CLLE)
E)
extraction
method,
heat
input
and
time
may
be
the
key
differences
(
ext
meth)
exhaustive
cleaning:
soap,
water,
chromic
acid,
water,
NaOH,
water,
muffle
oven
(
rig
clean)
standard
cleaning
(
std
clean)
D)
contamination
on
glassware
surfaces
(
cleaning)
add
carbon
trap
to
condenser
top
and
Teflon
tape
all
joints
(
C
trap)
standard
lab
setup
(
std
cond)
C)
contamination
in
lab
air
pulled
in
through
CLLE
condenser
(
lab
air)
commercial
HPLC
reagent
water
(
com
H2O)
standard
deionized
water
(
DI)
B)
contamination
in
reagent
water
(
reagent
H2O)
special
methanol
surrogate
spike
solution
(
methanol)
standard
acetone
surrogate
spike
solution
(
acetone)
A)
use
of
acetone
as
spike
solvent
(
spike
solvent)
­
level
+
level
Factors
Table
2.
Factors
and
Levels
from
Final
Factorial
Design
Set
wrap
extraction
glassware
in
aluminum
foil
(
foil
wrap)
standard
exposure
to
light
(
std
light)
D)
exposure
to
fluorescent
light
(
light
exposure)
special
methanol
surrogate
spike
solution
(
methanol)
standard
acetone
surrogate
spike
solution
(
acetone)
C)
use
of
acetone
as
spike
solvent
(
spike
solvent)
commercial
HPLC
reagent
water
(
com
H2O)
standard
deionized
water
(
DI)
B)
contamination
in
reagent
water
(
reagent
H2O)
brand
X
dichloromethane,
residue
analysis
grade
(
X)
brand
Y
dichloromethane
residue
analysis
grade
(
Y')
A)
contamination
in
dichloromethane
(
DCM)
­
level
+
level
Factors
The
results
of
the
factorial
calculations
indicated
the
most
important
factors
and
factor
interactions
were
spike
solvent,
spike
solvent
&
light
exposure
interaction,
light
exposure.
A
second
group
of
interaction
effects
were
smaller
and
all
related
to
the
DCM
source;
DCM
source
&
spike
solvent,
DCM
source
&
light
exposure
and
DCM
source
&
spike
solvent
&
light
exposure.

The
largest
effects
were
caused
by
the
acetone
surrogate
spike
solvent
and
exposure
to
lab
light.
The
large
two
factor
interaction
also
suggests
that
the
lab
light
has
a
catalytic
effect
on
the
chemical
reactions
which
produce
the
interfering
contaminants.
The
three
factor
interaction
between
acetone,
extraction
method
(
CLLE)
and
dichloromethane
postulated
in
the
first
factorial
was
still
present,
although
represented
as
the
two
factor
interaction
between
DCM
source
and
spike
solvent.
This
interaction
was
now
clear
of
the
alias
with
the
reagent
water
factor.
Although
this
interaction
was
measurable
it
was
small
compared
to
other
effects.
Also,
remember
that
the
DCM
grade
from
supplier
Y
was
changed
from
HPLC
grade
to
residue
analysis
grade.
This
grade
appeared
to
be
cleaner
and
probably
reduced
the
measured
differences
between
DCM
sources.
There
was
also
an
interaction
between
DCM
source
and
lab
light
exposure.
This
was
consistent
with
the
postulated
catalytic
effect
of
light
exposure.
The
three
factor
interaction
between
spike
solvent,
light
exposure
and
DCM
source
had
a
small
but
distinguishable
effect.
Lastly,
no
effect
from
the
two
different
reagent
water
sources
was
seen,
thus
lab
water
contamination
was
not
likely.
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
108
There
were
significant
interactions
between
the
3
main
factors
(
multiplicative
effect
rather
than
additive).
The
spike
solvent
and
light
exposure
are
the
main
problem
sources.
DCM
source
is
important
but
secondary.
Since
the
interactions
between
spike
solvent,
DCM
source
and
light
exposure
appear
strong
this
explains
why
our
early
one
factor
at
a
time
work
was
confusing.

Spike
solvent
shows
up
as
the
most
important
factor,
which
means
that
either
the
acetone
(
or
something
in
the
acetone)
is
reacting
(
light
catalyzed)
with
other
impurities
present
in
the
extraction
system.
These
impurities
are
probably
present
in
the
DCM
since
the
final
concentration
of
contaminants
correlates
with
DCM
source.
Recent
experience
with
a
different
acetone
supplier
showed
even
greater
levels
of
blank
contamination.
One
solvent
supplier
indicated
this
was
quite
likely
since
different
acetone
manufacturing
process
produce
different
secondary
components
in
the
acetone.

The
easiest
solution
to
implement
was
switching
from
acetone
to
methanol
surrogate
and
matrix
spike
solutions.
Simply
changing
the
spike
solvent
made
a
dramatic
difference
as
illustrated
by
Figure
3.
This
extraction,
with
methanol
spiking
solutions,
was
performed
under
the
same
conditions
as
Figure
1.

Figure
3.
OCP/
PCB
chromatogram
of
continuous
liquid­
liquid
extract
(
methanol)

HERBICIDES
The
phenoxy
acid
herbicide
prep
and
analysis
(
Method
8150/
1)
relies
on
"
intentional
chemistry"
at
two
key
steps,
hydrolysis
of
esters
and
methylation
of
acids
to
esters.
The
basic
method
chemistry
was
developed
for
carboxylic
acids
and
their
esters.
Dinoseb,
a
phenol,
is
a
more
recent
addition
to
the
method.
It
appears
that
dinoseb
reacts
with
acetone
during
the
hydrolysis
step
for
solid
samples.
The
reaction
and
its
products
are
currently
unknown.
Shortening
the
hydrolysis
time
can
improve
dinoseb
recovery
but
the
hydrolysis
of
herbicide
esters
becomes
less
efficient
and
results
in
poor
analyte
recovery
for
many
other
herbicides.
However,
substitution
of
methanol
for
acetone
in
spiking
solutions
and
eliminating
acetone
from
the
extraction
improves
dinoseb
recovery.

SEMIVOLATILE
BNAS
A
one
to
one
mixture
of
dichloromethane
and
acetone
(
DCM/
acetone)
is
recommended
by
EPA
methods
when
extracting
low
level
solid
samples
(
Methods
3540
&
3550).
Primary
amines
in
the
TCL
and
Appendix
IX
BNA
lists
show
reduced
recovery
when
concentrated
in
dichloromethane
/
acetone
compared
to
dichloromethane
alone.
Similar
trends
have
been
seen
in
various
Method
Detection
Limit
studies.
The
secondary
and
tertiary
amines
do
not
seem
to
be
affected.
If
anything
the
secondary
and
tertiary
amine
recoveries
are
slightly
higher
when
acetone
is
present.

Similar
phenomena
were
reported
by
McNally
(
Anal.
Chem.
1993,
65,
596).
Primary
aromatic
amines
showed
reduced
recovery
when
concentrated
by
KD
when
using
DCM/
acetone.
Concentration
with
a
Zymark
TurboVap
improved
analyte
recovery.
Presumably
this
is
because
it
used
a
lower
concentration
temperature.
They
suggested
that
analyte
losses
may
be
largely
due
to
ion
exchange
or
hydrogen
bonding
adsorption
on
glassware
surfaces.

One
senior
Quanterra
chemist
suggested
that
acetone
may
be
reacting
with
the
primary
amine
to
form
the
imine
while
being
heated
in
the
concentration
step.
The
general
equation
is
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
109
Imine
peaks
were
found
in
the
chromatogram
for
most
amines
that
showed
reduced
recovery.
Figure
7
shows
two
of
the
imine
peaks
that
are
visible
on
the
total
ion
chromatogram.
The
peaks
at
8.0
and
22.6
min.
are
the
imine
forms
of
aniline
and
3,3'­
dichlorobenzidine.
The
mass
spectra
and
fragment
assignments
are
shown
in
Figures
8­
10.
Since
3,3'­
dichlorobenzidine
is
a
diamine
it
shows
both
the
mono­
imine
and
di­
imine
forms.

Based
on
analyte
recovery
2­
Nitroaniline
does
not
react
under
these
conditions
and
4­
nitroaniline
appears
to
only
react
slightly.
a
,
a
-
Dimethyl
phenethylamine
has
very
poor
chromatographic
peak
shape
and
is
a
consistent
calibration
problem,
thus
it
is
not
surprising
that
it
does
not
follow
the
same
trend
as
the
remainder
of
the
primary
amines.

Figure
5.
Secondary
Amine
Recovery
from
DCM
and
DCM/
acetone
KD
Concentrations
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
110
Figure
4.
Primary
Amine
Recovery
from
DCM
and
DCM/
acetone
KD
Concentrations
Figure
7.
Imine
Peaks
in
BNA
Chromatogram
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
111
Figure
6.
Tertiary
Amine
Recovery
from
DCM
and
DCM/
acetone
KD
Concentrations
Figure
9.
Mass
Spectrum
of
Mono­
Imine
Form
of
3,3'­
dichlorobenzidine
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
112
Figure
8.
Mass
Spectrum
of
Imine
Form
of
Aniline
Figure
10.
Mass
Spectrum
of
Di­
Imine
Form
of
3,3'­
dichlorobenzidine
Imine
peaks
and
spectra
were
found
for:
4­
chloroaniline,
3­
nitroaniline,
o­
toluidine,
p­
phenylene
diamine,
1­
naphthylamine,
2­
naphthylamine,
4­
aminobiphenyl,
3,3'­
dimethylbenzidine
(
and
di­
imine)
and
N­
nitro­
o­
toluidine.

Since
a
polar
co­
solvent
is
needed
to
aid
extraction
from
wet
samples
simply
eliminating
the
acetone
may
reduce
extraction
efficiency
for
all
nonpolar
analyses.
It
may
be
possible
to
improve
amine
recovery
by
reducing
the
percentage
of
acetone
in
the
solvent
mixture
or
substituting
an
alternative
polar
solvent.
Methanol
would
not
be
a
good
substitute
since
its
boiling
point
is
significantly
higher
than
acetone.
This
would
greatly
extend
KD
concentration
time
and
raise
the
concentration
temperature
which
might
cause
other
unintended
reactions.
Also,
if
too
much
methanol
was
present
in
the
final
concentrated
BNA
extract,
chromatographic
performance
would
be
degraded.
Using
a
polar
tertiary
amine
such
as
dimethylethyl
amine
instead
of
acetone
should
eliminate
the
conversion
of
primary
amines
to
their
respective
imine
forms.
However,
this
solvent
has
practical
constraints
related
to
purity,
availability
and
cost
at
this
point
in
time.

CONCLUSION
The
three
examples
above
suggest
that
acetone
is
an
active
participant
in
"
accidental
chemistry"
reactions
which
degrade
analytical
results
based
on
standard
EPA
sample
preparation
methods
for
some
target
analyses.
Substituting
methanol
produces
better
results
for
the
organochlorine
pesticide
and
dinoseb
examples.
A
good
means
of
preventing
the
conversion
of
primary
amines
to
imines
is
not
available
at
present.
However,
knowing
that
it
is
a
systematic
limitation
of
the
preparation
methods
can
aid
in
data
interpretation.

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
DETERMINATION
OF
NITROAROMATIC,
NITRAMINE,
AND
NITRATE
ESTER
EXPLOSIVES
IN
WATER
USING
SOLID
PHASE
EXTRACTION
AND
GC­
ECD
Marianne
E.
Walsh,
Chemical
Engineer
U.
S.
Army
Cold
Regions
Research
and
Engineering
Laboratory,
Hanover,
New
Hampshire
03755­
1290
Thomas
Ranney,
Staff
Scientist
Science
and
Technology
Corporation,
Hanover,
New
Hampshire
03755
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
113
ABSTRACT
SW­
846
Method
8330,
the
current
USEPA
method
for
the
analysis
of
14
nitroaromatic
and
nitramine
explosives
and
co­
contaminants,
uses
a
liquid
chromatograph
(
LC)
equipped
with
a
UV
detector.
In
many
environmental
laboratories,
gas
chromatographs
(
GCs)
are
the
most
commonly
used
instruments
because
the
majority
of
SW­
846
methods
for
organics
are
gas
chromatographic
methods.
The
desire
to
make
maximum
use
of
gas
chromatography
naturally
leads
to
attempts
to
substitute
GCs
for
LCs
when
analyzing
for
explosives.
However,
quantitative
analysis
of
explosives
by
gas
chromatography
is
complicated
by
the
thermal
lability
of
some
of
the
analyses,
particularly
the
nitramines.
We
have
found,
by
using
high
linear
carrier
gas
velocities,
deactivated
injection
port
liners,
and
short
wide­
bore
capillary
columns,
that
the
Method
8330
analyses
plus
nitroglycerin,
PETN,
and
dinitroaniline
may
be
analyzed
quantitatively
by
GC­
ECD
(
gas
chromatography­
electron
capture
detector).

Water
samples
are
preconcentrated
with
solid
phase
extraction
(
SPE)
and
acetonitrile
(
AcN)
elusion
using
either
Waters
Porapak
RDX
Cartridges
or
Empore
Styrene­
divinylbenzene
membranes.
Typically
for
Method
8330,
acetonitrile
extracts
are
mixed
with
water
prior
to
analysis
by
LC.
The
AcN
extracts
may
be
analyzed
directly
by
GC­
ECD.
Method
detection
limits
in
the
low
part
per
trillion
range
were
obtained
for
most
analyses.

Several
SPE­
AcN
well
water
extracts
from
military
sites
in
the
US
and
Canada
were
analyzed
by
GC­
ECD
and
LC.
Correlation
coefficients
between
the
GC­
ECD
and
LC
concentration
estimates
for
the
analyses
most
frequently
detected
 
HMX
(
Octahydro­
1,3,5,7­
tetranitro­
1,3,5,7­
tetrazocine),
RDX
(
Hexahydro­
1,3,5­
trinitro­
1,3,5­
triazine),
TNT
(
2,4,6­
Trinitrotoluene),
and
TNB
(
1,3,5­
Trinitrobenzene)
were
all
greater
than
0.99.

The
GC
method
provides
greater
sensitivity
than
LC,
but
accurate
calibration
is
more
difficult.
The
UV
detector
used
for
the
LC
analysis
has
much
greater
linear
range
than
the
ECD
used
for
GC
analysis.
In
addition,
the
GC
instrumentation
requires
more
care
than
the
LC.
Specifically,
the
injection
port
liner
must
be
changed
frequently
to
maintain
accurate
determination
of
the
nitramines.

Perhaps
the
most
valuable
asset
of
the
GC
determination,
when
used
in
conjunction
with
LC,
is
the
ability
to
confirm
analyte
presence
based
on
two
different
physical
properties:
vapor
pressure
with
GC
and
polarity
with
LC.
When
detection
is
ambiguous
using
LC,
confirmation
by
GC
will
be
very
useful.

Documentation
and
performance
data
will
be
submitted
to
the
Office
of
Solid
Waste
for
consideration
of
this
method
as
a
standard
for
inclusion
in
SW­
846.

INTRODUCTION
The
current
USEPA
method
for
the
analysis
of
nitroaromatic
and
nitramine
explosives
and
co­
contaminants,
SW846
Method
83301,
involves
extraction
of
water
samples
using
either
salting­
out
or
solid
phase
extraction
and
analysis
of
the
acetonitrile
extract
using
HPLC­
UV.
Certified
reporting
limits2
range
from
0.03
to
0.3
µ
g/
L,
and
are
sufficiently
low
for
determining
if
water
quality
criteria
are
met
for
most
of
the
analyses
for
which
criteria
have
been
determined.

Because
of
the
prevalence
of
gas
chromatographs
in
environmental
labs,
an
alternative
method
for
explosives
based
on
GC
would
provide
another
option
for
analysis.
Some
of
the
8330
analyses
are
already
included
in
current
GC
SW846
methods3.
These
include
the
nitroaromatics
NB,
2,4­
DNT,
2,6­
DNT,
1,3­
DNB,
1,3,5­
TNB,
and
the
isomers
of
NT
(
Table
1).
The
physical
properties
of
some
of
the
other
Method
8330
analyses,
principally
the
nitramines,
would
lead
one
to
believe
that
GC
analysis
would
be
impractical.
High
melting
points,
low
vapor
pressure,
and
thermal
lability
are
characteristic
of
the
nitramines.
For
example,
the
melting
point
of
HMX
is
275
°
C
4,
and
it
is
reported
to
decompose
prior
to
boiling.
In
addition,
the
vapor
pressure
of
HMX
(
10­
14
torr)
is
well
below
what
is
typical
for
GC
analyses.
Nonetheless,
explosives,
including
the
nitramines,
have
been
analyzed
by
GC
for
several
years,
primarily
for
forensic
applications5.
For
the
most
part,
quantitative
results
have
been
limited
to
the
nitroaromatics.

Hable
et
al.
6
were
the
first
to
report
quantitative
GC
determination
of
HMX
in
water.
The
nitroaromatics
2,4­
DNT,
2,6­
DNT,
and
TNT
were
extracted
using
toluene,
and
the
more
polar
nitramines
HMX
and
RDX
were
extracted
from
a
separate
subsample
with
glass
distilled
iso­
amyl
acetate.
Successful
GC
analysis
was
obtained
using
deactivated
injection
port
liners,
high
injection
port
temperatures,
and
short,
wide­
bore
capillary
columns.
Another
factor
was
the
elimination
of
contact
between
the
analyses
and
metal
parts
of
the
injector.
Elution
of
intact
HMX,
not
a
thermal
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
114
degradation
product,
from
the
GC
column
was
confirmed
by
GC/
MS.
The
certified
reporting
limits
were
similar
to
those
obtained
using
Method
83307
for
RDX,
TNT,
and
2,4­
DNT,
significantly
higher
for
HMX,
and
lower
for
2,6­
DNT.

The
goal
of
our
work
was
to
develop
a
GC
method
that
includes
all
the
Method
8330
analyses
in
a
single
extraction
step,
and
that
uses
commercially
available
and
routinely
used
instrumentation.
We
included
other
analyses
as
well
that
might
be
present
in
explosives­
contaminated
water.
We
added
3,5­
dinitroanliline,
the
biotransformation
product
of
TNB,
and
the
nitrate
esters
NG
and
PETN.
To
complement
Method
8330,
we
sought
to
use
a
compatible
sample
preparation
method
so
that
a
single
extract
could
be
subjected
to
two
methods
of
analysis,
thereby
allowing
direct
comparisons
of
concentration
estimates
obtained
by
the
two
methods
and
providing
another
method
for
analyte
confirmation.

Table
1.
Current
SW­
846
gas
chromatographic
methods1,
which
include
some
of
the
Method
8330
analyses.
Estimated
Quantitation
Limits,
if
reported,
are
listed
in
parenthesis
next
to
the
method
number.

8091,
8260,
8270
(
10
µ
g/
L)
8091,
8270
(
10
µ
g/
L)
8091,
8270
(
10
µ
g/
L)
8091,
8270
(
20
µ
g/
L)
8270
(
10
µ
g/
L)
8091
NB
2,4­
DNT
2,6­
DNT
1,3­
DNB
1,3,5­
TNB
o­,
m­,
p­
NT
SW­
846
Methods
(
Estimated
Quantitation
Limit)

8091:
Nitroaromatics
and
Cyclic
Ketones:
Capillary
Column
Technique.
8260:
Volatiles
Organic
Compounds
by
GC/
MS:
Capillary
Column
Technique.
8270:
Semivolatile
Organic
Compounds
by
GC/
MS:
Capillary
Column
Technique.

EXPERIMENTAL
Calibration
Standards.
Analytical
standards
were
prepared
from
SARM
(
standard
analytical
reference
materials)
obtained
from
the
U.
S.
Army
Environmental
Center,
Aberdeen
Proving
Ground,
Maryland.
Stock
solutions
were
prepared
in
acetonitrile.
Calibration
standards
were
prepared
in
acetonitrile
over
the
concentration
ranges
shown
in
Table
2.

Matrices.
Blank
matrices
used
for
spike
recovery
and
method
detection
limit
studies
were
reagent
grade
(
Type
1)
water
(
MilliQ,
Milipore)
and
groundwater
from
a
domestic
well
in
Weathersfield,
Vermont.
Field­
contaminated
samples
were
obtained
from
Louisiana
AAP
(
Doyline,
Louisiana),
Kansas
AAP
(
Parsons,
Kansas),
Umatilla
Army
Depot
(
Hermiston,
Oregon),
and
DREV
(
Defence
Research
Establishment
Valcartier,
Quebec),

Sample
Preparation.
For
each
sample,
up
to
1000
mL
of
water
was
preconcentrated
using
Solid
Phase
Extraction
according
to
manufacture's
directions.
Both
Waters
Sep­
Pak
Vac
Porapak
RDX
Cartridges
and
Empore
SDB­
RPS
47­
mm
Membranes
were
used.
The
solid
phases
were
eluted
with
4
mL
of
acetonitrile,
and
each
extract
was
directly
injected
into
the
GC­
ECD.
When
necessary,
field
sample
extracts
were
diluted
with
acetonitrile
so
that
peak
heights
would
be
bracketed
by
calibration
standards.

Instrument.
We
configured
the
GC
based
on
the
work
of
Hable
et
al.
6.
The
GC
parameters
were:
GC:
HP
5890
with
electron
capture
detector
(
Ni63)
Column:
J
and
W
DB­
1,
0.53­
mm
ID,
1.5­
or
3.0­
µ
m
film
(
6
m)
Injection
Port
Liner:
Restek
Direct
Injection
Uniliner
(
deactivated)
Injection
Port
Temperature:
250
°
C
(
varied
from
200
to
300
°
C)
Injection
Volume:
1
µ
L
Carrier:
Hydrogen
(
linear
velocity
varied
from
30
to
185
cm/
s)
Makeup:
Nitrogen
(
38
mL/
min.)
Oven
Program:
100
°
C
for
2
min.,
10
°
C
per
min.
ramp
to
200
°
C,
20
°
C
per
min.
to
250
°
C,
5
min.
hold.
Detector
Temperature:
300
°
C
Temperature
programs
for
confirmation
columns
are
given
in
captions
later
in
this
paper.
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
115
Table
2.
Concentration
ranges
for
calibration
standards
diluted
with
acetonitrile.

Set
1:
DNB,
2,6­
DNT,
2,4­
DNT,
TNB,
TNT,
4­
Am­
DNT,
2­
Am­
DNT
Set
2:
3,5­
DNA,
Tetryl
Set
1:
NB,
RDX
Set
1:
o­
NT,
m­
NT,
p­
NT,
HMX
Set
2:
PETN
Set
2:
NG
0.5
to
500
2.5
to
500
1.0
to
1000
5
to
5000
25
to
5000
50
to
10,000
Conc.
(
µ
g/
L)
Range
RESULTS
AND
DISCUSSION
Initial
Tests.
Initially,
we
configured
the
GC
based
on
the
work
of
Hable
et
al.
6
except
that
we
used
a
15­
m
DB­
1
column
as
provided
by
the
manufacturer.
With
the
exception
of
HMX,
which
did
not
produce
a
peak,
the
8330
analyses
eluted
as
individual
peaks,
indicating
that
this
column
provides
adequate
resolution
for
these
analyses.
However,
the
additional
analyte
PETN,
which
has
a
vapor
pressure
almost
identical
to
that
of
RDX,
coeluted
with
RDX.

We
analyzed
blanks
of
commercially
available
toluene
and
iso­
amyl
acetate,
the
two
extraction
solvents
used
by
Hable
et
al.
6.
We
also
analyzed
blanks
of
acetone
and
acetonitrile
(
AcN)
after
passage
through
SPE
cartridges
(
Waters
Porapak
RDX).
While
the
chromatogram
of
iso­
amyl
acetate
contained
numerous
large
peaks,
the
chromatograms
for
the
other
solvents
had
no
significant
peaks.
Because
acetonitrile
produced
no
background
interference
and
Method
8330
specifies
acetonitrile
for
extraction,
we
pursued
the
use
of
solid
phase
extraction
with
acetonitrile
elusion
as
described
below.

Figure
1.
Calibration
standard
(
from
set
1
in
Table
2)
analyzed
using
DB­
1
at
two
carrier
gas
velocities.
Higher
linear
velocities
resulted
in
higher
HMX
peak
heights.

We
experimented
with
different
temperature
programs
and
injected
a
high
concentration
solution
of
HMX.
With
a
high
temperature
isothermal
run,
HMX
eluted
as
a
broad
jagged
peak.
We
next
shortened
the
GC
column
to
6
m
and
found
that
HMX
now
eluted
as
a
sharp
peak.
This
dramatic
improvement
was
not
attributable
to
total
time
in
the
GC,
but
rather
to
the
decreased
column
length,
thus
less
surface
area
to
which
the
analyte
was
exposed.
The
retention
time
for
HMX
was
the
longest
of
all
the
analyses
(
Fig.
1a).
Late
elusion
was
expected
because
of
the
extremely
low
vapor
pressure
of
HMX.
As
we
adjusted
the
linear
velocity,
we
noticed
that
the
HMX
peak
height
changed
significantly
with
changes
in
linear
velocity
(
Fig.
1),
so
we
systematically
changed
the
linear
velocity
to
document
this
effect.
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
116
Effect
of
Carrier
Gas
Linear
Velocity.
Optimum
linear
velocity
for
peak
resolution
is
26
cm/
s
when
using
hydrogen
carrier
gas
and
a
0.53­
mm
ID
column.
Restek
recommends
twice
the
optimum
linear
velocity
when
using
the
Direct
Injection
Uniliners.
We
tested
the
effect
of
increasing
carrier
gas
linear
velocity
over
the
range
30­
185
cm/
s
and
found
a
significant
increase
in
response
from
HMX,
RDX,
NG,
and
PETN.
For
example,
no
peak
for
HMX
was
observed
at
the
lowest
linear
velocity
tested.
The
linear
velocity
was
increased
to
55
cm/
s
and
HMX
eluted
as
a
sharp
peak.
Thereafter,
the
HMX
peak
height
approximately
doubled
with
each
doubling
of
the
linear
velocity
(
Fig.
2).
Some
degradation
in
peak
resolution
did
occur.
With
increasing
carrier
gas
linear
velocity,
the
peak
for
dinitroaniline
merged
with
the
peak
for
4­
amino­
DNT.

Effect
of
Injection
Port
Temperature.
Hable
et
al.
6
found
increased
HMX
response
with
increasing
injection
port
temperature,
and
recommended
an
injection
port
temperature
of
270
°
C
for
the
determination
of
TNT,
DNTs,
RDX,
and
HMX.
We
reexamined
the
effect
of
injection
port
temperature
at
high
linear
velocity
(
133
cm/
s)
for
the
8330
analyses
plus
NG,
PETN,
and
DNA.
We
found
that
the
optimum
temperatures
were
different
for
the
different
analyses.
The
lowest
temperatures
tested
(
200­
220
°
C)
resulted
in
the
highest
response
for
the
nitrotoluenes
and
nitrate
esters.
Higher
temperatures
(
250­
270
°
C)
were
best
for
HMX,
RDX,
the
amino­
DNTs,
and
DNA.
However,
even
at
250
°
C,
responses
for
the
other
analyses
were
at
least
90%
of
the
maximum
responses.
Therefore,
we
recommend
an
injection
port
temperature
of
250
°
C.

Figure
2.
Effect
of
carrier
gas
linear
velocity
on
peak
height
of
HMX
in
a
500­
µ
g/
L
solution.

Calibration.
For
all
of
the
analyses,
the
calibration
factors
(
CF)
decreased
with
increasing
analyte
concentration
(
Fig.
3);
therefore,
a
linear
model
relating
response
to
concentration
is
not
appropriate.
SW­
8463
lists
four
options,
in
order
of
increasing
difficulty,
for
non­
linear
calibration
data:
adjust
the
instrument
or
perform
instrument
maintenance;
narrow
the
calibration
range
until
response
is
linear
(<
20%
RSD
for
CF);
use
a
linear
calibration
model
that
does
not
pass
through
origin;
use
a
calibration
curve
or
non­
linear
model.

ECDs
typically
have
a
narrow
linear
range
(
approximately
40­
fold),
with
a
dynamic
range
of
about
1000­
fold8.
The
calibration
data
we
observed
fell
within
these
specifications.
The
shape
of
the
curve
of
peak
height
data
for
2,6­
DNT
(
Fig.
3)
over
the
range
0.508­
508
µ
g/
L
was
representative
for
other
analyses.
Increasing
the
make­
up
gas
was
suggested
as
a
potential
means
to
increase
the
linear
range.
Therefore,
we
tried
increasing
the
make­
up
gas
from
37
to
69.5
mL/
min.
This
increase
did
not
improve
linearity;
it
only
decreased
the
GC
response.
Next,
we
narrowed
the
calibration
range
to
the
five
lowest
standards;
in
general,
the
calibration
factors
for
most
of
the
analyses
were
within
20%
RSD.
This
very
limited
linear
range
of
the
ECD
is
a
disadvantage
compared
to
the
HPLC­
UV,
which
has
a
broad
linear
range.
For
GC­
ECD,
sample
extracts
will
need
to
be
diluted
within
the
proper
calibration
range.
For
samples
with
multiple
analyses
at
varying
concentrations,
a
single
extract
may
require
several
determinations
at
different
dilution
factors.

Instability
of
Low
Concentration
Standards.
The
calibration
standards
for
TNB
and
TNT
were
unstable
when
left
at
room
temperature
in
amber
autosampler
vials.
Previous
stability
studies
had
shown
that
these
analyses
were
stable
for
several
days
in
acetonitrile9.
However,
the
standards
in
this
previous
study
were
much
higher
in
concentration
(
3
mg/
L).
We
found
that
analyte
loss
was
most
noticeable
at
the
lower
concentrations
(
50
vs.
500
µ
g/
L),
and
that
the
loss
differed
with
different
brands
of
acetonitrile
and
was
slowed
by
refrigeration
of
the
solution
(
Fig.
4).
We
were
particularly
concerned
about
this
instability
because
samples
and
standards
could
potentially
sit
in
an
autosampler
for
several
hours
close
to
a
heated
injection
port
and
GC
oven
vent.
The
autosampler
we
used
(
HP
6890)
was
designed
so
that
a
coolant
could
be
circulated
through
the
tray
containing
the
sample
vials.
With
this
modification,
the
standards
were
stable
over
a
typical
12­
hour
analytical
shift.
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
117
Figure
3.
Typical
calibration
curve
obtained
by
Figure
4.
Decrease
in
peak
height
observed
in
10­
µ
g/
L
GC­
ECD.
TNT
calibration
standard
stored
in
vials
at
three
different
temperatures.

Confirmation
Columns.
We
have
tested
four
0.53­
mm
ID
columns
for
suitability
as
confirmation
columns.
In
order
of
increasing
polarity,
these
columns
are
J
and
W
DB­
1301
(
6%
cyanopropylphenyl
methylpolysiloxane),
J
and
W
DB­
17
(
50%
phenyl
methylpolysiloxane),
a
Restek
RTX­
200
(
Crossbond
trifluoropropyl
methylpoly­
siloxane),
and
a
Restek
RTX­
225
(
50%
cyanopropylmethyl­
50%
phenyl
methylpolysiloxane).
The
DB­
1301
was
not
acceptable
because
TNB
coeluted
with
TNT
and
DNB
coeluted
with
2,6­
DNT.
The
DB­
17
was
not
suitable
because
TNB
coeluted
with
TNT.
The
Restek
RTX­
200
resolved
the
8330
analyses
at
low
linear
velocity,
but
HMX
is
not
detected
(
Fig.
5a).
At
high
linear
velocity,
HMX
was
detected
(
Fig.
5b),
but
PETN
coeluted
with
RDX
and
DNA
coeluted
with
4­
Am­
DNT.
Finally,
on
the
RTX­
225,
tetryl
co­
eluted
with
RDX
and
HMX
was
not
detected
(
Fig.
5c).
However,
in
subsequent
analysis
of
well­
water
samples
from
Louisiana
AAP,
we
found
this
column
to
be
excellent
for
confirmation
of
the
amino­
DNTs.
Thus,
for
confirmation,
extracts
must
be
analyzed
under
the
appropriate
conditions.
Both
the
RTX­
200
and
RTX­
225
look
promising
for
confirmation
because
the
elution
order
of
several
analyses
is
the
reverse
of
that
on
the
DB­
1.
For
example,
2,6­
DNT
and
DNB
reverse
order
as
do
TNT
and
TNB.
In
addition,
RDX
elutes
after
the
amino­
DNTs,
whereas
it
elutes
before
the
amino­
DNTs
on
the
DB­
1.

We
also
tested
an
Alltech
MultiCapillary
SE­
54
(
5%
phenyl
methylpolysiloxane)
column.
These
columns
are
only
1
m
long
and
are
composed
of
a
bundle
of
over
900
liquid­
phase
coated
40­
µ
m
capillaries.
They
provide
rapid
analysis
of
pesticides,
and
accommodate
high
carrier
gas
velocities,
so
we
reasoned
that
they
might
be
suitable
for
the
analysis
of
explosives.
We
tested
numerous
chromatographic
conditions
and
found
that
the
column
was
suitable
for
the
analysis
of
NB,
the
nitrotoluenes,
DNB,
and
the
DNTs.
Resolution
of
the
other
analyses
was
poor,
the
peaks
for
TNB,
TNT,
and
RDX
were
uncharacteristically
small,
and
HMX
did
not
elute
at
all.
Here
again,
large
column
surface
area,
not
total
time
in
the
GC,
may
contribute
to
HMX
loss.

Splitless
injection
Port.
One
objective
of
this
method
development
was
to
use
standard
laboratory
equipment.
We
use
a
packed
column
injection
port
that
was
modified
to
accept
a
0.53­
mm
ID
capillary
column.
(
Conversion
kits
designed
to
convert
most
packed
column
chromatographs
are
commercially
available
from
a
variety
of
sources
including
HP,
J
and
W,
Restek.)
Some
GCs
may
be
equipped
solely
with
Split/
Splitless
injection
ports.
Although
splitless
injection
is
routinely
used
for
trace
analysis,
this
injection
technique
is
not
appropriate
for
reactive
or
high
boiling
compounds
such
as
explosives
because
of
adsorption,
condensation,
and
discrimination
in
the
injection
port.
However,
injection
port
liners
are
available
that
allow
direct
injections
to
be
made
in
a
splitless
capillary
port.
We
tested
a
Restek
Uni­
liner
that
is
designed
for
HP
5890
split/
splitless
injection
port
and
found
no
difference
in
response
than
that
obtained
previously.
Therefore,
either
configuration
appears
to
be
acceptable
for
this
analysis.

Initial
Spike
Recovery.
A
spike
recovery
study
was
performed
using
both
the
Empore
SDB­
RPS
(
47­
mm­
diameter)
disks
and
the
Water
Sep­
Pak
Vac
Porapak
RDX
cartridges.
Both
of
these
solid
phase
extraction
devices
have
been
used
successfully
for
the
preconcentration
of
explosives
from
water
prior
to
HPLC­
UV7.
Also,
prior
to
the
availability
of
commercially
packed
cartridges,
Richard
and
Junk10
demonstrated
the
feasibility
of
preconcentrating
explosives
from
water
onto
solid
styrene
divinylbenzene
resins
in
the
field;
analyses
were
eluted
in
the
laboratory
and
determined
by
GC­
ECD.
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
118
a.
RTX­
200,
LV=
70
cm/
s,
oven
100
°
C
for
1.2
min.,
5
°
C/
min.
to
140
°
C,
1
°
C/
min.
to
160
°
C,
20
°
C/
min.
to
250
°
C,
5
min.
hold.
Injector
250
°
C.
Detector
290
°
C.

b.
RTX
200,
LV
=
122
cm/
s,
oven
150
°
C
for
1
min.,
20
°
C/
min.
to
250
°
C,
5
min.
hold.
Injector
270
°
C.
Detector
290
°
C.

c.
RTX
225,
LV
=
90
cm/
s,
oven
100
°
C
2
min.,
10
°
C/
min.
to
220
°
C,
8
min.
hold.
Injector
220
°
C
(
column
upper
limit).
Detector
250
°
C.

Figure
5.
Chromatograms
obtained
following
injection
of
a
standard
(
set
1
from
Table
2)
onto
confirmation
columns.

Using
both
membranes
and
cartridges,
we
preconcentrated
duplicate
samples
spiked
at
5
µ
g/
L
for
most
of
the
analyses
(
Table
3).
We
divided
each
acetonitrile
extract
and
analyzed
each
by
GC­
ECD
and
HPLC­
UV.
We
found
good
recovery
for
all
the
analyses
by
both
methods.
In
general,
repeatability
was
better
using
HPLC­
UV.

Field
Samples.
We
analyzed
several
solid
phase
extracts
of
water
samples
collected
from
various
explosives­
contaminated
sites.
These
included
extracts
from
LAAP
in
which
500­
mL
samples
were
preconcentrated
using
Porapak
RDX
cartridges
and
eluted
with
5
mL
AcN.
These
extracts
were
prepared
and
analyzed
by
HPLC
at
the
U.
S.
Army
Engineer
Waterways
Experiment
Station
(
Vicksburg,
Mississippi).
Water
samples
from
Umatilla
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
119
Army
Depot
and
DREV
were
preconcentrated
at
CRREL
using
either
cartridges
or
Empore
membranes,
with
the
HPLC
analysis
performed
at
CRREL.
All
GC­
ECD
analysis
was
done
at
CRREL.

Concentration
estimates
obtained
by
the
two
methods
of
determination
for
the
most
commonly
found
analyses
(
HMX,
RDX,
TNT,
TNB,
and
2,4­
DNT)
(
Table
4)
compared
favorably
for
most
samples
over
a
wide
range
of
concentrations.
Discrepancies
between
the
two
methods
of
analysis,
however,
do
exist.
The
GC
appeared
to
underestimate
the
concentration
of
RDX
in
some
of
the
low
concentration
samples.
However,
the
ECD
is
a
more
selective
detector,
so
this
apparent
underestimation
may
not
be
real.
Secondly,
tetryl
was
detected
by
GC
in
some
LAAP
extracts,
but
not
by
HPLC.
We
suspect
that,
when
we
analyze
a
tetryl
standard
by
GC,
the
peak
we
observe
actually
corresponds
to
a
thermal
degradation
product
of
tetryl,
possibly
n­
methyl­
picramide11.
Several
LAAP
water
samples
are
also
contaminated
with
picric
acid,
and
a
co­
contaminant
of
picric
acid
is
potentially
the
source
of
the
peak
we
observe
on
the
GC.
Finally,
2,6­
DNT
was
detected
by
GC­
ECD
in
almost
every
sample
that
contained
2,4­
DNT.
These
isomers
often
co­
elute
on
the
LC­
18
separation
specified
in
Method
8330.
However,
these
isomers
can
be
resolved
on
other
LC
columns12,
specifically
those
with
3­
µ
m
phase
particles,
which
are
less
rugged
for
routine
analysis
of
large
number
of
samples.

Table
3.
Recovery
and
repeatability
of
GC
and
HPLC
determinations
of
analyte
concentrations
in
spiked
water
samples.

a.
Empore
SDB­
RPS
47­
mm
diameter
3.7%
3.9%
8.1%
9.8%
0.2%

5.2%
2.2%
106%
102%
107%
115%
103%

105%
93%
5.26
10.2
5.18
5.48
10.3
10.3
46.9
5.45
10.6
5.62
6.04
10.3
10.8
45.9
5.06
10.2
5.04
5.01
10.0
10.1
50.1
HPLC­
UV
DNB
2,6­
DNT
and
2,4­
DNT**
TNB
TNT
RDX
4­
Am­
2,6­
DNT
and
2­
Am­
4,6­
DNT**
HMX
9.4%
8.7%
6.1%
1.7%
1.9%
2.4%
5.1%
9.5%
5.6%
90%
92%
91%
85%
93%
94%
87%
109%
96%
4.35
4.48
4.50
4.25
4.63
9.32
4.28
5.22
47.0
4.77
4.88
4.78
4.33
4.72
9.55
4.51
5.74
49.7
5.06
5.08
5.12
5.04
5.01
10.0
5.06
5.02
50.1
GC­
ECD
DNB
2,6­
DNT
2,4­
DNT
TNB
TNT
RDX
4­
Am­
2,6­
DNT
2­
Am­
4,6­
DNT
HMX
RPD
(%)
Average
Recovery
(%)
Membrane
2
Membrane
1
Spiked
Conc
(
µ
g/
L)
Found
Concentration
(
µ
g/
L)
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
120
b.
Waters
Sep­
Pak
Vac
Porapak
RDX
Cartridges
1.1%
0.3%
0.7%
0.4%
3.3%

0.4%
1.3%
113%
108%
113%
119%
123%

105%
111%
5.70
11.0
5.67
5.99
12.1
10.6
56.2
5.76
11.0
5.71
5.97
12.5
10.6
55.5
5.06
10.16
5.04
5.01
10.0
10.1
50.1
HPLC­
UV
DNB
2,6­
DNT
and
2,4­
DNT**
TNB
TNT
RDX
4­
Am­
2,6­
DNT
and
2­
Am­
4,6­
DNT**
HMX
11.1%
8.3%
4.6%
3.8%
3.7%
1.8%
9.6%
8.1%
1.6%
98%
100%
96%
96%
103%
106%
95%
101%
136%
4.66
4.87
4.80
4.73
5.07
10.6
4.58
4.85
67.7
5.20
5.29
5.03
4.92
5.26
10.8
5.05
5.26
68.8
5.06
5.08
5.12
5.04
5.01
10.0
5.06
5.02
50.1
GC­
ECD
DNB
2,6­
DNT
2,4­
DNT
TNB
TNT
RDX
4­
Am­
2,6­
DNT
2­
Am­
4,6­
DNT
HMX
RPD
(%)
Average
Recovery
(%)
Cartridge
2
Cartridge
1
Spiked
Conc
(
µ
g/
L)
Found
Concentration
(
µ
g/
L)

**
Peak
not
resolved.

Table
4.
Concentration
estimates
obtained
for
the
most
commonly
found
analyses
by
HPLC
and
GC­
ECD
for
water
samples
collected
at
explosives­
contaminated
sites.

a.
HMX
0.10
0.60
0.59
0.21
13
26
110
109
179
147
217
280
308
285
1378
1842
0.20
0.29
0.22
0.31
19
26
97
116
141
182
216
219
250
251
1300
1860
Cartridge
Membrane
Membrane
Cartridge
Cartridge
Cartridge
Cartridge
Cartridge
Cartridge
Cartridge
Cartridge
Cartridge
Cartridge
Cartridge
Cartridge
Cartridge
KSS
AAP
Umatilla
Umatilla
Umatilla
LAAP
DREV
DREV
LAAP
Umatilla
LAAP
LAAP
DREV
DREV
DREV
LAAP
LAAP
GC­
ECD
HPLC
SPE
Method
Source
HMX
Conc.
(
µ
g/
L)
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
121
b.
RDX
0.2
0.20
0.95
1.0
0.56
0.28
0.6
4.9
5.1
5.2
7.0
1.2
37.7
34.7
44.1
59O
1973
2241
3640
8175
20833
0.2
0.27
1.6
1.7
2.0
2.4
3.6
4.9
5.2
6.5
6.7
8.9
29.8
30.9
33.8
845
1430
2060
3710
11800
23400
Cartridge
Membrane
Cartridge
Cartridge
Cartridge
Cartridge
Cartridge
Membrane
Membrane
Cartridge
Cartridge
Cartridge
Cartridge
Cartridge
Cartridge
Cartridge
Cartridge
Cartridge
Cartridge
Cartridge
Cartridge
KSS
AAP
Umatilla
KSS
AAP
DREV
DREV
LAAP
LAAP
Umatilla
Umatilla
Umatilla
Umatilla
LAAP
DREV
DREV
DREV
LAAP
LAAP
LAAP
LAAP
LAAP
LAAP
GC­
ECD
HPLC
SPE
Method
Source
RDX
Conc.
(
µ
g/
L)

c.
TNT
0.5
0.1
0.3
1.2
142
233
405
2876
3721
7781
12168
0.3
0.4
0.5
2.4
152
241
390
2430
2890
7500
10500
Cartridge
Cartridge
Cartridge
Cartridge
Cartridge
Cartridge
Cartridge
Cartridge
Cartridge
Cartridge
Cartridge
LAAP
LAAP
LAAP
LAAP
LAAP
Umatilla
LAAP
LAAP
LAAP
LAAP
LAAP
GC­
ECD
HPLC
SPE
Method
Source
TNT
Conc.
(
µ
g/
L)

d.
TNB
0.02
0.4
1.0
33.8
34.2
1128
782
11991
10640
0.1
1.0
1.9
15.6
22.3
649
742
9110
9150
Cartridge
Cartridge
Cartridge
Cartridge
Cartridge
Cartridge
Cartridge
Cartridge
Cartridge
LAAP
LAAP
LAAP
LAAP
LAAP
LAAP
LAAP
LAAP
LAAP
GC­
ECD
HPLC
SPE
Method
Source
TNB
Conc.
(
µ
g/
L)
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
122
e.
2,4­
DNT
0.07
0.06
0.05
0.15
0.36
11.8
18.7
33.6
126
84.8
341
<
d
<
d
<
d
<
d
0.69
10.7
24.5
46.8
127
142
442
Cartridge
Cartridge
Cartridge
Cartridge
Cartridge
Cartridge
Cartridge
Cartridge
Cartridge
Cartridge
Cartridge
LAAP
LAAP
LAAP
LAAP
LAAP
LAAP
LAAP
LAAP
LAAP
LAAP
LAAP
GC­
ECD
HPLC
SPE
Method
Source
2,4­
DNT*
Conc.
(
µ
g/
L)

*
In
several
extracts,
2,6­
DNT
was
detected
by
GC­
ECD,
but
not
by
HPLC.

Almost
all
the
extracts
from
field
samples
required
dilution
prior
to
GC­
ECD
analysis
so
that
peak
heights
would
fall
within
the
linear
calibration
range.
Dilution
actually
appeared
to
improve
the
accuracy
of
the
GC
determination
of
HMX
when
several
samples
were
run
sequentially.
We
suspect
that
dilution
served
to
"
cleanup''
the
extracts
and
slowed
the
buildup
of
non­
volatile
co­
extracted
contaminants
that
deposit
in
the
injection
port
liner.
Accurate
determination
of
HMX
required
that
the
injection
port
liner
be
changed
or
cleaned
frequently.
We
changed
the
liner
each
time
we
replaced
the
injection
port
septum,
at
least
every
50
injections.

Method
Detection
Limits.
Method
Detection
Limits
(
MDLs)
were
computed
from
the
standard
deviation
of
the
mean
concentration
of
10
replicate
spiked
water
samples
and
the
appropriate
Student's
t
value.
The
MDLs
were
lowest
for
the
di­
and
tri­
nitroaromatics,
and
were
all
well
below
current
water
quality
criteria
(
Table
5).
If
the
analyte
of
most
interest
is
2,6­
DNT,
the
MDL
could
be
lowered
by
preconcentrating
a
greater
volume
of
water.
We
limited
the
volume
we
preconcentrated
to
prevent
breakthrough
of
HMX
and
RDX.
2,6­
DNT
is
well
retained
on
both
solid
phases,
and
the
volume
of
water
preconcentrated
is
more
likely
limited
by
practical
considerations
such
as
time
or
possible
plugging
of
the
solid
phase.

Based
on
the
Certified
Reporting
Limits
for
Method
8330,
and
using
the
generalization
that
CRLs
are
similar
to
MDLs,
the
GC­
ECD
appears
to
have
lower
detection
limits
(
Table
5).
Given
that
MDLs
are
highly
matrix­
and
laboratory­
specific,
more
data
from
multiple
labs
are
needed
prior
to
generalizing
about
the
magnitude
of
the
improvement
in
detection
capability
in
actual
practice.

Table
5.
Method
detection
limits
(
µ
g/
L)
for
GC­
ECD
method,
certified
reporting
limits
(
µ
g/
L)
for
the
HPLC
method7,
and
water
quality
criteria.

1.0
a
40
a,
0.007
b
50
a,
0.1
b
2.0
a
400
a
0.036
0.044
0.051
0.13
0.124
0.055
0.83
0.33
0.20
0.37
0.23
0.032
0.085
0.042
0.068
0.27
0.046
0.24
0.21
0.10
0.13
0.12
0.0025
0.0025
0.0106
0.0027
0.0196
0.0174
0.0030
0.0044
0.0100
0.0081
0.235
0.351
0.294
0.325
0.225
0.275
0.066
0.0036
0.0029
0.0092
0.0066
0.0144
0.0044
0.0026
0.0030
0.0094
0.0041
0.475
0.352
0.297
0.225
0.257
0.349
0.046
DNB
2,6­
DNT
2,4­
DNT
TNB
TNT
RDX
4­
Am­
2,6­
DNT
2­
Am­
4,6­
DNT
Tetryl
HMX
NB
o­
NT
m­
NT
p­
NT
NG
PETN
DNA
Membrane
Cartridge
Membrane
Cartridge
Water
Quality
Criteria
(
µ
g/
L)
CRLs
by
HPLC
MDLs*
by
GC­
ECD
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
123
*
1L
of
water
preconcentrated
to
4.0
mL
AcN
aEPA
Lifetime
Health
Advisory
Number
bEPA
number
for
increased
cancer
risk
of
1.0
x
10­
6
SUMMARY
A
gas
chromatographic
method
for
the
analysis
of
explosives
in
water
was
developed
to
serve
as
an
alternative
to
the
current
HPLC
SW846
Method
8330.
Water
samples
are
preconcentrated
using
solid
phase
extraction,
and
the
acetonitrile
extracts
are
directly
injected
onto
a
short
(
6­
m)
DB­
1
analytical
column.
High
linear
carrier
gas
velocities
resulted
in
higher
peak
heights
for
the
nitramines
and
nitrate­
esters,
the
most
thermally
labile
analyses.
Detection
limits
ranged
from
0.003
to
0.5
µ
g/
L.

Analysis
of
extracts
from
field
samples
showed
good
agreement
between
the
GC­
ECD
and
the
standard
HPLC
method.

Potential
advantages
over
the
current
HPLC
method
include
lower
detection
limits,
improved
chromatographic
resolution,
and
the
use
of
instrumentation
most
commonly
found
in
environmental
labs.
Disadvantages
of
the
GC
method
include
non­
linear
calibration,
limited
dynamic
range
of
the
detector,
and
increased
attention
to
instrument
maintenance
(
i.
e.,
frequent
changes
of
the
injection
port
liner).

Combined
use
of
GC­
ECD
and
HPLC
will
provide
an
improved
method
for
analyte
confirmation
because
chromatographic
separations
are
based
on
different
physical
properties
(
vapor
pressure
and
polarity)
and
the
detectors
are
based
on
different
principles
(
electronegativity
and
UV
absorption).

ACKNOWLEDGMENTS
The
authors
gratefully
acknowledge
funding
for
this
work
provided
by
the
U.
S.
Army
Engineer
Districts,
Kansas
City
(
Dick
Medary)
and
Omaha
(
Chung­
Rei
Mao
and
Kevin
Coats),
and
the
U.
S.
Army
Environmental
Center
(
Martin
Stutz
and
George
Robitaille).
Water
samples
and
extracts
from
LAAP
were
provided
by
the
U.
S.
Army
Engineer
Waterways
Experiment
Station.
Technical
reviews
were
provided
by
Thomas
F.
Jenkins
and
Martin
Stutz.

REFERENCES
1.
USEPA
(
1994)
Test
Methods
for
Evaluating
Solid
Waste,
Physical/
Chemical
Methods.
SW­
846
Update
II,
Office
of
Solid
Waste,
Washington,
DC.
2.
Hubaux
and
Vos
(
1970)
Analytical
Chemistry,
42:
849­
855.
3.
USEPA
(
1995)
Test
Methods
for
Evaluating
Solid
Waste,
Physical/
Chemical
Methods.
SW­
846
Update
III,
Office
of
Solid
Waste,
Washington,
DC.
4.
Meyer,
R.
(
1987)
Explosives.
VCH
Verlagsgesellschaft
mbH.
Weinheim,
Germany.
5.
Yinon
and
Zitrin
(
1993)
Modern
Method
and
Applications
in
Analysis
of
Explosives.
John
Wiley
and
Sons
Ltd.,
West
Sussex,
England.
6.
Hable,
M.,
C.
Stern,
C.
Asowata,
and
K.
Williams
(
1991)
Journal
of
Chromatographic
Science.
29:
131­
135.
7.
Jenkins,
T.
F.,
P.
H.
Miyares,
K.
F.
Myers,
E.
McCormick,
and
A.
B.
Strong
(
1992)
Comparison
of
cartridge
and
membrane
solid­
phase
extraction
with
salting­
out
solvent
extraction
for
preconcentration
of
nitroaromatic
and
nitramine
explosives
from
water.
CRREL
Special
Report
92­
25.
U.
S.
Army
Cold
Regions
Research
and
Engineering
Laboratory,
Hanover,
New
Hampshire.
8.
McNair,
H
M.
and
E.
J.
Bonelli
(
1969)
Basic
Gas
Chromatography.
Consolidated
Printers,
Berkeley,
California.
9.
Jenkins,
T.
F.
P.
W.
Schumacher,
M.
E.
Walsh
and
C.
F.
Bauer
(
1988)
Development
of
an
analytical
method
for
the
determination
of
explosive
residues
in
soil.
Part
II:
Additional
development
and
ruggedness
testing.
CRREL
Report
88­
8.
U.
S.
Army
Cold
Regions
Research
and
Engineering
Laboratory,
Hanover,
New
Hampshire.
10.
Richard,
J.
J.
and
G.
A.
Junk
(
1986)
Analytical
Chemistry,
58:
723­
725.
11.
Tamiri,
T.
and
S.
Zitrin
(
1986)
Journal
of
Energetic
Materials.
4:
215­
237.
12.
Walsh,
M.
E.,
T.
F.
Jenkins,
P.
S.
Schnitker,
J.
W.
Elwell,
and
M.
H.
Stutz
(
1993)
Evaluation
of
SW846
Method
8330
for
characterization
of
sites
contaminated
with
residues
of
high
explosives.
CRREL
Report
93­
5.
U.
S.
Army
Cold
Regions
Research
and
Engineering
Laboratory,
Hanover,
New
Hampshire.
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
124
COMPREHENSIVE,
QUANTITATIVE,
CONGENER­
SPECIFIC
PCB
ANALYSIS:
WHEN
IS
IT
REQUIRED
AND
WHAT
IS
NECESSARY
TO
ACHIEVE
IT?

George
M.
Frame,
Research
Chemist
General
Electric
Corporate
R&
D
Center,
Bldg.
K1,
Rm
3B32,
P.
O.
Box
8,
Schenectady,
New
York,
12301­
0008
ABSTRACT
Current
regulatory
analyses
for
PCBs
specify
reporting
as
equivalents
of
the
commercial
mixtures
(
Aroclors,
EPA
methods
8081,
draft
8082)
or
short
lists
of
priority
congeners
(
European
BCR,
7
priority
PCBs,
EPA
draft
method
1668).
Studies
which
must
unravel
mechanisms
which
alter
congener
distributions,
such
as
microbial
or
photolytic
dechlorination,
aerobic
microbial
degradation,
or
pyrolytic
rearrangement
and
thermal
synthesis,
require
complete
quantitative
characterization
of
PCB
congener
distributions
by
the
title
procedure
(
abbreviated
CQCS).

CQCS
methods
such
as
Mullin's
EPA
"
Green
Bay
Calibration"
and
the
G.
E.
"
DB­
1
118­
peak
HRGC
System"
have
employed
Aroclor
mixtures
as
congener
standards,
but
these
are
restricted
to
specific
HRGC
columns
and
Aroclor
lots.
To
meet
the
need
for
greater
flexibility
in
developing
CQCS
PCB
analysis
methods,
the
author
has
organized
several
consortia
to
obtain
retention
data
for
all
209
PCB
congeners
on
20
different
HRGC
systems
and
complete
congener
distributions
for
17
Aroclor
mixtures.

From
these
data
this
paper
evaluates
how
to
best
choose
systems
for
CQCS
PCB
analysis,
illustrating
with
actual
data,
various
HRGC
systems
employing
a
variety
of
detectors,
which
have
been
optimized
for
maximum
performance.
From
the
data
in
these
studies,
the
optimum
distribution
of
all
the
congeners
into
a
set
of
9
mixtures
appropriate
for
calibrating
a
wide
variety
of
CQCS
PCB
analyses
has
been
designed
and
is
now
marketed
by
a
commercial
distributor.
A
method
for
the
efficient
employment
of
these
primary
standard
sets
by
combining
the
5
mixtures
comprising
most
of
the
congeners
found
in
Aroclors
mixtures
is
evaluated.
The
most
detailed
and
comprehensive
studies
to
date
of
Aroclor
congener
distributions
has
been
made
using
these
standards
and
HRGC
systems
optimized
for
maximum
congener
resolution.
The
pitfalls
of
using
Aroclors
as
quantitative
standards
for
CQCS
PCB
analyses
become
evident
upon
inspection
of
the
variation
in
distributions
of
congeners
among
different
lots
of
the
same­
numbered
Aroclors.

The
initial
results
of
a
new
collaborative
study
on
the
PCB
congener
detection
limits
and
effective
linear
ranges
of
a
variety
of
HRGC
detectors
(
ECD,
EI­
MS­
SIM,
medium
and
high
resolution
EI­
MS­
SIM,
full­
scan
ion­
trap
MS)
are
described.
The
results
illustrate
the
tradeoffs
in
selectivity,
sensitivity,
cost,
ease
of
use,
and
accuracy
among
the
different
detectors
when
applied
to
CQCS
PCB
analyses.
The
latest
model
bench­
top
instrumentation,
exemplified
by
the
new
Hewlett­
Packard
5973
GC­
MS
instrument,
the
Varian
Saturn
2000
Ion­
Trap
MS
instrument,
and
the
new
Hewlett­
Packard
micro
ECD
detector,
provide
order­
of­
magnitude
improvements
in
sensitivity
and
linearity
for
this
application.
When
these
detectors
are
combined
with
HRGC
capillary
columns
chosen
from
the
database
survey
to
provide
optimal
resolution
of
the
most
congeners,
and
the
new
comprehensive
congener
calibration
sets
are
used
as
primary
standards,
CQCS
PCB
analysis
becomes
a
practical
reality.

INTRODUCTION
Polychlorinated
biphenyls
(
PCBs)
are
comprised
of
209
chlorine­
substituted
biphenyl
compounds
known
as
congeners.
They
are
often
referred
to
by
their
IUPAC
or
Ballschmiter
and
Zell
(
BZ)
numbers.
The
matrix
diagram
of
Figure
1
relates
the
chlorine­
substitution
pattern
on
each
of
the
two
phenyl
rings
to
the
BZ
numbering
system.
It
also
displays
the
ring
position
numbering
conventions
that
underlie
the
more
formal,
complete
IUPAC
nomenclature,
and
the
correlation
between
the
20
different
ring
substitution
patterns
and
the
209
different
BZ­
numbered
congeners.
Only
about
150
of
these
congeners
are
observed
in
the
commercial
mixtures
formerly
manufactured
by
catalytic
chlorination
of
biphenyls
(
Aroclors
(
Monsanto)
in
the
USA)).
The
shading
of
the
cells
in
the
matrix
highlights
both
those
congeners
which
dominate
in
Aroclors
due
to
highly
favored
and
stable
ring
substitution
patterns
as
well
as
those
not
seen
due
to
ring
substitutions
which
are
very
difficult
to
produce
and
are
vulnerable
if
formed
to
rapid
further
chlorination.

The
methods
of
choice
for
PCB
analysis
are
based
on
separation
of
a
majority
of
the
congeners
by
high
resolution
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
125
gas
chromatography
(
HRGC)
on
long
capillary
columns,
with
detection
by
sensitive
and
selective
detectors
such
as
the
electron
capture
detector
(
ECD)
or
mass
spectrometry
in
the
selected­
ion­
monitoring
mode
(
MS­
SIM).
Even
these
powerful
techniques
are
insufficient
to
measure
all
the
congeners
typically
found
in
an
Aroclor
mixture
or
the
altered
patterns
resulting
from
environmental
processes
such
as
microbial
degradation
or
dechlorination,
photolysis,
or
degradation
or
de
novo
synthesis
by
thermal
processes.
If
an
attempt
is
made
to
correctly
assign
all
the
congeners
eluting
or
coeluting
in
each
resolved
HRGC
peak
and
to
quantify
the
total
PCB
content
of
each
resolved
peak,
the
analysis
can
be
described
as
"
A
comprehensive,
quantitative,
congener­
specific
PCB
analysis".
This
will
be
abbreviated
in
this
paper
as
a
CQCS
PCB
analysis.
This
most
demanding
form
of
PCB
analysis
should
be
distinguished
from
the
less
complete
"
congener­
specific"
methods
which
specify
only
a
limited
list
of
target
congeners,
selected
both
for
the
ease
of
resolution,
the
availability
of
suitable
standards,
and
relevance
to
a
particular
question.
Examples
of
the
latter
include
the
7
European
Bureau
of
Community
Reference
(
BCR)
target
congeners
28,
52,
101,
118,
138,
153,
and
180;
or
the
list
of
14
congeners
whose
abundance
and
ability
to
achieve
a
planar
"
dioxin­
like"
conformation
qualify
them
for
inclusion
in
EPA
draft
method
1668.
Several
of
these
lists,
and
the
calibration
sets
that
support
the
analyses
are
tabulated
in
the
bottom,
unshaded
cells
of
Figure
2.
These
congener­
specific
methods
contrast
with
the
earlier
EPA
methods
608,
8080,
8081,
and
draft
method
8082,
where
the
calibration
is
against
the
best
matching
Aroclor
mixture(
s),
and
results
are
reported
as
equivalents
of
particular
numbered
Aroclors.
[
The
Aroclor
mixtures
are
typically
numbered
from
1221
to
1262,
where
12
indicates
the
12
carbon
atoms
in
biphenyl,
hence
a
PCB,
and
the
last
2
digits
the
weight
percentage
of
chlorine
in
the
mixture.]

Figure
1.
Matrix
of
PCB
IUPAC
No.
s
vs
Phenyl
Ring
Cl
Substitution
Positions
+
Aroclor
Abundances
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
126
Figure
2.
Top/
Down:
PCB
coeluting
in
6
Systems
[
Shaded
Cells]
Bottom/
Up:
PCB
Congener
"
Short
Lists"
[
Clear
Cells]

DISCUSSION
In
contrast
to
the
"
short­
list"
congener­
specific
PCB
analytical
methods
alluded
to
in
the
introduction,
CQCS
analysis
will
be
required
to
unravel
the
mechanisms
which
result
in
the
environmental
or
human
alterations
of
the
initially
produced
Aroclor­
like
PCB
distributions.
These
tend
to
be
very
congener
specific,
and
distinctive
to
the
particular
process(
es)
which
may
be
acting.
For
example
microbial
dechlorination
in
anaerobic
sediments
proceeds
primarily
by
selective
removal
of
chlorine
atoms
from
positions
meta­
or
para­
to
the
other
phenyl
ring.
By
contrast
photolytic
dechlorination
is
dominated
by
removal
of
the
ortho­
chlorines.
These
processes
will
often
produce
some
congeners
not
generally
present
in
the
commercial
Aroclor
mixtures.
The
disadvantage
of
this
is
that
the
standard
congener­
specific
PCB
methods
often
do
not
take
account
of
their
possible
presence
and
lack
suitable
calibration
standards
for
them.
However,
if
they
can
be
recognized
and
quantified,
they
are
distinctive
markers
for
the
presence
and
extent
of
these
processes.
Partial
degradation
processes
(
e.
g.
aerobic
microbial
or
thermal)
can
selectively
remove
particular
congeners,
without
the
appearance
of
the
corresponding
product
PCBs
which
are
observed
in
the
dechlorination
cascades.
Highly
resolving
CQCS
analytical
procedures
are
needed
to
characterize
these.
Some
incomplete
thermal
degradations
or
de
novo
syntheses
of
PCB
congeners
will
display
a
"
scrambling"
of
the
chlorine
substitutions
around
the
rings,
again
resulting
in
characteristic
congener
patterns
which
require
CQCS
PCB
analyses
for
their
full
elucidation.
It
is
an
often
overlooked
feature
of
many
industrial
processes
which
involve
phenyl
rings
in
the
presence
of
chlorine
or
chlorinating
reagents
together
with
high
temperature
and
catalytic
conditions,
that
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
127
PCBs
may
form
by
coupling
reactions
of
chlorinated
phenyl
rings.
The
resultant
congener
distributions
will
be
radically
different
from
those
obtained
from
the
catalytic
chlorination
of
biphenyls
used
to
produce
Aroclors,
and
may
easily
escape
detection
if
one
naively
attempts
to
recognize
their
presence
by
comparison
to
Aroclor
congener
distributions.

A
CQCS
PCB
analytical
procedure
which
is
capable
of
quantifying
all
209
congeners
or
even
the
subset
of
about
150
found
in
the
Aroclors
in
a
single
pass
remains
a
"
Holy
Grail",
whose
attainment
is
ever
more
closely
approached,
but
may
remain
forever
beyond
the
grasp
of
PCB
analytical
crusaders.
The
new
technique
of
comprehensive
two­
dimensional
GC
may
eventually
accomplish
this,
but
the
neither
necessary
pair
of
columns
nor
adequate­
quantitation
software
have
been
demonstrated
for
this
application.
Even
the
development
of
the
present,
limited
CQCS
methods
has
been
a
laborious
and
time
consuming
task
for
those
laboratories
which
have
attempted
it.
This
accounts
for
its
employment
being
limited
to
applications
where
it
is
indispensable
and
researchers
have
the
resources
of
time
and
money
to
support
its
development.
To
expedite
this
process,
the
author
organized
in
1994
a
consortium
of
12
labs
to
obtain
retention
order
and
coelution
information
on
all
209
PCB
congeners
on
27
HRGC
systems
spanning
20
different
GC
stationary
phase
compositions.
The
resultant
database
is
due
to
be
published
in
July
1997
1,2.
Data
from
18
of
the
systems
enabled
measurement
of
levels
of
all
significant
congeners
in
6
different
Aroclors.
Knowledge
of
all
congeners
present
in
the
Aroclors
enabled
evaluation
of
the
retention
database
information
to
identify
GC
phases
especially
suitable
for
measuring
the
greatest
percentage
of
the
Aroclor
congeners.
It
was
possible
to
process
the
database
information
to
design
the
minimum
number
of
mixtures
(
5
of
"
Aroclor
congeners"
and
4
of
"
non­
Aroclor
congeners")
which
could
calibrate
systems
employing
the
12
most
useful
GC
phases
without
significant
interference
by
coelutions.
The
information
in
the
database
was
compiled
to
enable
identification
of
congeners
within
each­
mixture
on
any
of
the
12
phases
by
simply
observing
their
order
of
elusion.
AccuStandard
Inc.,
of
New
Haven,
CT,
one
of
the
2
licensed
synthesizers
of
PCB
congeners
in
the
USA,
and
the
only
one
to
offer
all
209,
immediately
prepared
the
mixtures
and
marketed
them
with
the
requisite
tables
of
elusion
orders.
This
drastically
reduced­
the
costs
in
materials
and
time
to
assign
congeners
to
peaks
and
calibrate
proposed
new
systems
for
CQCS
PCB
analysis.

As
a
result
of
the
database
study,
several
columns
especially
suitable
for
the
2
alternative
HRGC
methods
for
CQCS
PCB
analysis
were
identified.
These
methods
are:

1.
Separation
on
a
single
column
with
detection
by
GC­
MS­
SIM,
which
can
provide
in
many
cases
independent
measurement
of
coeluting
congeners
whose
chlorination
level
differs
by
one
chlorine.
The
method
enjoys
the
high
selectivity
of
MS
detection
(
although
isomers
(
same
Cl#)
of
PCBs
in
general
cannot
be
distinguished),
but
it
is
usually
not
as
sensitive
as
the
ECD
detector.

2.
A
single
injection
split
to
a
suitably
chosen
pair
of
HRGC
columns,
each
measuring
the
effluent
with
the
generally
more
sensitive
and
partially
selective
(
not
as
much
as
GC­
MS)
electron
capture
detector
(
ECD).

The
congener
resolution
capability
(
implied
by
a
listing
of
Aroclor
congeners
not
quantifiable)
in
three
successively
more
capable
systems
of
category
2
are
displayed
in
the
first
3
columns
of
Figure
2,
while
the
next
3
columns
do
the
same
for
three
improved
versions
of
category
1.
The
title
cells
for
each
column
list
the
GC
phases
employed,
and
immediately
below
are
the
numbers
of
the
references
containing
more
detailed
information
on
the
systems.
The
IUPAC
numbers
of
PCBs
not
uniquely
measurable
due
to
coelutions
are
displayed
in
the
shaded
cells,
where
the
more
heavily
shaded
ones
represent
major
Aroclor
congeners
(
2
­
12
Wt%)
as
in
Figure
1.
Note
that
System
3
approaches
holy
grail
status
most
closely,
but
that
the
remaining
coelutions
are
almost
all
major
Aroclor
components.
Its
dual­
column
/
ECD
technique
substitutes
confirmation
of
congener
identity
by
a
second
column
elusion
time
for
System
6'
s
confirmation
by
characteristic
molecular
ion
masses
in
the
MS­
SIM
measurement.
In
the
author's
experience,
the
data
from
a
system
like
that
of
System
6
are
more
easily
reduced
and
amenable
to
automated
processing
than
are
the
multiple
categories
of
congener
measurements
in
a
system
like
that
of
System
3.
The
shorter
linear
range
of
older
model
ECDs
makes
it
harder
to
cover
the
1000­
fold
range
of
congener
concentrations
one
may
wish
to
quantify
in
a
complex
Aroclor­
derived
mixture.
This
disadvantage
may
be
alleviated
by
the
new,
even
more
sensitive
and
linear
micro
ECD
just
introduced
by
Hewlett­
Packard.
The
dual
column
ECD
systems
employ
the
least
expensive
instrumentation,
but
the
differential
is
only
on
the
order
of
~
30%
less.

While
HP
is
keeping
the
ECD­
based
systems
in
the
game
with
its
new
detector,
it
is
also
providing
stiff
competition
for
them
with
the
new
bench­
top
5973
GC­
MS
system,
which
lowers
the
effective
measurement
range
for
PCBs
nearly
an
order
of
magnitude
from
that
of
predecessor
models.
The
Varian
Saturn
2000
ion­
trap,
full­
scan,
bench­
top
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
128
GC­
MS
instrument
provides
comparable
data
in
a
similar
price
range,
and
provides
full
scan
data
and
options
for
MS/
MS
data
acquisition.
The
more
expensive
and
complex
high
resolution
MS
instrumentation
which
has
provided
the
greatest
sensitivity
and
selectivity
for
classical
methods
of
trace
dioxin,
PCDF,
and
trace
coplanar
(
e.
g.,
PCB
126,
cf
EPA
draft
method
1668)
PCB
analyses
by
isotope
dilution
MS,
may
not
be
as
suitable
for
routine
CQCS
PCB
analysis
(
if
we
are
willing
to
consider
CQCS
as
routine!)
as
the
cheaper
and
easier
to­
use
benchtop
instrumentation.
Isotope
dilution
HRGC­
HRMS
will
still
continue
to
be
the
method
of
choice
for
precise
quantitation
of
short
lists
of
critical
congeners
such
as
the
trace
levels
of
PCBs
77,
126,
and
169
whose
high
toxic
equivalency
factors
(
TEFs)
sometimes
dominate
such
calculation.
This
latter,
more
demanding
instrumentation
provides
much
better
discrimination
against
the
effects
of
unsuspected
contaminants
in
complex
environmental
samples.

Figure
3
displays
a
comparison
of
the
detection
limits
and
linear
response
ranges
of
the
old
and
new
HP
GC­
MS­
SIM
benchtop
instruments
and
the
old
HP
GC­
ECD
instrumentation
in
the
context
of
CQCS
PCB
analysis.
Average
values
for
3
to
6
congeners
at
each
chlorination
level
from
1
to
9
chlorines
are
displayed.
The
data
were
obtained
from
an
18­
level
serially
diluted
standard
curve
vs
2
internal
standards
at
20
pg
each
over
a
range
of
nearly
6
orders
of
magnitude.
The
lower
dots
represent
the
signal/
noise
=
3.3
detection
limits
and
the
bars
the
±
12%
linearity
range.
The
ranges
within
each
chlorination
level
are
numbered
and
keyed
left
to
right.
Note
the
lower
sensitivity
of
the
ECDs
to
the
less
chlorinated
PCBs.
They
gradually
approach
and
surpass
the
performance
of
the
MS
instruments
as
chlorination
level
increases.
Note
how
the
5973
MS
holds
its
own
vs
the
old
ECD
over
most
of
the
chlorination
levels,
and
performs
almost
10­
fold
better
than
its
predecessor
5972.
Note
the
improvement
conferred
in
extending
the
linear
range
to
lower
levels
by
performing
Gaussian
smoothing
on
noisy
low
level
signals
before
performing
peak
integration.
This
results
primarily
from
the
smoother
baseline
allowing
the
integration
software
to
more
consistently
place
the
start,
stop
and
baseline
levels
of
peaks
which
are
close
to
the
noise
limit.
This
graph
represents
preliminary
results
from
a
more
extensive
study
which
will
also
evaluate
the
Varian
Saturn
2000
ion­
trap
GC­
MS,
and
several
GC­
HRMS
systems,
and
the
new
micro­
ECD.
A
second
leg
with
PCBs
at
the
same
levels
in
1000
ppm
complex
petroleum
oil
contaminated
samples
will
be
run
to
evaluate
the
advantages
of
HRMS
instrumentation
in
dealing
with
highly
contaminated
samples.

As
a
follow­
up
to
the
Aroclor
distribution
database
study2
a
second
collaboration
employed
one
example
of
dual
column/
ECD
System
3
and
2
examples
of
single
column/
MS­
SIM
System
6
to
measure
complete
congener
distributions
in
17
Aroclor
lots
against
the
9
AccuStandard
calibration
mixtures3.
This
demonstrated
the
effectiveness
of
these
optimized
systems
and
the
new
calibration
mixes
in
expediting
CQCS
PCB
analysis.
The
results
revealed
the
pairs
of
Aroclor
1248
and
1254
lots
which
were
compared
differed
substantially.
This
highlights
a
weakness
in
CQCS
methods
which
propose
calibration
against
Aroclor
mixtures.
They
are
convenient,
and
represent
distributions
similar
to
those
often
observed,
but
they
are
only
as
good
as
the
tables
of
distribution
provided
for
them,
and
they
only
work
if
one
uses
the
same
or
a
similar
lot.
They
offer
little
help
if
"
non­
Aroclor"
congeners
are
encountered.
One
Aroclor
1254
lot
was
found
to
have
major
elevations
in
the
relative
amounts
of
non­
and
monoortho
chloro
substituted
congeners
when
compared
with
the
more
common
Aroclor
1254
lots.
This
has
been
found
to
represent
a
different
mode
of
manufacture
for
these
unusual
lots.
This
finding
has
important
implications
both
for
the
employment
of
such
lots
as
analytical
standards
and
for
studies
of
Aroclor
toxicity.

To
further
simplify
analytical
calibration
with
primary
standards
in
CQCS
PCB
analysis,
the
5
AccuStandard
"
Aroclor
congener
mixtures"
were
combined
and
then
serially
diluted
into
a
6
level
standard
curve
covering
a
100­
fold
concentration
range.
The
resultant
isomer
coelutions
in
the
standard
curve
when
Aroclors
were
profiled
on
a
GC­
MS­
SIM
system
designed
on
the
model
of
System
6
required
correction
for
the
combined
response
factors
present,
which
were
independently
measured
in
a
prior
experiment.
When
these
and
other
corrections
for
fragment
ion
contributions
from
coeluting
higher
homologs
(
higher
chlorination
level
PCBs)
in
both
the
samples
and
standards
were
applied
to
the
concentration
calculations,
results
for
Aroclor
peak
distributions
comparable
to
those
obtained
in
reference3
could
be
obtained
with
much
less
effort.
The
use
of
a
multilevel
inclusive
standard
curve
is
believed
to
improve
the
quantitation
more
the
uncertainties
arising
from
application
of
corrections
for
the
presence
of
isomer
or
homolog
coelutions
in
the
standard
mixture
and
in
the
samples.

SUMMARY
A
comprehensive
database
of
HRGC
PCB
retention
times
and
Aroclor
congener
distributions,
optimized
columns,
new
ECD
or
MS
detectors,
and
appropriately
designed
comprehensive
sets
of
primary
congener
standards
all
combine
to
greatly
reduce
the
effort
required
to
develop
and
validate
CQCS
PCB
analyses.
There
is
no
longer
an
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
129
excuse
for
neglecting
to
perform
such
analyses
when
studying
processes
which
give
rise
to
non­
standard
congener
distributions,
especially
those
that
give
rise
to
important
"
process
signature"
congeners
which
are
minor
or
missing
in
the
commercial
Aroclor
products.
Designing
a
PCB
analysis
to
accommodate
a
CQCS
measurement
of
all
resolvable
peaks
and
running
the
necessary
standards
to
support
eventual
quantitation
of
congeners
in
addition
to
those
specified
in
many
of
the
short
list
methods
can
increase
the
value
of
data
acquired
in
a
wide
range
of
environmental
studies.
This
should
be
recommended
to
researchers
who
cannot
predict
whether
the
levels
of
some
congener
initially
of
no
interest
may
not
later
on
become
critical
to
answering
a
new
question
or
resolving
an
as­
yet­
unidentified
problem.

Figure
3.
Comparison
of
PCB
Detectors'
Sensitivity
and
Linearity
REFERENCES
1.
Frame,
G.
M.,
Fresenius'
J.
Anal.
Chem.
1997
(
July),
in
Press
2.
Frame,
G.
M.,
Fresenius'
J.
Anal.
Chem.
1997
(
July),
in
Press
3.
Frame,
G.
M.,
Cochran,
J.
W.,
and
Bøwadt,
S.
S.
J.
High
Res.
Chromatogr.
1996,
19,
657­
668
4.
Larsen,
B.,
Cont,
L.,
Montanarella,
L.,
and
Platzner,
J.
J.
Chromatogr.
1995,
708,
115­
129
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
130
PROPOSED
US
EPA
METHOD
8320:
A
RISK
ASSESSMENT
METHOD
FOR
SECONDARY
EXPLOSIVES
Walter
Murray,
Yves
Tondeur,
Chris
Enterline,
Henry
Gruelich,
and
Jerry
Roach
Triangle
Labs,
Inc.,
801
Capitola
Drive,
Durham,
NC
27713
919­
544­
5729
Currently,
nitramines
and
nitroaromatics
are
measured
in
environmental
matrices,
such
as
groundwater
and
soil,
using
the
standard
US
EPA
Method
8330
(
SW­
846).
This
analytical
methodology
involves
a
sample
extraction
step
followed
by
HPLC
fractionation
using
a
uV
detector
for
quantification.
This
approach
is
useful
for
monitoring
a
group
of
74
secondary
explosives
and
associated
degradation
products
and
precursors
at
ppm
concentrations
in
samples
obtained
during
the
initial
characterization
phase
of
a
remediation
project,
as
well
as
for
manufacturing,
storage,
testing,
and
DOD
facilities.

What
we
are
proposing
is
not
an
alternate
but
complementary
method
for
analyzing
the
same
14
compounds
in
matrices
ranging
from
groundwater
to
soil/
sediment,
and
expanded
to
include
biological
tissues,
air,
and
ash/
residue.
The
detection
limits
of
this
new
methodology
­­
which,
once
approved,
will
be
given
the
name
of
EPA
Method
8320
­­
are
in
the
ppb
range.

Proposed
Method
8320
uses
isotope­
dilution,
in
which
isotopically­
labeled
compounds
are
introduced
into
the
sample
prior
to
extraction.
These
labeled
internal
standards
provide
extraction
efficiency
information,
recovery­
corrected
analyte
concentrations
and
qualitative
identification.
Isotope­
dilution
HPLC
coupled
with
negative­
ion
atmospheric
pressure
chemical
ionization
tandem
mass
spectrometry
(
MS/
MS)
of
explosives
offers
lower
detection
limits
than
ever
before.

The
validation
data
from
Method
8320
in
a
series
of
matrices
will
be
presented.
Also,
various
aspects
and
value/
benefits
of
the
new
methodology
will
be
discussed
in
terms
of
precision,
accuracy
and
reliability.
A
draft
version
of
the
method
was
submitted
to
the
EPA
during
summer
1996
for
consideration
and
inclusion
into
the
SW­
846
compendium
of
methods.
It
is
concluded
that
Method
8320
is
more
valuable
for
risk
assessment
studies,
with
reporting
levels
two­
to­
three
orders
of
magnitude
lower
than
for
Method
8330.

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
FAST
PRESCREENING
OF
WATER
AND
SOIL
SAMPLES
USING
SOLID­
PHASE
MICROEXTRACTION
(
SPME)

Tamra
L.
Schumacher
Lancaster
Laboratories,
2425
New
Holland
Pike,
Lancaster,
PA
17605­
2425
Phone
(
717)
656­
2300
EXT
1275.

The
objective
of
this
presentation
is
to
discuss
a
relatively
new
prescreening
approach
for
volatile
compounds.

Many
EPA
protocols
require
purge
and
trap
(
PT)
analysis
coupled
with
gas
chromatography
(
GC)
and
gas
chromatography­
mass
spectrometry
(
GC­
MS)
for
the
analysis
of
environmental
samples.
These
PT
techniques
can
be
costly
and
inefficient
to
run
due
to
problems
of
carry­
over
and
cross
contamination.
As
autosamplers
become
more
prevalent
in
PT
analyses,
the
inefficiencies
due
to
system
contamination
become
increasingly
problematic.

SPME
prescreening
methods
have
been
developed
to
determine
VOC
content
in
3
­
4
minutes
without
the
use
of
cryogens.
The
SPME
methods
which
have
been
optimized,
require
a
standard
gas
chromatograph
and
flame
ionization
detector
(
GC­
FID)
equipped
with
an
SPME
autosampler.
The
methods
developed
for
both
water
and
soil
matrices
use
SPME
headspace
sampling.
Problems
due
to
carry­
over
are
virtually
non­
existent.
Short
lengths
of
narrow
bore
capillary
columns
with
relatively
high
phase
ratios
yield
sufficient
chromatographic
resolution
and
loading
capacity
for
these
prescreening
applications.
The
use
of
FID
provides
stability
and
a
broad
linear
range.
The
system
is
easy
to
use,
maintenance
is
minimal,
and
system
troubleshooting
and
repairs
are
performed
easily
in­
house.
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
131
The
supporting
data
for
this
method
includes
comparisons
between
SPME
and
PT
data
for
a
number
of
water
and
soil
samples.
Technique
strengths
and
representative
standard
and
sample
chromatograrns
will
also
be
presented.

In
conclusion,
this
SPME
prescreening
technique
has
been
found
to
be
a
helpful
tool
in
processing
of
VOC
samples
in
a
time
efficient
manner.

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
ON­
SITE
ANALYSIS
OF
EXPLOSIVES
IN
SOIL:
EVALUATION
OF
THIN­
LAYER
CHROMATOGRAPHY
FOR
CONFIRMATION
OF
ANALYTE
IDENTITY
Sae­
Im
Nam,
Research
Chemist,
Daniel
C.
Leggett,
Research
Chemist,
and
Thomas
F.
Jenkins,
Research
Chemist
U.
S.
Army
Cold
Regions
Research
and
Engineering
Laboratory,
72
Lyme
Road,
Hanover,
New
Hampshire
03755­
1290
Martin
H.
Stutz,
Senior
Chemist
U.
S.
Army
Environmental
Center,
Aberdeen
Proving
Ground,
Maryland
21010­
5401
ABSTRACT
Two
colorimetric­
based
methods
are
commonly
used
for
on­
site
analysis
of
explosives
in
soil.
For
the
TNT
method,
acetone
soil
extracts
are
reacted
with
base
to
produce
reddish
Janowsky
anions.
For
the
RDX
method,
acetone
extracts
are
acidified
and
reacted
with
zinc
metal
to
reduce
RDX
to
nitrous
acid,
which
is
further
reacted
with
a
Griess
reagent
to
produce
a
reddish
product.
In
both
cases,
concentrations
are
estimated
using
absorbance
measurements
at
540
or
507
nm,
respectively.
The
limitations
on
positive
analyte
identification
with
these
procedures
are
that
the
TNT
method
also
reacts
with
other
polynitroaromatics,
such
as
TNB
and
DNT,
and
the
RDX
method
reacts
with
other
nitramines
(
HMX)
and
nitrate
esters
(
NG
and
PETN).
The
ability
to
qualitatively
differentiate
among
the
various
analyses
that
produce
positive
responses
would
greatly
enhance
the
usability
of
these
methods.

This
study
investigated
the
use
of
thin­
layer
chromatography
(
TLC)
as
a
simple,
on­
site
method
to
confirm
the
identity
of
analyses
detected
using
the
colorimetric
procedures.
Separations
using
both
laboratory­
grade
and
locally
available
solvents
were
developed.
The
combination
of
petroleum
ether:
isopropanol
(
4:
1)
provided
the
best
separation
for
the
nitroaromatics,
and
petroleum
ether:
acetone
(
1:
1)
produced
the
best
separation
for
the
nitramines
and
nitrate
esters.
Various
types
of
visualization
schemes
were
also
investigated.
The
most
sensitive
were
TiCl3
with
dimethylaminocinnamaldehyde
(
DMACA)
for
the
nitroaromatics,
and
the
Griess
reagent
with
UV
exposure
for
the
nitramines.
The
major
limitation
of
TLC
confirmation
analysis
is
that
it
does
not
currently
provide
an
analyte
detection
capability
comparable
to
the
colorimetric
tests.
Using
plates
with
a
preconcentration
zone
and
high
ratios
of
soil
to
solvent,
detection
levels
of
about
10
mg/
kg
seem
attainable.

INTRODUCTION
Environmental
concerns
over
explosives
contamination
in
soil
have
resulted
in
the
determination
of
the
extent
of
this
contamination
at
numerous
Department
of
Defense
installations.
Laboratory
analytical
methods
were
developed
to
enable
the
determination
of
the
most
commonly
found
components
of
explosives,
such
as
2,4,6­
trinitrotoluene
(
TNT),
hexahydro­
1,3,5­
trinitro­
1,3,5­
triazine
(
RDX),
and
related
impurities
and
environmental
transformation
products
in
the
soil
matrix
(
Jenkins
et
al.
1989,
U.
S.
EPA
Method
8330).
On­
site
methods
for
TNT
and
RDX,
the
most
commonly
encountered
contaminants
(
Walsh
et
al.
1993),
were
also
developed
to
provide
a
more
expedient
means
of
rapidly
characterizing
these
sites
prior
to
extensive
laboratory
analyses
(
Jenkins
and
Walsh
1992,
Teaney
and
Hudak
1994).
Overall,
the
use
of
on­
site
methods
has
been
successful
in
providing
rapid
site
characterization
at
explosives­
contaminated
sites.

The
most
commonly
used
pair
of
on­
site
methods
for
TNT
and
RDX
in
soil
is
based
on
research
conducted
at
the
U.
S.
Army
Cold
Regions
Research
and
Engineering
Laboratory
(
CRREL).
These
methods
are
based
on
the
production
of
colored
products
when
acetone
soil
extracts
are
reacted
with
the
appropriate
reagents.
In
the
field
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
132
screening
methods
by
Jenkins
(
1990)
and
Walsh
and
Jenkins
(
1991),
TNT
and
RDX,
respectively,
are
converted
to
color­
specific
compounds
that
are
quantified
spectrophotometrically.
Kits
containing
the
associated
reagents
and
supplies
are
commercially
available
from
EnSys
Corporation
(
now
Strategic
Diagnostics,
Inc.,
Newark,
Delaware).
In
the
TNT
method,
acetone
soil
extracts
are
reacted
with
strong
base
to
produce
reddish­
colored
Janowsky
anions
when
TNT
is
present.
Reddish­
colored
anions
are
also
produced,
however,
when
1,3,5­
trinitrobenzene
(
TNB)
or
N­
methyl­
N­
2,4,6­
tetranitrobenzenamine
(
tetryl)
is
present,
and
a
bluish­
colored
anion
is
produced
when
2,4­
dinitrotoluene
(
2,4­
DNT)
is
present
(
Jenkins
and
Walsh
1991).
Thus
a
positive
response
on
the
TNT
test
does
not
unequivocally
prove
that
TNT
is
present,
because
several
other
polynitroaromatics
can
give
a
similar
response.

For
the
RDX
test,
soil
extracts
are
first
acidified
with
acetic
acid
and
reacted
with
zinc
to
reduce
any
RDX
present
to
nitrous
acid,
and
the
resulting
solution
is
reacted
with
a
Griess
reagent
to
produce
a
reddish­
colored
azo
dye.
Reddish­
colored
azo
dyes
are
also
produced
if
other
nitramines,
such
as
octahydro­
1,3,5,7­
tetranitro­
1,3,5,7­
tetrazocine
(
HMX)
or
tetryl,
or
organonitrate
esters
such
as
nitroglycerin
(
NG),
or
pentaerythritol
tetranitrate
(
PETN)
are
present.
In
addition,
nitrate
and
nitrite
ion,
if
not
removed
using
an
anion
exchange
column
prior
to
reaction
with
zinc,
will
also
respond.
The
ion
exchanger
is
specified
in
the
CRREL­
developed
method,
but
is
not
recommended
for
routine
use
by
EnSys.

For
both
of
these
methods,
the
intensity
of
the
color
is
directly
proportional
to
the
concentration
of
the
analyte
of
interest,
and
concentrations
are
determined
by
measuring
absorbance
at
540
nm
for
TNT
and
at
507
nm
for
RDX.
Method
detection
limits
for
TNT
and
RDX
in
soil
samples
using
these
methods
are
1.1
mg/
kg
and
1.4
mg/
kg,
respectively.

Often
the
capability
of
the
TNT
test
to
detect
other
polynitroaromatics
can
be
quite
useful.
For
example,
in
a
recent
study
in
Sparks,
Nevada,
areas
contaminated
with
2,4­
DNT
were
detected
using
this
test
(
Jenkins
et
al.
1996).
Likewise
the
capability
of
the
RDX
test
to
determine
HMX
concentrations
was
recently
demonstrated
at
an
active
anti­
tank
range
at
Valcartier,
Quebec
(
Jenkins
et
al.
in
press).
It
is
important
to
be
able
to
discriminate
among
the
various
compounds
that
respond
to
these
tests
because
cleanup
levels
for
the
various
explosives
can
be
set
at
somewhat
different
concentrations.
Therefore
it
would
be
quite
useful
if
a
simple,
inexpensive,
on­
site
method
were
available
to
qualitatively
determine
which
of
the
potentially
detectable
analyses
are
giving
rise
to
the
colored
reaction
products
from
either
the
TNT
or
RDX
tests.

The
objective
of
this
work
is
to
evaluate
the
use
of
conventional
TLC
as
an
adjunct
to
current
on­
site
colorimetric
methods
for
TNT,
RDX,
and
related
compounds.
This
work
evaluates
the
ability
of
TLC
to
separate
TNT
and
the
other
common
polynitroaromatic
compounds,
and
RDX
and
the
other
commonly
encountered
nitramines
and
nitrate
esters
in
a
cost­
effective
and
timely
manner.

MATERIALS
AND
METHODS
TLC
materials
A
basic
thin­
layer
chromatography
starter
kit
was
purchased
from
Alltech
Associates,
Inc.
(
Deerfield,
Illinois).
The
starter
kit
included
a
20­
x
20­
cm
TLC
tank
with
glass
lid,
20­
x
20­
cm
tank
liners,
microcap
(
microcapillary)
dispensers
(
for
sample
spotting),
disposable
spray
box,
spotting
template,
reagent
spray
unit
with
glass
jar,
and
20­
x
20­
cm
Adsorbosil
Plus
1
TLC
plates.
Additional
glass­
backed
plates
consisting
of
EM
silica
gel
60
F254
(
20
x
20
cm,
250
µ
m),
EM
silica
gel
60
F254
with
preconcentration
zone
(
20
x
20
cm,
250
µ
m),
EM
silica
gel
60
with
preconcentration
zone
(
20
x
20
cm,
250
µ
m),
EM
HPTLC
silica
gel
60
F254
with
preconcentrated
and
prechanneled
zones
(
10
x
10
cm,
200
µ
m),
and
Adsorbosil
HPTLC
with
phosphor
and
preconcentrated
zone
(
10
x
10
cm,
150
µ
m)
were
purchased
from
EM
Science,
Gibbstown,
New
Jersey.
A
multiband
(
254
and
366
nm)
UV
lamp
with
a
viewing
box
was
obtained
from
UVP,
Inc.,
in
San
Gabriel,
California.

Standards
Analytical
standards
for
TNT,
TNB,
tetryl,
2,4­
DNT,
4­
A­
DNT,
2­
A­
DNT,
RDX,
HMX,
NG,
and
PETN
were
prepared
from
Standard
Analytical
Reference
Materials
(
SARM)
obtained
from
the
U.
S.
Army
Environmental
Center,
Aberdeen
Proving
Ground,
Maryland.

Solvents
Commercial­
grade
solvents
were
purchased
from
local
paint
and
hardware
stores.
These
solvents
consisted
of
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
133
Sunnyspec
Paint
Thinner,
Sterling
Lynsol,
Sterling
Thin­
X
(
red),
Sterling
VM
&
P
Naphtha,
Savogram
Deglosser,
Parks
VM
&
P
Naphtha,
Recorder
Paint
Thinner,
PPG
Industries
Duracryl,
Sterling
Acetone,
Ace
Paint
Thinner,
Woolworth
brand
70%
Isopropanol,
Sterling
Solvent
Alcohol,
and
3M
General
Purpose
Adhesive
Cleaner.

Visualization
A
number
of
different
visualization
methods
were
evaluated
as
a
part
of
this
work.
These
include
use
of
visual
fluorescence
while
irradiating
with
UV
light,
and
application
of
a
number
of
chemical
reagents
to
air­
dried
TLC
plates.
A
description
of
the
method
of
preparation
for
the
various
chemical
visualizing
reagents
is
presented
elsewhere
(
Nam
in
press).

General
thin­
layer
chromatography
procedures
The
developing
tank
was
prepared
by
the
addition
of
the
mobile
phase
(
200
mL)
and
equilibrated
for
approximately
30
to
40
minutes
or
until
the
tank
liner
had
been
saturated.
The
TLC
plates
were
prepared
for
sample
spotting
by
designating
the
line
of
origin
(
about
2­
3
cm
above
the
bottom
of
the
plate)
and
the
solvent
front
line
(
10
cm
from
the
line
of
origin).
Using
capillary
micropipettes,
samples
were
spotted
along
the
line
of
origin
approximately
1
cm
apart.
Spotting
volumes
ranged
from
0.5
to
30
µ
L,
but
in
most
cases,
were
spotted
1
µ
L
at
a
time.
The
plates
were
then
placed
in
the
developing
tank
(
containing
the
freshly
prepared
mobile
phase)
and
developed
in
an
ascending
manner
until
the
mobile
phase
had
reached
the
solvent
front
line.
The
plates
were
removed
from
the
tank
and
either
air­
dried
or
dried
with
hot
air
from
a
heat
gun
prior
to
observation.
The
fluorescence­
containing
plates
were
observed
under
the
UV
lamp
(
set
at
254
nm)
and/
or
sprayed
with
visualizing
agents.
Non­
fluorescence­
containing
plates
were
sprayed
with
visualizing
agents.
The
position
of
the
resulting
non­
fluorescing
spots
(
observed
under
the
UV
light)
or
colored
spots
were
marked
and
the
retention
factor
(
Rf)
values
were
determined
by
dividing
the
distance
traveled
by
the
compound
by
the
distance
traveled
by
the
solvent
front.
When
analyzing
soil
samples,
analyses
were
identified
by
comparing
the
Rf
values
to
the
Rf
values
of
the
standards,
which
were
spotted
on
the
same
plate.

RESULTS
AND
DISCUSSION
Separation
of
nitroaromatics
using
laboratory­
grade
and
commercial­
grade
solvents
Numerous
mobile
phase
systems
were
tested
to
determine
the
best
solvent
or
combination
of
solvents
that
would
result
in
a
distinguishable
separation
of
nitroaromatic
compounds,
such
as
TNT,
TNB,
DNT,
tetryl,
and
the
isomers
of
amino­
DNTs.
The
evaluated
mobile
phase
systems
using
laboratory­
grade
solvents
included
hexane:
chloroform
(
4:
1),
chloroform,
petroleum
ether:
acetone
(
3:
1
and
2:
1),
petroleum
ether:
isopropanol
(
4:
1),
xylene,
and
Stoddard
solution:
isopropanol
(
2:
1
and
1:
1).
These
solvents
were
chosen
according
to
their
eluting
strength.
The
combination
of
four
parts
petroleum
ether
with
one
part
isopropanol
resulted
in
the
best
distinguishable
separation
of
nitroaromatic
compounds
compared
to
the
other
evaluated
mobile
phase
systems
(
Table
1).

Commercial­
grade
paint
thinners
and
alcohols,
which
are
readily
available
in
any
local
hardware
or
paint
store,
were
also
evaluated
for
their
ability
to
separate
nitroaromatic
compounds.
However,
due
to
high
water
content
in
commercially
available
alcohols
(
i.
e.,
Woolworth
brand
70%
Isopropanol
and
Sterling
Solvent
Alcohol),
all
mobile
phases
prepared
with
these
alcohols
resulted
in
very
little
or
no
movement
of
the
nitroaromatic
compounds.
However,
when
commercially
available
paint
thinners
were
mixed
with
laboratory­
grade
isopropanol
(
4:
1),
compounds
such
as
TNT,
DNT,
TNB,
and
tetryl
could
be
distinguished
from
each
other
(
Table
2).
Commercial
brands
such
as
Recorder
Paint
Thinner,
Sterling
VM
&
P
Naphtha,
Sterling
Thin­
X
Paint
Thinner,
and
Ace
Paint
Thinner
were
among
the
tested
brands
that
have
resulted
in
good­
to­
fair
separation
of
TNT,
DNT,
TNB,
and
tetryl.
However,
in
most
cases,
the
development
time
took
longer
than
when
the
mobile
phase
was
composed
purely
of
laboratory­
grade
solvents.

Separation
of
nitramines
and
nitrate
esters
using
laboratory­
grade
and
commercial­
grade
solvents
The
solvent
system
of
petroleum
ether
and
acetone
(
1:
1)
was
found
to
be
very
effective
in
separating
the
nitramines,
such
as
RDX
and
HMX,
and
in
producing
a
fair
separation
of
the
nitrate
esters,
PETN
and
NG
(
Table
3).
RDX,
PETN,
and
NG
were
also
effectively
separated
with
petroleum
ether:
isopropanol
(
4:
1),
but
HMX
failed
to
move
from
the
line
of
orlgln.

Mobile
phase
solvents
consisting
of
commercial­
brand
acetone
and
some
paint
thinners
(
1:
1)
were
also
effective
in
producing
a
good
separation
of
RDX
and
HMX
(
Table
4).
However,
the
separation
of
PETN
and
NG
was
not
as
clear,
thus
resulting
in
very
similar
Rf
values.
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
134
Table
1.
Separation
of
nitroaromatics
with
laboratory­
grade
solvents.

0.14
0.23
0.30
0.27
±
0.02
0.55
0.62
0.16
0.22
0.30
0.23
±
0.02
0.55
0.62
0.1
0.20
0.27
0.33
0.37
±
0.02
0.62
0.67
0.28
0.48
0.40
0.45
0.52
±
0.02
0.63
0.65
0.47
0.36
0.40
0.46
0.58
±
0.02
0.71
0.72
0.50
0.52
0.42
0.47
0.64
±
0.02
0.75
0.74
0.54
1234567
4­
A­
DNT
Rf
±
S.
D.
2­
A­
DNT
Rf
±
S.
D.
Tetryl
Rf
±
S.
D.
2,4
­
DNT
Rf
±
S.
D.
TNB
Rf
±
S.
D.
TNT
Rf
±
S.
D.
Solvent
systems
Solvent
system
key:
1)
Chloroform
2)
Petroleum
ether:
Acetone
(
3:
1)
3)
Petroleum
ether:
Acetone
(
2:
1)
4)
Petroleum
ether:
Isopropanol
(
4:
1),
n
=
3
5)
Stoddard
solution:
Isopropanol
(
2:
1)
6)
Stoddard
solution:
Isopropanol
(
1:
1)
7)
Xylene
Table
2.
Separation
of
nitroaromatic
compounds
with
commercial­
brand
paint
thinners
and
laboratory­
grade
isopropanol
(
4:
1).

0.56
±
0.01
0.84
±
0.01
0.61
±
0.04
0.43
±
0.04
0.43
±
0
0.40
0.44
±
0
0.49
±
0.02
0.54
±
0.01
0.83
±
0.01
0.59
±
0.04
0.39
±
0.01
0.40
±
0.01
0.35
0.42
±
0.02
0.45
±
0.01
0.68
±
0.05
0.85
±
0
0.65
±
0.04
0.52
±
0.04
0.57
±
0.02
0.48
0.53
±
0
0.62
±
0.02
0.65
0.86
±
0.01
0.71
0.61
±
0.04
0.63
±
0.01
0.55
±
0
0.64
±
0.01
0.66
±
0.01
0.76
±
0.04
0.88
±
0.01
0.69
±
0.02
0.67
±
0.05
0.68
±
0
0.60
±
0
0.68
±
0
0.76
±
0.01
0.78
±
0.04
0.88
±
0.02
0.74
±
0.01
0.70
±
0.07
0.72
±
0.02
0.65
±
0.00
0.74
±
0.01
0.79
±
0.01
12345678
4­
A­
DNT
Rf
±
S.
D.
2,
A­
DNT
Rf
±
S.
D.
Tertyl
Rf
±
S.
D.
2,4­
DNT
Rf
±
S.
D.
TNB
Rf
±
S.
D.
TNT
Rf
±
S.
D.
Solvent
systems
Solvent
system
key:
1)
3M
Adhesive
cleaner:
Isopropanol,
n
=
2
2)
Duracryl:
Isopropanol,
n
=
2
3)
Deglosser:
Isopropanol,
n
=
2
4)
Recorder
paint
thinner:
Isopropanol,
n
=
2
5)
Parks
VM
&
P
naphtha:
Isopropanol,
n
=
2
6)
Sterling
VM
&
P
naphtha:
Isopropanol,
n
=
2
7)
Sterling
Thin­
X
paint
thinner:
Isopropanol,
n
=
2
8)
Ace
paint
thinner:
Isopropanol,
n
=
2
Table
3.
Separation
of
nitramines
and
nitrate
esters
with
laboratory­
grade
solvents.

0.79
±
0.02
0.84
±
0.03
No
movement
0.30
±
0.02
3
Petroleum
ether:
Isopropanol
(
4:
1)
0.88
±
0.02
0.91
±
0.03
0.62
±
0.01
0.72
±
0.01
3
Petroleum
ether:
Acetone
(
1:
1)
NG
Rf
±
S.
D.
PETN
Rf
±
S.
D.
HMX
Rf
±
S.
D.
RDX
Rf
±
S.
D.
n
Solvent
systems:
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
135
Table
4.
Separation
of
nitramines
and
nitrate
esters
with
commercial­
brand
solvents.

0.59
±
0.05
0.61
±
0.04
0.36
±
0.09
0.48
±
0.07
Ace
paint
thinner:
Sterling
acetone
(
1:
1)
0.60
±
0.03
0.59
±
0.03
0.43
±
0.08
0.53
±
0.03
Sterling
Thin­
X:
Sterling
acetone
(
1:
1)**
0.63
±
0.04
0.65
±
0.06
0.42
±
0.02
0.53
±
0.01
Sterling
VM
&
P
naphtha:
Sterling
acetone
(
1:
1)*
0.69
±
0.06
0.73
±
0.05
0.45
±
0.03
0.54
±
0.01
Parks
VM
&
P
naphtha:
Sterling
acetone
(
1:
1)*
0.78
0.80
±
0.05
0.53
±
0.04
0.58
±
0.04
3M
Adhesive
cleaner:
Sterling
acetone
(
1:
1)*
NG
Rf
±
S.
D.
PETN
Rf
±
S.
D.
HMX
Rf
±
S.
D.
RDX
Rf
±
S.
D.
Solvent
systems:

*
n=
2
**
n
=
3
Evaluation
of
TLC
plates
TLC
plates
used
in
this
study
were
all
glass
plates
precoated
with
silica
gel.
The
differing
features
of
these
plates
included
different
commercial
brands,
fluorescence
vs.
nonfluorescence,
and
preconcentration
zone
vs.
no
preconcentration
zone.
The
preconcentration
area,
located
on
the
bottom
of
the
plate,
is
made
up
of
inert
material
that
is
meant
to
absorb
sample
solvent.
The
various
commercial
fluorescent
and
non­
fluorescent
plates
were
found
to
produce
identical
separations.
However,
the
plates
having
preconcentration
zones
did
give
better
analyte
resolution
when
the
spotting
volume
exceeded
5
µ
L.
This
result
was
in
agreement
with
Rabel
and
Palmer
(
1992)
and
Hauck
and
Mack
(
1990),
who
report
enhanced
resolution,
reproducibility,
and
recovery
of
analyses
spotted
on
preconcentration
zones.

Evaluation
of
high­
performance
thin­
layer
chromatography
(
HPTLC)
plates
HPTLC
plates,
like
conventional
TLC
plates,
are
usually
coated
with
various
binders
to
hold
sorbent
material
together.
However,
the
dimensions
of
HPTLC
plates
are
approximately
half
the
size
of
conventional
plates.
The
particle
sizes
of
the
sorbent
material
are
much
smaller
and
the
size
distribution
of
these
particles
is
much
tighter.
HPTLC
plates
are
also
thinner
and
the
surface
is
more
uniform
than
conventional
plates.

HPTLC
plates
were
evaluated
here
to
determine
if
the
separation
and
resolution
were
better
than
with
conventional
TLC
plates.
Two
different
brands
of
HPTLC
plates
were
evaluated
(
EM
and
Adsorbosil).
Both
brands
had
preconcentration
zones,
with
the
EM
plates
also
having
channeled
zones
while
the
Adsorbosil
did
not.
The
compounds
were
spotted
along
the
preconcentration
zone
using
microcapillary
dispensers.
Due
to
the
size
of
the
plates
(
10
x
10
cm)
and
thinner
thickness
(
150­
200
µ
m),
the
developing
time
was
usually
between
10­
15
minutes,
half
the
development
time
of
standard
TLC
plates.
The
HPTLC
plates
also
required
less
mobile
phase
volume
compared
to
standard
TLC
plates.
However,
when
the
compounds
were
visualized
with
UV
light
or
with
visualizing
agents,
the
separation
and
resolution
of
compounds,
including
nitroaromatics,
nitramines,
and
nitrate
esters,
were
similar
to
the
standard
TLC
plates
whether
or
not
they
had
channeled
zones
(
Table
5).

Evaluation
of
visualizing
agents
The
simplest
method
for
visualizing
nitroaromatics
and
nitramines
was
viewing
the
developed
plate
under
shortwave
(
254
nm)
UV
light
(
Grover
and
Hoffsommer
1973,
Malotky
and
Downes
1983,
McCormick
et
al.
1978,
and
Zou
et
al.
1994).
We
found
the
fluorescence­
containing
plates
to
have
bright
green
backgrounds
with
light­
to­
dark
spots
representing
nitroaromatic
and
nitramine
compounds.
Following
the
UV
viewing,
nitroaromatics
could
be
further
distinguished
by
placing
approximately
1
µ
L
of
the
EnSys
TNT
developer
on
the
dark
spots.
The
EnSys
TNT
developer
reacts
with
the
nitroaromatic
compounds
to
form
Meisenheimer
complexes
and
results
in
the
following
color
formations:
purple
for
TNT,
orange
for
tetryl,
light
yellow
for
4­
A­
DNT,
orange
for
TNB,
light
yellow
for
2­
ADNT,
and
light
green
for
2,4­
DNT.
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
136
Table
5.
Comparison
of
HPTLC
plates
and
TLC
plates.

0.72
±
0.01
0.62
±
0.01
0.91
±
0.03
0.88
±
0.02
0.65
±
0.05
0.57
±
0.06
0.82
±
0.04
0.79
±
0.04
0.30
±
0.02
No
movement
0.84
±
0.03
0.79
±
0.02
0.32
±
0.02
No
movement
0.82
±
0.02
0.75
±
0.02
RDX
HMX
PETN
NG
0.64
±
0.02
0.58
±
0.02
0.52
±
0.02
0.37
±
0.02
0.23
±
0.02
0.27
±
0.02
0.83
±
0.03
0.79
±
0.02
0.73
±
0.03
0.68
±
0.01
0.57
±
0
0.62
±
0.02
TNT
TNB
2,4­
DNT
Tetryl
2­
A­
DNT
4­
A­
DNT
TLC
Rf*
±
S.
D.
HPTLC
Rf*
±
S.
D.
TLC
Rf*
±
S.
D.
HPTLC
Rf*
±
S.
D.
Compounds
Petroleum
ether:
Acetone
(
1:
1)
Petroleum
ether:
Isopropanol
(
4:
1)

*
n=
3
A
number
of
other
chemical­
based
visualization
procedures
were
evaluated
and
are
fully
discussed
elsewhere
(
Nam
in
press).
The
following
are
the
results
for
what
we
consider
to
be
the
best
candidate
visualization
systems.

TiCl3
and
DMACA
This
combination
worked
well
in
visualizing
nitroaromatic
compounds.
The
developed
plate
is
initially
sprayed
with
the
TiCl3
reagent,
which
reduces
the
nitroaromatic
compounds
to
amines.
When
the
plate
is
dried,
it
is
sprayed
with
DMACA
(
dimethylaminocinnamaldehyde),
which
reacts
with
the
amines
to
form
Schiff
bases.
Compounds
sprayed
with
DMACA
became
purple.
Color
development
was
immediate
and
was
best
viewed
when
the
plate
was
still
wet.

Griess
reagent
and
UV
exposure
Spraying
with
Griess
reagent
followed
by
UV
light
exposure
for
approximately
30
minutes
resulted
in
pink­
colored
spots
for
RDX,
HMX,
PETN,
and
NG.
However,
when
the
plates
were
dried
in
a
110
°
C
oven
for
20
minutes,
RDX
and
HMX
yielded
light
blue
spots
while
PETN
and
NG
had
lime
green
spots.

Estimation
of
detection
capability
A
preliminary
estimation
of
the
minimum
detectable
level
for
each
explosive
(
Table
6)
was
determined
by
spotting
different
volumes
(
ranging
from
0.5
to
30
µ
L)
of
each
standard
solution.
The
concentration
of
the
various
standard
solutions
ranged
from
5
to
2000
mg/
L.
The
HPTLC
plates
did
not
significantly
enhance
detectability
of
analysts.

Testing
of
soil
samples
collected
from
the
field
Soil
samples
collected
from
the
field
(
ranging
from
ammunition
plants
to
firing
ranges)
were
analyzed
using
the
conventional
TLC
methods
to
determine
or
confirm
the
accuracy
of
their
separation
procedures.
The
concentration
of
the
analyses
in
these
soil
samples
was
previously
determined
by
standard
HPLC
methods.
In
most
cases,
soil
samples
were
extracted
with
acetone
(
1:
5)
and
the
filtered
extracts
were
spotted
on
TLC
plates.
Soil
from
Umatilla
Army
Depot,
which
contained
716
µ
g
of
TNT
per
gram
of
soil,
was
extracted
and
spotted
(
10
x
1
µ
L).
This
yielded
a
visible
spot
under
UV
light
that
corresponded
to
the
same
Rf
value
as
the
spot
from
a
standard
solution
of
TNT.
A
soil
sample
from
Hawthorne
Ammunition
Plant
containing
an
array
of
explosives,
including
HMX
(
2.4
mg/
g),
RDX
(
8.1
mg/
g),
TNB
(
0.088
mg/
g),
DNB
(
0.002
mg/
g),
TNT
(
13.9
mg/
g),
and
2,4­
DNT
(
0.007
mg/
g),
was
extracted.
When
1
µ
L
of
the
extract
was
spotted
and
developed,
two
spots
were
visible
under
UV
light,
corresponding
to
TNT
and
RDX.
Soil
collected
from
Defence
Research
Establishment,
Valcartier
(
DREV)
had
tested
positive
for
RDX
using
the
RDX
that
area
was
HMX.
When
10
µ
L
of
the
DREV
soil
extract
was
spotted
and
developed
using
the
solvent
system
of
petroleum
ether:
acetone
(
1:
1),
a
single
pink
spot
appeared
after
spraying
the
plate
with
Griess
reagent
and
exposing
the
plate
under
UV
light
for
30
minutes.
The
Rf
value
corresponded
to
HMX,
and
the
identity
was
confirmed
by
standard
HPLC
methods.
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
137
Table
6.
Lowest
level
of
visualization.

2/
2
Griess
and
UV
exp.
0.4
NG
2/
2
Griess
and
UV
exp.
0.4
PETN
Nitrate
esters
3/
3
2/
2
HPTLC*/
UV
Griess
and
UV
exp.
0.1
0.2
HMX
4/
7
1/
1
8/
8
UV
Griess
and
UV
exp.
HPTLC*/
UV
0.1
RDX
Nitramines
4/
4
8/
8
3/
3
UV
TiCl3
and
DMACA
HPTLC*/
UV
0.1
TNB
1/
1
1/
1
1/
1
TiCl3
and
DMACA
EDA:
DMSO
(
1:
1)
HPTLC*
TiCl3
and
DMACA
0.01
0.05
Tetryl
4/
4
8/
8
4/
4
1/
1
UV
TiCl3
and
DMACA
HPTLC*/
UV
HPTLC*/
TiCl3
and
DMACA
0.1
TNT
Nitroaromatics
Frequency
Visualizing
agent
Lowest
level
(
µ
g)
Compound
*
High­
performance
thin­
layer
chromatography
plates
In
most
cases,
soil
samples
collected
from
the
field
contained
a
high
concentration
of
TNT
with
very
low
levels
of
other
nitroaromatics.
When
soil
extracts
from
these
samples
were
spotted,
due
to
the
high
TNT
concentration,
other
nitroaromatics
could
not
clearly
be
identified.
To
determine
the
effectiveness
of
the
TLC
method
in
separating
nitroaromatic
compounds
in
extracts
from
field
samples,
soil
from
Savanna
Army
Depot,
which
contained
similar
levels
of
TNT
(
14
µ
g/
g)
and
TNB
(
9.4
µ
g/
g)
were
used.
Soil
samples
were
extracted
in
acetone
as
described
above,
but
to
maximize
the
extract
concentration,
the
soilto­
solvent
ratio
was
increased
to
1:
2
(
1
gm
of
soil
to
2
mL
of
acetone).
A
volume
of
20
µ
L,
which
was
equivalent
to
0.14
µ
g
of
TNT
and
0.094
µ
g
of
TNB,
was
spotted
and
developed
in
the
solvent
system
of
Sterling
VM
&
P
Naphtha
and
laboratory­
grade
isopropanol
(
4:
1).
Plates
were
sprayed
with
TiCl3,
NaNO2,
and
Bratton­
Marshall
reagent
(
Jork
et
al.
1994).
Two
spots,
light
purple
in
color,
were
identified
as
TNT
and
TNB.
These
results
seemed
to
indicate
that
TLC
could
be
used
to
separate
and
distinguish
different
explosive
components
in
actual
field
samples
when
appropriate
solvent
system
and
visualizing
procedures
are
utilized.

Recommendations
for
specific
separations
and
visualizing
reagents
The
solvent
system
of
petroleum
ether
and
isopropanol
(
4:
1)
is
recommended
to
separate
various
species
of
nitroaromatic
compounds,
including
TNT,
TNB,
and
DNT.
The
most
sensitive
visualizing
agents
tested
for
nitroaromatics
are
TiCl3
followed
by
DMACA
spray.
For
the
separation
of
nitramines
such
as
RDX
and
HMX,
and
nitrate
esters
PETN
and
NG,
the
solvent
system
of
petroleum
ether
and
acetone
(
1:
1)
is
recommended
with
visualization
with
Griess
reagent
followed
by
UV
exposure.
Optimal
separation
occurs
with
laboratory­
grade
solvents;
however,
in
cases
where
laboratory­
grade
solvents
are
not
readily
available,
commercially
available
solvents
such
as
paint
thinner
and
acetone
may
be
substituted.
If
sensitivity
is
not
an
issue,
commercially
available
solvents
will
be
more
cost­
effective
and
more
readily
available
in
the
field.
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
138
SUMMARY
AND
CONCLUSIONS
For
the
purposes
of
this
evaluation,
the
conventional
TLC
approach
was
used
for
the
following
reasons:
1)
the
conventional
TLC
techniques
involve
minimal
equipment,
thus
making
them
field­
portable
and
inexpensive,
and
2)
the
purpose
of
this
report
was
to
evaluate
a
method
that
can
be
used
in
conjunction
with
on­
site
colorimetric
methods.
The
results
indicate
that
TLC
methods
could
indeed
be
used
to
separate
various
components
of
explosives
such
as
TNT,
TNB,
DNT,
RDX,
HMX,
PETN,
and
NG
from
soil
samples.
Using
appropriate
solvent
systems,
such
as
petroleum
ether
and
isopropanol
(
4:
1)
and
petroleum
ether
and
acetone
(
1:
1),
nitroaromatic,
nitramine,
and
nitrate
ester
compounds
could
be
effectively
separated.
In
most
cases,
commercial­
brand
solvents
(
readily
available
in
hardware
stores)
were
also
effective
in
giving
good­
to­
fair
separation
of
components
of
explosives.
However,
as
mentioned
at
the
onset
of
this
report,
the
detection
capability
of
conventional
TLC
methods
is
poor
and
remains
the
major
limitation
of
this
method.
The
conventional
TLC
method
evaluated
in
this
report
is
capable
of
detecting
0.1
µ
g
of
TNT
or
RDX
with
either
UV
light,
TiCl3
spray
followed
by
DMACA,
or
Griess
reagent
followed
by
UV
exposure.
This
is
equivalent
to
spotting
a
volume
of
1
µ
L
of
sample
extract
containing
100
µ
g/
mL
of
TNT
or
RDX,
and
if
the
sample
extract
was
prepared
using
the
soil­
to­
solvent
ratio
used
in
the
on­
site
colorimetric
methods
(
20
g
of
soil
to
100
mL
of
acetone),
the
concentration
of
TNT
or
RDX
in
soil
would
correspond
to
approximately
500
µ
g/
g.
This
is
about
500
times
above
the
minimum
detection
limit
for
TNT
(
1.1
µ
g/
g)
and
RDX
(
1.4
µ
g/
g)
colorimetric
on­
site
tests.
Even
if
the
maximum
spotting
volume
of
30
µ
L
is
used,
the
detection
capability
remains
at
about
17
µ
g/
g
if
the
soil­
to­
solvent
ratio
is
maintained
at
20
g
and
100
mL.
If
a
larger
soil­
to­
solvent
ratio
is
used
to
obtain
an
extract
for
TLC
analysis,
the
detection
capability
could
be
further
improved.
More
experiments
aimed
at
optimizing
detection
capability
by
either
concentrating
sample
extracts
and/
or
utilizing
higher
soil­
to­
solvent
ratios
are
needed
to
fully
assess
the
practical
limit
of
detection
for
the
TLC
method.

ACKNOWLEDGMENTS
Funding
for
this
research
was
provided
by
the
U.
S.
Army
Environmental
Center,
Aberdeen
Proving
Ground,
Maryland,
Martin
H.
Stutz,
Project
Monitor.
This
publication
reflects
the
personal
views
of
the
authors
and
does
not
suggest
or
reflect
the
policy,
practices,
programs,
or
doctrine
of
the
U.
S.
Army
or
Government
of
the
United
States.
The
authors
acknowledge
Marianne
Walsh
for
providing
useful
review
comments
on
this
manuscript.

REFERENCES
Glover,
D.
J.,
and
J.
C.
Hoffsommer
(
1973)
Thin­
layer
chromatographic
analysis
of
HMX
in
water.
Bulletin
of
Environmental
Contamination
&
Toxicology.
10:
302­
304.
Hauck,
H.
E.,
and
M.
Mack
(
1990)
Precoated
plates
with
concentrating
zones:
A
convenient
tool
for
thin
layer
chromatography.
LC­
GC,
8:
88­
96.
Jenkins,
T.
F.
(
1990)
Development
of
a
simplified
field
method
for
the
determination
of
TNT
in
soil.
USA
Cold
Regions
Research
and
Engineering
Laboratory,
Special
Report
90­
38,
Hanover,
New
Hampshire.
Jenkins,
T.
F.,
and
M.
E.
Walsh
(
1992)
Development
of
field
screening
methods
for
TNT,
2,4­
DNT,
and
RDX
in
soil.
Talanta,
39:
419­
428.
Jenkins,
T.
F.,
and
M.
E.
Walsh
(
1991)
Field
screening
method
for
2,4­
dinitrotoluene
in
soil.
USA
Cold
Regions
Research
and
Engineering
Laboratory,
Special
Report
91­
17,
Hanover,
New
Hampshire.
Jenkins,
T.
F.,
M.
E.
Walsh,
P.
W.
Schumacher,
P.
H.
Miyares,
C.
F.
Bauer,
and
C.
L.
Grant
(
1989)
Liquid
chromatographic
method
for
determination
of
extractable
nitroaromatic
and
nitramine
residues
in
soil.
Journal
of
AOAC,
72:
890­
899.
Jenkins,
T.
F.,
C.
L.
Grant,
G.
S.
Brar,
P.
G.
Thorne,
T.
A.
Ranney,
and
P.
W.
Schumacher
(
1996).
Assessment
of
sampling
error
associated
with
collection
and
analysis
of
soil
samples
at
explosives­
contaminated
sites.
USA
Cold
Regions
Research
and
Engineering
Laboratory,
Special
Report
96­
15,
Hanover,
New
Hampshire.
Jenkins,
T.
F.,
C.
L.
Grant,
P.
G.
Thorne,
M.
E.
Walsh,
S.
Thiboutot,
G.
Ampleman,
and
T.
A.
Ranney
(
in
press).
Assessment
of
sampling
error
associated
with
collection
and
analysis
of
soil
samples
at
a
firing
range
contaminated
with
HMX.
USA
Cold
Regions
Research
and
Engineering
Laboratory,
Special
Report,
Hanover,
New
Hampshire.
Jork,
H.,
W.
Funk,
W.
Fischer,
and
H.
Wimmer
(
1994)
Thin­
Layer
Chromatography:
Reagents
and
Detection
Methods
 
Physical
and
Chemical
Detection
Methods:
Activation
Reactrons,
Reagent
Sequences,
Reagents
Il.
Volume
l
b.
Weinheim.
VCH.
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
139
Malotky,
L.,
and
S.
Downes
(
1983)
Explosive
analysis
kit.
Proceedings,
1st
Symposium
on
Analysis
and
Detection
of
Explosives,
FBI
Academy,
Quantico,
Virginia,
pp.
63­
65.
McCormick,
N.
G.,
J.
H.
Cornell,
and
A.
M.
Kaplan
(
1978)
Identification
of
biotransformation
products
from
2,4­
dinitrotoluene.
Applied
and
Environmental
Microbiology,
35:
945­
948.
Nam,
S.
I.
(
in
press)
On­
site
analysis
of
explosives
in
soil:
Evaluation
of
thin­
layer
chromatography
for
confirmation
of
analyte
identity.
USA
Cold
Regions
Research
and
Engineering
Laboratory,
Special
Report,
Hanover,
New
Hampshire.
Rabel,
F.,
and
K.
Palmer
(
1992)
Advantages
of
using
thin­
layer
plates
with
concentration
zones.
American
Laboratory,
92:
20BB.
Teaney,
G.
J.,
and
T.
T.
Hudak
(
1994)
Development
of
an
enzyme
immunoassay­
based
field
screening
system
for
the
detection
of
RDX
in
soil
and
water.
Proceedings
of
the
87th
Annual
Meeting
and
Exhibition,
Air
&
Waste
Management
Association,
Cincinnati,
Ohio,
94­
RP143.05,
15
pp.
U.
S.
EPA
Method
8330
(
1995)
Nitroaromatics
and
nitramines
by
HPLC.
Second
Update
SWS46
Method
8330.
Walsh,
M.
E.,
and
T.
F.
Jenkins
(
1991)
Development
of
a
field
screening
method
for
RDX
in
soil.
USA
Cold
Regions
Research
and
Engineering
Laboratory,
Special
Report
91­
7,
Hanover,
New
Hampshire.
Walsh,
M.
E.,
T.
F.
Jenkins,
P.
S.
Schnitker,
J.
W.
Elwell,
and
M.
H.
Stutz
(
1993).
Evaluation
of
SW846
Method
8330
for
characterization
of
sites
contaminated
with
residues
of
high
explosives.
USA
Cold
Regions
Research
and
Engineering
Laboratory,
CRREL
Report
93­
5.
Zou,
H.,
S.
Zhou,
X.
Hu,
Y.
Zhang,
and
P.
Lu
(
1994)
Identification
of
2,4,6­
trinitrotoluene
and
its
biodegradation
products
by
thin­
layer
chromatography.
Journal
of
Planar
Chromatography
Modern
TLC,
7:
461­
463.
WTQA
'
97
­
13th
Annual
Waste
Testing
&
Quality
Assurance
Symposium
140
