THE DETERMINATION OF SYNTHETIC ORGANIC COMPOUNDS IN WATER
BY PURGE AND SEQUENTIAL TRAPPING CAPILLARY COLUMN GAS CHROMATOGRAPHY
by
Thomas A. Bellar
and
James J. Lichtenberg
Physical and Chemical Methods Branch
Environmental Monitoring and Support Laboratory
U. S. Environmental Protection Agency
Cincinnati, Ohio 45268
ENVIRONMENTAL MONITORING AND SUPPORT LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U. S. ENVIRONMENTAL PROTECTION AGENCY
CINCINATI, OHIO 45268
January 1985
-------
THE DETERMINATION OF SYNTHETIC ORGANIC COMPOUNDS IN WATER
BY PURGE AND SEQUENTIAL TRAPPING CAPILLARY COLUMN GAS CHROMATOGRAPHY
Thomas A. Bellar and James J.LIchtenberg
INTRODUCTION
Based upon the limited data obtained from early purge and trap methods
development reports, it was generally concluded that a single programmed
temperature packed column could elute all of the compounds efficiently
extracted by common purge and trap operations. As a result, several genera-
tions of purge and trap instruments were developed that were designed to
operate solely with highly efficient packed column gas chromatographs. As
purge and trap methods evolved with such instrumentation and as GC/MS survey
data were evaluated, it became apparent that the limiting factor for a broad
spectrum purge and trap analysis is not the extraction step but the in-
ability of a single packed column to resolve and elute all of the commonly
occurring synthetic organic compounds extracted from water by inert gas
purging. This observation is substantiated by current Agency methodologies
where a common set of extraction and trapping parameters are used for
several different methods, the primary difference being the utilization of a
variety of specific detectors, packed columns with unique resolving powers
or complex temperature programs. Only through the use of all of these tools
do packed columns provide the qualities for sensitive, accurate and precise
methods for a wide variety of compounds.
-1-
-------
Capillary columns have long been used to resolve complex mixtures of
apolar compounds over an extremely wide boiling range. Moreover, with the
development of chemically Inert glass and fused silica columns, 1t 1s now
possible to simultaneously analyze extracts containing compounds of widely
varying polarities and coexisting organic acids and bases. Inert capillary
columns and gas chromatographs, primarily designed for capillary columns,
are now commercially available and are common to many laboratories. The
unique properties of capillary systems and their acceptance by commercial
laboratories suggest that their application to purge and trap methodology is
practical and could significantly improve the quality of purge and trap
data. Their utilization could result in a single method capable of resolv-
ing complex mixtures of reactive compounds over a wide boiling range well
beyond the capabilities of a single packed column.
The experimental design of this study had three main objectives: the
first, to develop and document a thorough understanding of the mechanics and
limitations of purge and trap capillary column gas chromatography; secondly,
to develop a simple automated analytical approach for the analysis of a
number of synthetic organic compounds, and finally, to determine single
laboratory method detection limits, accuracy, precision and sample stability
data in order to determine if the approach can be used as a method for the
analysis of diverse organic chemicals in drinking water related samples.
EXPERIMENTAL
Almost every variable encountered in purge and trap operations has a
direct effect upon the accuracy and precision of the method. For this
-2-
-------
reason, many of the basic purge and trap parameters which have been palns-
takenly optimized over the years for packed column operations were utilized,
whenever possible. For capillary column gas chromatography, the purging and
trapping functions did not require modification. However, significant
modifications were required for the sample desorption injection step before
acceptable capillary column performance could be obtained.
For a packed column system during desorption, an inert gas flowing at a
rate between 20 and 40 mL/min backflushes the trap for approximately four
minutes while the trap is flash heated tn 180°C. The trapped components
are released from the sorbent as the temperature is elevated and are
transferred into the packed column by the inert gas. Low boiling apolar
compounds leave the trap as a sharp spike while higher boiling compounds
elute as broad tailing peaks. For packed columns with internal diameters
larger than 2 mm and theoretical plate values of less than 1500 plates/m,
compounds can be injected under isothermal conditions contained in 10-15 ml
of gas without adversely affecting the performance of the column. At a flow
rate of 30 mL/min low boiling materials exit the trap contained in
approximately 15 mL of gas. The result is acceptable peak geometries for
desorption injections of compounds even when the column is operated at
temperatures at which the low boiling compounds are mobile (i.e., chromato-
graphic separations begin at injection). Higher boiling compounds such as
aromatic hydrocarbons elute from the trap over longer periods of time, e.g.
120 seconds, and, therefore, are presented to the column in a volume of gas
approaching 60 mL. If the chromatographic column temperature is high enough
for such compounds to be mobile at injection, the desorption profile
-3-
-------
projects through the chroma to graphic column resulting 1n poor chromato-
graphlc peak geometries and a loss of resolution. Packed column purge and
trap methods avoid this problem through temperature programming. An Initial
column temperature 1s selected so that the low boiling, Ideally injected,
sample components are mobile and allowed to separate as they pass through
the column. At this relatively low column temperature, the higher boiling
compounds are immobile and remain trapped on the first few cm of the column
packing during desorption. Subsequently, as the temperature of the column
1s raised through temperature programming, the higher boiling components
become mobile and elute from the column as well defined peaks.
In direct contrast, for proper capillary column operations the sample
must be injected into a capillary column contained in a microvolume of gas.
The internal diameter, linear gas flow, and film thickness of the capillary
column all have a direct limit on the maximum volume of desorb gas in which
the analyte can be contained before it has an adverse effect upon the per-
formance of the capillary column. For current, commercially available glass
capillaries, this volume of gas varies from about 50 to 500 uL. It is
evident, therefore, that simply attaching a purge and trap unit designed for
packed column operations to a capillary column will result in poor quality
gas chroma to grams.
This limitation has been resolved by two differing desorption/injection
approaches. These approaches are commonly referred to as "cryofocusing" and
"sequential trapping." Cryofocusing is a condition where the desorbed
compounds are cold trapped in the analytical column or in a pre-capillary
-4-
-------
column at a temperature between 100'C and 150*C below the normal elutlon
temperature of those compounds. Under this condition, the low and high
boning compounds are Immobile and are contained 1n a very short section of
the capillary column. After 100 percent transfer, the cooled area 1s heated
and the compounds are released and separated by the column. Sequential
trapping is a procedure where the trap normally used for packed column
operations ("A" trap) is desorbed into a second microbore trap ("B" trap).
The "B" trap 1s in turn backflushed and desorbed into the capillary column
operated at temperatures where the ideally desorbed compounds are mobile and
the non-ideally desorbed compounds are cold trapped. Through temperature
programming all of the desorbed compounds elute as ideal peaks.
The primary advantages of cryofocusing are:
1. The modifications to existing purge and trap packed systems are
generally inexpensive and within the technical abilities of most
laboratories.
2. Almost all commercially available capillary columns can be used since
the volume of gas required for quantitative transfer does not affect the
peak geometries of cold trapped compounds.
Some of the disadvantages or precautions one should consider for such an
approach include:
1. The need to use liquid nitrogen or liquid carbon dioxide to cool the
capillary column down to cryofocusing temperatures.
2. The possibility of ice crystals forming within the trapping region of
the column resulting in restriction or total blockage of flow and
unpredictable, non-quantitative sample transfer.
-5-
-------
3. If the sample forms an aerosol as 1t 1s desorbed from the trap, H will
not be effectively trapped 1n an open tubular column.
4. Ideally, the stationary phase should still be a liquid at the reconsti-
tution temperature.
5. Variable retention data oftentimes result when cryogenic operations are
performed due to the adverse effect they have upon the oven temperature
and flow controllers.
The advantages of sequential trapping are:
1. Once the operational parameters are optimized the unit can be automated
to perform like current packed column purge and trap instruments
resulting in qualitative and quantitative data with outstanding accuracy
and precision.
2. Large volumes of coolants are not required, operational expenses are
lower and the unit can be set in remote locations for unattended
operation.
The disadvantages are:
1. The sequential trapping operation must be carefully optimized to
transfer and reconstitute all of the compounds of interest. It may not
be possible to include both extremely volatile compounds and high
boilers in a single analysis.
2. Microbore traps are difficult to prepare.
-6-
-------
3. For a broad spectrum analysis sequential trapping operations are only
possible using thick-film, widebore capillary columns (0.5 to 0.75 mm
Internal diameters).
4. New purge and trap equipment specifically designed for sequential trap-
ping must be purchased because extensive modifications are required to
update most packed column purge and trap systems.
For this study the sequential trapping system was selected to perform
all of the experiments, because the equipment was commercially available and
the approach appeared to be the greatest challenge.
EQUIPMENT USED
A Chemical Data Systems Model 320 (CDS-320) concentrator with the capil-
lary option was used for the purge and trap operations. The purge and trap
unit was attached to a Hewlett Packard Model 5730A gas chromatograph equip-
ped with a C02 subambient column accessory and flame ionization detectors
(FID). All of the retention, peak width and area data were gathered with a
Hewlett Packard 3388A integrator. Three Supelco glass capillary columns
were used: Grade AA SE-30-60 m long, 0.75 mm ID with a film thickness (df)
of 1.0 um and a reported coating efficiency of 115%; Grade AA, SE-30
Bonded—60m long, 0.75 mm ID with a df of 1 um and a coating efficiency of
101%; Grade AA SE-54-30 m long, 0.50 mm ID with an unknown film thickness
and unknown coating efficiency. A 50m, 0.50 mm ID Superox ALL column from
All tech and an Analabs 50m, 0.50 mm ID SE-30 column were
-7-
-------
also briefly evaluated. Information on the coating efficiency and film
thickness was not available for the latter two columns. A Supelco direct
Injection capillary column Inlet conversion kit was used to evaluate column
performance and system activity.
SYSTEM EVALUATION
The initial evaluation of the assembled analytical system was designed
to determine the performance of the capillary column gas chromatograph, the
transfer line between the column and the purge and trap unit, and the
sequential trapping and desorption operations. The purpose of the evalua-
tion was to determine what types of compounds can be handled quantitatively
by the entire system without regard for whether or not they can be purged
from water. For this evaluation, a neutral reactives test mixture (test
mixture) supplied with the Supelco SE-30 capillary column was used. The
column manufacturer also supplied a chromatogram of the mixture generated by
the column used for this study operated under optimum conditions. The
supplied chromatogram was considered to be a "primary chromatogram" and each
system component was systematically optimized whenever possible to ulti-
mately generate a chromatogram of similar quality.
The ends of the 60m X 0.75mm glass column coated with SE-30, 1.0 wm df
were straightened, deactivated and installed in the FID gas chromatograph.
Prior to attaching the purge and trap unit to the gas chromatograph, proper
column installation was confirmed by on-column injections of the test
mixture. For the initial evaluation the widebore capillary column direct
-8-
-------
Injection conversion kit was Installed 1n the gas chroma tograph to allow
direct volatilization Injections of liquid and gaseous samples Into the
capillary column. Using helium as a carrier gas, the linear gas flow
through the column was adjusted to 20 cm/second at 115*C. Helium was used
as the make-up gas to increase the total flow into the FID to 40 mL/minute.
Volumes between 0.25 and 1.0 uL of the neat test mixture were injected into
the glass capillary column. The peak geometries and relative peak heights
of the resulting chroma to gram closely duplicated the primary chroma to gram.
The concentation of each analyte in the test mixture was such that a
properly operated FID would generate nearly equal response to each analyte
if the entire system is equally inert to each of the test components. The
installation test recommended by the column manufacturer compares the peak
heights of the reactive components to those of the non-reactive components.
To compensate for peak width changes obtained during isothermal operations,
a continuous curve is drawn connecting the apex of each of the non-reactive
peaks (n-alkanes) in the resulting chromatogram. The percent response of
each reactive analyte is calculated by dividing the theoretical height of
the reactive compound by the observed peak height and multiplying by 100.
This test assumes that the n-alkanes generate ideal peaks and, therefore,
becomes a means of monitoring losses of reactive compounds and peak tailing
effects. Percent response values for the methyl silicone column in excess
of 70% are considered "good" by the column manufacturer and are representa-
tive of an inert column and proper installation. The calculated results of
the initial column installation evaluations and those supplied by the column
manufacturer (primary chromatogram) are shown in Table 1.
-9-
-------
The peak geometry for- each of the n-alkanes was sharp and symmetrical.
The calculated response values for each of the reactive analytes (Instal-
lation chroma to gram) was found to be 1n excess of 70 percent and comparable
to those obtained by the column manufacturer indicating proper chro-
matographic system performance. It should be noted that the primary
chromatogram was generated using a splitting injector operating with a split
ratio of 50:1. Through this simple test it was also shown that contrary to
narrow-bore capillary column operations, volumes between 0.25 and 1.0 »L of
the test mixture can be directly injected into widebore capillary columns at
temperatures significantly above the boiling point of the solvent without
adversely affecting the performance of the column.
The purge and trap unit was then attached to the gas chromatograph
exactly according to the operators manual. The heated metal transfer line
was attached directly to the 0.75 mm capillary column bypassing the capil-
lary injector. The linear velocity through the capillary column was
adjusted to 20 cm/second at 115*C using the mass flow controller supplied
with the purge and trap unit. The design of the COS-320 purge and trap unit
allows liquid injections to be made into the unit at two points through
heated injectors so that each of the trap/desorption functions can be moni-
tored. Volumes of the test mixture between 0.25 and 1.0 uL were injected
into the COS-320 column injector to evaluate the performance of the heated
metal transfer line and the CDS-320 column injector, both operated at
200*C. The resulting chromatogram was of poor quality. The n-alkanes
tailed indicating the possibility of excessive internal volumes, cold spots,
or reactive surfaces between the injector and the column. The percent
-10-
-------
response values for the reactive compounds calculated from the chromatogram
are listed 1n Table 1 under COS transfer line chromatogram. The appearance
of the chromatogram and the calculated data show that the Injector and/or
the transfer line are detrimental for Isothermal capillary column analyses
of compounds of similar polarities and boiling range. In an effort to
resolve this problem, the transfer line was modified by threading a section
of 0.32 inn ID fused silica capillary column coated with OV-1 through the
transfer line. One end of the fused silica line was attached directly to the
CDS-320 column injector using a zero dead volume reducing fitting and
graphite ferrules and the other end was connected directly to the glass
capillary using a capillary column butt-end connector. The test was again
repeated and the resulting chromatogram generated sharp, symmetrical peaks
for the alkanes and, with the possible exception of the alcohol, also for
the reactive analytes. The calculated response values for the test mixture
using the modified transfer line also appear in Table 1. The reactivity of
the system to alcohol (recovery 67 percent) appears to be due to the active
sites located within the injector. All of the other test compounds provided
recoveries and peak geometries nearly identical to the primary and instal-
lation chroma to grams. For an initial evaluation of the assembled purge and
trap-capillary column system, the test solution was injected into the "A"
trap through the trap injector. With the exception of the modified transfer
line, the unit was operated as received using the sequential purge and trap
conditions recommended in the CDS operators manual. The resulting chromato-
gram provided unusually wide symmetrical peaks for the n-alkanes indicating
that the volume of desorb gas containing the analytes was excessive,
adversely affecting the performance of the capillary column. The reactive
-11-
-------
analytes were present but low 1n yield and the late elutlng compounds
appeared as doublets 1n the chroma togram. The percent response values
appear 1n Table 1 under sequential trapping chromatogram and are, at best,
estimates because of the poor quality of the chromatogram. Attempts to
improve peak geometry through simple desorption parameter modifications did
not improve the quality of the chroma to grams. Based upon observations of
these and other chroma to grams, it was apparent that extensive modifications
to the purge and trap system and optimization of the various parameters
would be required before reliable multi-residue quantitative analyses could
be performed upon water samples. Progressive experiments were then designed
to determine acceptable trap internal diameters, the boiling range of
neutral compounds that can be desorbed into an isothermally operated wide-
bore capillary column, the sorbents and conditions best suited for capillary
column purge and trap operations, the parameters required for quantitative
sequential trap operations and the selection of a capillary column and the
temperature program best suited for the analysis of purgeable compounds.
ANALYTE BOILING RANGE AND TRAP INTERNAL DIAMETERS
From the sequential trapping chromatogram it was evident that at least
one major problem was occurring. The analytes selected for this study were
contained in an excessive volume of desorb gas for proper injection into the
isothermally operated capillary column. Based upon packed column experi-
ences, it was assumed that the boiling points of the test analytes were too
high, the internal volume of the trap was too large, or the trap sorbent
-12-
-------
could not be heated fast enough to generate sharp, symmetrical desorptlon
peaks.
To evaluate these possibilities, the following series of experiments was
performed: 150 uL injections of a gaseous standard solution containing
1.0 uL of n-pentane, n-hexane, n-heptane, n-octane, n-nonane, and n-decane/L
of air (n-alkane mix) were injected into the trap injector on the COS-320.
The injected sample was swept into the "A" trap (23*C) for 11 minutes with
helium flowing at 40 ml/minute. Trap "A" was then heated to 180*C and
backflushed into the "8° trap for 2 minutes with helium flowing at 20
mL/minute. The "B" trap was then backflushed at 180*C into the analytical
column with helium flowing at 10 mL/minute for 120 seconds. The "A" trap,
common to most EPA purge and trap methods, contained only Tenax, was 23 cm
long and had an internal diameter of 2.67 nro. Three different "B" traps
were evaluated: the trap supplied with the unit (a 2.667 mm I.D. stainless
steel tube containing 23 cm of Tenax) and two traps fabricated in the
laboratory (a 1.651 mm ID copper tube containing 23 on of Tenax, and a 1.8
mm I.D. glass lined stainless steel tube containing 23 cm of Tenax). The
gas chroma to graphic column was maintained at 70*C for 8 minutes, then
programmed at 8*/min to 100*C. Chromatographic column conditions were
selected so that most of the compounds would elute under the 70*C isothermal
conditions while n-decane, a compound already shown to have adversely
affected isothermal peak geometries, would elute as the column is programmed.
The purpose of the column program was to determine if mild temperature
programming would improve the peak geometry of dec ane. Triplicate analyses
were performed using each set of traps and the results were compared
-13-
-------
to direct Injections of the n-alkane mix Into the column through the CDS 320
column Injector. Retention data, peak area, peak width at half height
(Intergrator value) and peak height 1n mm (hand measurement) were recorded
and averaged. Table 2 lists the resulting data.
With the exception of n-nonane, the retention data from the trapped
injections are uniform and differ from direct injections by about 24 * 2
seconds. For some unknown reason, n-nonane leaves the trap before the other
analytes and differs by 12 seconds. With the equipment used the internal
vo'iumes are s.ra11, the linear gas velocities are high and all surfaces are
heated; therefore, the difference in retention times between the trapped
materials and direct injection are primarily due to the time it takes to
heat the Tenax sufficiently to release the compounds to the backflush flow.
!t is interesting to note that the thermal conductivity of the trap tubing,
copper vs. glass-lined stainless steel, did not influence the retention data.
Peak area comparisons between the on-column injections and the various
traps in Table 2 show that, with the exception of n-pentane, nearly identi-
cal areas were obtained for each trap and these, in turn, compare favorably
to direct injection areas.
Quantitative values for pentane were not obtained because under the
simulated purging conditions (11 minutes at 40 mL/minute) the retention
volume of pentane for trap "A" was exceeded, resulting in partial venting.
-14-
-------
Peak height and peak width at half height comparisons show that as the
Internal diameter of the "8" trap decreases, the peak geometries of the
peaks elutlng within the Isothermal area of the chromatogram more closely
approximates those of direct injection. These peak data also show that
decane, elating under programmed conditions, does not exhibit peak broaden-
ing effects from desorption indicating that a minimal column temperature
change (20-30*C) will sharpen the peak geometries of non-polar compounds.
The visual appearance of the chromatograms indicate that for isothermal
column gas chromotography the two narrow bore traps (< 2 mm ID) generate
acceptable (but not i Jeal) chromatograms while the wi deb ore trap (2.7 mm ID)
adversely affects the performance of the capillary column to the point where
closely eluting peaks may fuse.
To further determine what effects the "B" trap and chroma to graphic
conditions may have upon the quality of capillary column chromatogram of
polar compounds and higher boiling alkanes, additional test mixture injec-
tions were performed. Triplicate 1 uL aliquots were injected into the
CDS-320 column injector and into the "A" trap injector. Trap "A" in each
case was a standard 23 cm X 0.105" Tenax trap. The three previously
described "B" traps were further evaluated. The injected materials were
flushed into trap A, sequentially trapped on trap B and desorbed to the
column according to the previous experiment. The desorbed compounds were
separated isothermal!y at 115*C. The experiments were then repeated where
-15-
-------
the sample upon desorptlon from the "B" trap was reconstituted on the capil-
lary column under true cold trapping conditions. The capillary column was
programmed as follows: during desorptlon, the column was maintained at 20*C
for 2 minutes followed by a 32*/minute program (maximum rate) to 115*C. The
column was then maintained at 115*C until all of the compounds eluted. All
of the compounds eluted during the 115*C isothermal conditions.
The percent response values defined previously were calculated for each
of the reactive analytes, the area of each peak relative to n-decane was
determined and the number of theoretical plates per meter for n-tridecane
was calculated. The isothermal data (non-cold trapping) appear in Table 3
and the programmed (cold trapping) data appear in Table 4.
As noted in the primary chromatogram (Table 1), where similar injections
were performed, the quality of the isothermal chromatograms used to generate
the data in Table 3 are, at best, poor for the hydrocarbons and totally
unacceptable for the reactive compounds. Comparison of the n-decane ratio
data between direct injection and the various traps indicate that the purge
and trap operations are quantitative for the alkanes while there appear to
be losses for the reactives. It is interesting to note that the 2.7 mm trap
caused the retention time for 2,4-dimethylphenol to increase fusing it with
n-undecane. As in the case of the previous experiment with normal alkanes,
as the internal diameter of the trap decreases, the quality of the resulting
chroma togram improves. This is further emphasized by comparing the number
of theoretical plates per meter for n-tridecane.
-16-
-------
In direct contrast, Table 4 comparisons of the number of theoretical
plates obtained from the column using the various traps shows that, under
cold trapping conditions, the normal alkanes elute as Ideal peaks showing no
adverse affects from the sequential trapping operations or the variation in
traps. Comparing the area ratio data for direct injection to the various
traps shows that relative to the n-decane there is quantitative transfer of
the higher boiling alkanes and naphthalene. For each of the traps there was
a slight tailing for the alcohol peak resulting in response determinations
of 70% or less. Although peak geometries for the 2,6-dimethylphenol were
acceptable, there is a slight loss (-10 percent) relative to n-decane.
Other reactive compounds appear to be quantitatively transferred.
SORBENT SELECTION
Early packed column purge and trap methods development investigations
evaluated a large number of potential trap s or bents. From these studies,
traps packed with Tenax or combinations of Tenax, silica gel and activated
carbon were selected as best suited for the analysis of purgeable priority
pollutants by packed column purge and trap operations. Since these studies
a number of potential sorbents with unique properties have been developed.
The previously developed sorbent traps and a few potential sorbents were
evaluated to determine their applicability to capillary column multi-residue
purge and trap operations. The following sorbents were evaluated: Tenax
GC, Silica gel, Ambersorb XE-340, Molecular seive ELZ-115, and several
experimental Carbosieve-like products supplied by Supelco. For this study,
stainless steel 23 cm X 2.7 mm ID "A" traps were packed with 100 percent
-17-
-------
Tenax and 50 percent Tenax (Inlet) followed by 50 percent of each of the
above mentioned test materials. The "B" traps used for this study were 23
on x 1.8 mm 10 glass-lined stainless steel packed with the sane sorbents as
the "A" traps. The following conditions were selected: a 5 ml aqueous
solution was purged at 22-25*C with helium flowing at 40 mL/minute for 11
minutes, 150 uL of gaseous injections were made directly into trap "A."
During the purge cycle or injection, the "A" trap was maintained at room
temperature (25~30*C). For the "A" trap to "B" trap transfer, the "A" trap
was heated to 180*C and backflushed to the "B" trap with helium flowing at
20 "I/minute for 120 seconds. The "B" trap was at room temperature. For
desorption to the capillary column the "B" trap was rapidly heated to 180*C
while being backflushed with helium flowing at 10 ml/minute for 120 seconds.
A 60m 0.75 mm ID glass capillary column coated with SE-30 1 ym df was
selected as the analytical column.
The results of the investigations show that using common purge and trap
conditions, sorbents other than Tenax retained too much water during the
purge operations at 22-25*C for capillary column analyses. When the
retained water was desorbed into the capillary column it formed a continuous
liquid plug of water within the capillary column several on long. The
result was erratic detection of water soluble compounds that were flushed
through the column by the water plug and variable retention times for
non-polar compounds. The water plugs also extinguished the flame in the
detector. Conventional forward and reverse flow trap drying operations at
various trap temperatures were tried and found to be of no value. Because
of the water problem associated with other than Tenax traps, Tenax "A" and
-18-
-------
"8" traps were selected for all development work. Note: Since this evalua-
tion the COS 320 software was modified by the manufacturer to allow the "A"
trap to be maintained at elevated temperatures during the purge cycle, 1t 1s
likely that a trap temperature can be selected for other sorbents that will
allow water to be vented while the target compounds are concentrated as is
the case of Tenax operated at 22-25*C. It is important to add that during
this lengthy sorbent evaluation numerous "water plug" analyses were per-
formed using SE-30 bonded and non-bonded phases with no observed degradation
of either capillary column.
-19-
-------
SEQUENTIAL TRAP TRANSFER CONDITIONS
Several experiments were performed to determine the critical parameters
required to quantitatively backflush a variety of compounds from the "A"
trap to the "B" trap. This was accomplished by Injecting 1 v\_ volume of the
test mixture into the "A" trap injector as the purge gas was flowing at 40
mL/minute through the purging device filled with 5.0 ml of reagent water.
After 11.0 minutes the "A" trap was baclcflushed into the "B" trap. Oesorb
time, flow rate and temperature were evaluated as variables. Based upon
oast observations of the system performance an initial evaluation was
performed using a fixed flow rate of 15 mL/minute, and a fixed desorb
temperature of 200*C. Oesorb times were varied from 200 seconds down to 20
seconds. Table 5 lists the peak areas relative to decane obtained from a
direct column injection chromatogram and those obtained from the sequential
trapping operations using different transfer times. From these data it is
evident that all of the compounds are not released from the trap over the
same period of time. The lower boiling compounds are released first
followed by the higher boiling n-alkanes and finally the polar compounds.
Close examination of the chromatograms show that decane was almost quanti-
tatively transferred before 20 percent of the 2,6-dimethylaniline or
naphthalene was transferred. This table shows that at 15 mL/minute and
200*C it took a minimum of 100 seconds to quantitatively transfer the most
retentive of the test compounds from the "A" trap to the "B" trap. Extend-
ing the transfer times up to 200 seconds did not adversely affect the
quality of the data. A second set of trap conditions were then evaluated
where the trap temperature was elevated to 250*C, the flow was maintained at
-20-
-------
15 ml/minute, and the desorb times were varied as 1n the previous experiment.
It was found that raising the trap temperature decreases the quantitative
transfer time for the volatile reactive components and the n-alkanes from
100 seconds at 200*C to 80 seconds at 250*C with no evidence of thermal
breakdown.
A final set of trap conditions were evaluated where the trap desorb
temperature was 200*C and the transfer flow rate changed from 15 mL/minute
to 10 mL/minute. As before, the desorb times were varied. Decreasing the
flow rate during the trap transfer step had little effect upon the resulting
data. The transfer times and recoveries were nearly Identical to those
obtained at the 15 mL/minute flow rate.
From this series of experiments, it is demonstrated that all of the
compounds do not backflush from the sorbent trap as a sharp plug of material
but elute, depending upon their boiling point and polarity. Oesorb time and
desorb temperature appear to be the most important variables. Similar
studies involving trichlorobenzenes indicated that desorb time of 120
seconds is required for quantitative transfer. Based upon these observa-
tions and other data, backflushing the "A" trap at 180*C with a flow rate of
15 mL/min for 200 seconds was selected for subsequent studies. Higher trap
temperatures were not selected because previous method development research
associated Tenax trap failure with desorption temperatures in excess of
200*C and because excessive background peaks appeared in the FID blank
chromatograms whenever trap temperatures exceeded 200*C.
-21-
-------
TRAP TRANSFER TO COLUMN
A similar study was performed to determine the conditions required to
quantitatively transfer compounds from the 1.8 mm ID "B" trap (loaded
according to the previously described conditions) to the analytical column.
Since the desorb flow rate through the "B" trap supplies the carrier gas
flow to the analytical column, the flow must be adjusted to provide optimum
flow conditions for the analytical column and not optimum chromatographic
transfer conditions. For the column used in this evaluation, the flow was
fixed at 10 mL/minute. Similarly, as previously stated, trap desorb
temperatures in excess of 200*C are not desirable, therefore, only desorb
times were evaluated. Trap B desorb times in excess of 20 seconds were
adequate for quantitative transfer of the compounds evaluated. Extended
desorb times up to 120 seconds did not appear to adversely affect the
quality or the appearance of the chroma to gram. For these reasons desorb
times of 120 seconds were selected to insure maximum transfer of a wide
variety of analytes.
COLUMN SELECTION
During the course of this study several capillary columns were briefly
evaluated to determine which internal diameters and film thicknesses are
best suited for sequential purge and trap analyses. For these studies, a
1.7 mm ID copper "B" trap packed with 23 cm of Tenax desorbed at 200*C for
120 seconds was used. The "8" trap desorb flow rate was adjusted to provide
-22-
-------
a 20 to 40 cm/second linear velocity of helium through the test column. It
was found that 0.2 to 0.32 mm ID fused silica columns with film thicknesses
between 0.25 and 1.0 urn were of limited value because they could be used
only for compounds that are reconstituted on the capillary under nearly true
cold trapping conditions. Non-reconstituted compounds eluted from these
columns as poorly defined broad peaks.
For glass columns with a 0.5 ran ID and film thicknesses near 1 urn,
visual comparisons of chromatograms from both desorption and direct injec-
tions at isothermal temperatures showed that the desorption process
adversely affected the peak geometry of the compounds eluting in the initial
isothermal area of the chromatogram. Peak geometries improved after minimal
programming.
By far the best chromatograms were obtained using 0.75 mm ID glass
columns with film thicknesses near 1 »m. Visual comparisons of peaks elut-
ing within the initial isothermal area of the chromatograms were nearly
identical. Comparisons of the number of theoretical plates/meter for
desorption chromatograms to direct injection chromatograms show that the .
average for pentane through octane was 1500 theoretical plates for direct
injection and 1100 theoretical plates/meter thus, even for desorption
chromatograms for wide bore capillary columns, the desorption process can
have an adverse effect upon compounds eluting in the early isothermal area
of the chromatogram. Increasing the film thickness of the column could help
to minimize this problem, however, at the time these experiments were
-23-
-------
performed film thicknesses* 1n excess of 1 urn were not commercially avail-
able. Based upon these observations 0.75 mm 10, 1 un df columns were
selected. Furthermore, since such columns typically exhibit about 1/3 the
number of theoretical plates/meter as 0.25 mm 10 columns, 60 m column
lengths were chosen in order to obtain the resolving power required to
separate complex mixtures of synthetic organic compounds.
Selection of a liquid phase was based solely upon its commercial avail-
ability and resolution of complex mixtures of purgeable compounds of current
interest to the Agency. Three phases were evaluated: methyl silicone,
SE-54 and Carbowax 20M. Of these liquid phases, the metnyl silicone phases
were found to be superior. The SE-54 was unable to resolve many component
pairs that were easily resolved by the methyl silicone phase. The Carbowax
20M chr one to grams were lengthy and the early eluting components generated
broad fused peaks indicating the liquid phase was unsuitable (a solid) at
the temperatures required to separate the most volatile compounds tested.
A number of temperature programs were evaluated and the following was
found to be best suited for resolving complex mixtures of synthetic organic
chemicals of current interest to the Agency. The SE-30 column is maintained
at 10*C for 4 minutes and then prograimed at 4*C per minute to 210*C. The
column is held at 210*C until all of the compounds elute or just before the
next analysis. Helium is used as the carrier gas flowing at 20 cm/second
(measured at 115*C). If only compounds eluting after methylene chloride are
to be analyzed, then the initial column temperature was raised to 30-40*C.
-24-
-------
PROCEDURE EVALUATION
Through the assessment of the previously described critical parameters
and from previous purge and trap methods development programs, the condi-
tions listed in Table 6 were selected as the most promising combination for
determining a wide variety of synthetic organic chemicals in water by purge
and sequential trapping capillary column gas chroma to graphy.
In the preamble to the "National Interim Primary Drinking Water Regula-
tions; Control of Trihalomethanes (THMs) in Drinking Water; Final Rule,"-it
is stated that "to qualify for interim certification, laboratories will be
required to demonstrate their ability to analyze the Performance Evaluation
samples provided to them to within 20 percent of the "true value" for each
of the THMs as well as for the total of the THMs in the samples using at
least one of the approved methods." One of the initial evaluations of the
proposed method was to determine if the procedure could reliably generate
data within *20 percent of the true value for an actual Performance Evalua-
tion sample. The following experiment was designed to determine the
accuracy and precision of the proposed method while minimizing the degree of
operator skill required to perform the analysis. The CDS 320 controller was
programmed to automatically function exactly according to the parameters
described in Table 6. Primary dilutions of chloroform and dibromochloro-
methane at 10,000 ug/mL in methanol were obtained from the EMSL-Cincinnati
Repository for Toxic and Hazardous Materials. Methanolic dilutions of
bromodichloromethane and bromoform were prepared in-house according to
-25-
-------
USEPA Method 501.1 (Ref. 1). Two multi component methanollc secondary
dilutions were prepared from these primary standards. Dilution No. 1
contained 125 ng/ML of each trlhalomethane and Dilution No. 2 contained
500 ng/yl of each trihalomethane. Three aqueous standards were prepared by
spiking 1000 ml of reagent water with 20.0 yl of Dilution No. 1, 100 ml of
reagent water with 20.0 uL of Dilution No. 1, and 100 ml of reagent water
with 20.0 uL of Dilution No. 2. The aqueous standard solutions were
analyzed starting with the low level, 2.5 ug/L, followed by the mid-range,
25 ug/L, and finally the high range, 100 ug/L, trihalomethane standards.
The data system was calibrated for each trihalomethane using the three point
calibration curve. The response of each THM was linear and passed through
zero providing .999 or better coefficients of determination. Eight days
later an EMSL-Cincinnati Quality Check trihalomethane concentrate was
diluted according to instructions in reagent water and analyzed in order to
verify the validity of the 8-day old calibration date. (True value data are
supplied with EMSL-Cincinnati Quality Check Samples.) Each trihalomethane
was found to be within 10 percent of the reported value validating the
calibration curves for each THM. This system evaluation was followed by
reagent water analyses (system blanks) and replicate analyses of two
different trihalomethane Performance Evaluation samples (PE-1 and PE-2).
The true values of PE-1 and PE-2 were unknown at the time of analysis. Each
of the samples were diluted in reagent water according to instructions and
analyzed in quadruplicate. A second different Quality Control Sample was
analyzed between the PE-1 and PE-2 sample in order to monitor the continuing
performance of the system. Tables 7 and 8 list the resulting concentrations
-26-
-------
taken directly from the data system reports. Just prior to analyzing PE-2
Dilution No. 1, the quality control sample containing an unusually high
concentration (660 ug/L) of chloroform was analyzed. It 1s believed that
system memory (- 0.5 percent carry-over) caused a false high chloroform
value in the PE-2-1 analysis, therefore, the PE-2-1 chloroform data were
deleted as an operator generated outlier. It is likely that the PE-2-2
value should be deleted also but it was not. After reporting the concen-
trations, the true values were obtained. Tables 7 and 8 show that the
resulting data were accurate in that in no case did the average value differ
from the true value by more than 10 percent. Moreover, with the exception
of the PE-2-1 and PE-2-2 chloroform values, suspected to be accidently
contaminated, the precision of the procedure is such that at the 99 percent
confidence limit all of the THMs analyzed in the PE samples easily fall
within the 20 percent acceptance criteria. This clearly demonstrated that
the proposed procedure is capable of generating accurate and precise tr1-
halomethane data utilizing operator skills with a rating of only one (Ref 2).
ACCURACY AND PRECISION STABILITY STUDY TAP MATER
Two liters of Cincinnati tap water were dechlorinated by the addition of
200 mg of sodium thiosulfate. The resulting quenched tap water was allowed
to stand head-space free for 18 hours at room temperature to allow trihalo-
methane intermediates to decompose to provide stable THM values with time.
A complex spiking mixture of organic compounds in methyl alcohol was
prepared according to Table 9. The compounds were selected based upon the
-27-
-------
current and long range needs of various Agency programs, the ability of the
column to adequately resolve them for accurate measurement (retention data
on Table 9) and to obtain data for representative compounds defining a wide
range of purging efficiencies. Concentrations were selected so that the FID
would provide similar peak height signals under the analytical conditions
stated in Table 6. One liter of the quenched 18-hour-old tap water was
spiked with 100 yl_ of the spiking solution resulting in the concentrations
listed in Table 6. Twenty-four 40 ml septum seal purge and trap sample
bottles were randomly filled with the resulting spiked mixture. Six of the
bottles were sealed and stored at room temperature (Thio-22*C); six of the
bottles were sealed and stored at 4*C (Thio-4*C). Six bottles were acidi-
fied with two drops of HC1 (1+1) to give a pH of 1.8, sealed and stored at
22*C (Thio-HCl-22*C). 100 vl of HgCl2 (0.5g/100 ml 1n reagent water)
solution was added to each of six bottles and stored at 22*C (Thio-Hg-22*C).
Six bottles were filled with non-spiked quenched tap water sealed and stored
at 22*C (Thio-blank). On day zero (spike day) the gas chromatograph was
calibrated at a single concentration using a 20.0 uL aliquot of the spiking
mixture (Table 9), diluted to 100 mL in reagent water. Duplicate analyses
upon a Thio-blank, Thio-22"C, Thio-HCl-22"C and a single analysis upon a
Thio-Hg-22*C sample were performed. Over the next 24 days the Instrument
was recalibrated each analysis day and similar analyses were performed. The
results appear in Tables 10 through 15.
Recovery data were corrected for the average THM values found in the
Thio-blanks. The addition of HC1 was selected to evaluate its performance
-28-
-------
as a bloclde and a chemical stabilization agent. HgClg was Included 1n
the study to evaluate Its performance as a bloclde.
Table 10 lists the averaged results of all of the spike day data. These
include two Thio-22*C; two Thio-HCl-22*C and a single Thio-Hg-22*C analysis.
Most of the compounds provided accurate and precise recoveries for the spike
day analyses in all the sample matrices. Noteworthy exceptions are allyl
bromide, 2-chloroethyl vinyl ether, and pentachloroethane. Allyl bromide
rapidly disappeared from each of the sample matrices studied. Over the
fi-hour period of time represented by these data, an average recovery of only
40 percent was obtained with a 47 percent relative standard deviation.
2-chloroethylvinyl ether in the Thio-HCl-22* and Thio-Hg-22*C preserved
samples also disappeared. Pentachloroethane rapidly decomposed to form
tetrachloroethylene in the Thio-22*C and Thio-Hg-22*C matrices, but was
stable in the Thio-HC1-22*C matrix.
Table 11 lists the method accuracy and precision for samples stored up
to 24 days in the Thio-22*C matrix. Comparing the 24-day averaged recovery
data to the spike day (Table 10) and 18-day recoveries (Table 15) show that
most of the compounds evaluated are stable indicating little or no bio-
logical activity. It is important to note that previous studies have shown
that biological activity can develop in such samples (Ref. 3). The
following compounds were found to be unstable in this particular matrix:
allyl chloride, allyl bromide, cis and trans-l,3,-dichloropropene,
-29-
-------
1,1,2,2-tetrachloroethane and pentachloethane. Progressive losses of
hexachloroethane and styrene indicate they too may be lost upon storage but
based upon the precision of the analytical methodology not at a significant
rate for this overall method. Pentachloroethane and 1,1,2,2-tetrachloro-
ethane decomposed to form tetrachloroethylene and trichloroethylene,
respectively, providing the likelihood of false positive identifications if
samples are stored in this manner.
Table 12 lists the results of the quenched sample storage at 4*C. Com-
paring the 24-day averaged recoveries to the spike day and day 18 recoveries
show that the same analytes are affected as Table 11 but generally with
improved recoveries.
The addition of mercury to the matrix (Thio-Hg-22*C) appears to have a
detrimental effect upon sample storage. Table 13 shows that "in addition to
the compounds affected by simple 22*C storage, a total loss of 2-chloroethyl
vinyl ether was noted along with a significant increase in the concentration
of 1,2-dichlorethane with time.
The adjustment of the sample pH with HC1 was originally intended to
observe its properties as a biocide. The data in Table 14 show that the
addition of HC1 to the sample matrix effectively halted the decomposition of
tetrachloroethane to form trichloroethylene and pentachloroethane to form
tetrachloroethylene. Compared to spiked day recoveries the detrimental
effects of preservation with HC1 are the loss of 2-chloroethyl vinyl ether
and styrene.
-30-
-------
Table 15 compares the average study recoveries and the average of
duplicate analyses performed on day 17 or 18. It appears from these data
that for a general analytical method the best sample storage technique would
be a combination of preservation with HC1 and storage at 4*C,
-31-
-------
METHOD ACCURACY AND PRECISION AND ANALYTE STABILITY IN BIOLOGICALLY ACTIVE
RIVER WATER
A prestudy evaluation of spiked Ohio River water showed that the sample
of river water obtained for this phase of the evaluation demonstrated no
biological activity toward any of the compounds listed in Table 9 over a one
week period of time when stored at 22*C. It was not determined if one or
more of the compounds present in the spiking solution inadvertently acted as
a biocide or if the naturally occurring microbes were not accustomed to
degrading the target compounds. In an effort to rapidly generate a biologi-
cally active sample matrix for as many compounds as possible, Ohio River
water was inoculated with a mixture of commercially available bacterial
cultures adapted to digest fresh water wastes containing hydrocarbons and
polychlorinated biphenyls. For this experiment 2 mL of the bio-spiking
solution and 1998 mL of Ohio River water was added to a 2L separatory
funnel. Four hundred uL of the methanolic spiking solution described in
Table 9 was injected below the surface of water (resulting in a mixture
containing the compounds at two times the concentrations listed in Table 9).
The separatory funnel was sealed and mixed by inverting twice. Thirty purge
and trap septum seal vials were then filled to overflowing using the Teflon
stopcock on the separatory funnel to control the flow and to minimize
turbulence as the bottles were filled. Six bottles were sealed and immedi-
ately stored at 4*C. Six bottles were sealed and stored at 22*C. Six
bottles were spiked with 100 yL of Slime-Trol RX-34 solution, sealed and
stored at 22*C. Six bottles were spiked with 100 yL HgCU solution,
sealed and stored at 22*C and finally, six bottles were spiked with 5 drops
-32-
-------
of hydrochloric add solution (1+1), sealed and stored at 22*C. The bio-
spiking solution was prepared by adding 1.0 g of Sybron PCS culture, 1.0 g
• Tfci
of Sybron hydrocarbon culture, and 1.0 g of Polybac Hydrobac Cl culture
to 25 ml of reagent water. The solution was allowed to stand 18 hours at
22*C with air bubbling through the mixture before use. The Slime-Trol
solution was prepared by diluting 0.6 g of Slime-TrolA RX-34 (a commercial
water soluble biocide from Betz Paperchem, Inc.) to 10.0 ml using reagent
water (Ref. 4). The HgCl2 solution was prepared by dissolving 0.5 g of
HgCU in 100 ml of reagent water. Ohio River water blank analyses were
performed and were found to be free of any interferring compounds at levels
significant to this study.
On day zero (spike day) the gas chromatograph was calibrated using
spiked reagent water at concentrations identical to the levels used for the
study. Spike day calibration data showed that the 2-chloroethylvinyl ether
had disappeared from the spiking solution, therefore, it does not appear in
the study data. Midway through this study an instrumental problem developed
interrupting the planned frequency of analyses and adversely affected the
precision of the data. Tables 16 through 21 list the results of the study.
Individual compounds contained in the samples stored at 22*C do show sta-
tistically significant evidence of die-off due to chemical and biological
activity. Storage at 4*C retards both chemical and biological losses while
the addition of the three'biocides appears to halt biological activity. As
in the previous study, HC1 addition appears to be the best preservation
reagent studied since, in addition to acting as a biocide, it also retards
chemical decomposition associated with pentachloroethane and tetrachloro-
ethane. Samples preserved with Slime-Trol RX-34 demonstrate little
-33-
-------
advantage over mercury preserved samples. After about 18 days of storage
chromatograms of Slime-Trol RX-34 preserved samples provided no resolution
between bromodlchloromethane and 1,1,2-tr1chloroethylene Indicating that an
unknown compound was formed with time that eluted within this retention
area. The appearance of this compound prevented accurate measurement of
either compound. Some of the Ohio River water spiked data appears to con-
flict with the spiked tap water study in that the concentration of 1,2-
dichloroethane did not increase with time when the sample was preserved with
mercury. Also styrene disappears at a significant rate 1n the mercury
preserved samples and at a reduced rate in the HC1 preserved samples.
METHOD DETECTION LIMITS
Reagent water containing 50 mg/L of sodium thiosulfate was spiked with
the methanolic mixture listed in Table 9 at the rate of 5 WL of the spiking
mixture per liter of water. The concentration of each compound was selected
so that each peak in the chromatogram would be at least 5 times higher than
the average noise level at the most sensitive detector setting usable with
the system. Actual concentrations are 0.05 times those listed in Table 9.
Peak heights for the various analytes appeared in the chromatogram between 5
and 7 mm. Peaks normally occurring in reagent water analyses (system
blanks) attributable to system background were well resolved from the test
compounds. Four purge and trap sample bottles were filled and sealed with
the dilute mixture. The gas chromatograph was calibrated with a single
point calibration standard at 40 times the concentration of the method
detection limit spike. Once the system was calibrated, the contents of each
-34-
-------
sample bottle was analyzed 1n duplicate until seven analyses were performed.
Table 22 lists the resulting data and the calculated method detection limits
(Ref 5). From these data 1t 1s apparent that the method detection limit 1s
primarily dependent upon the sensitivity of the flame lonization detector to
the target compound. The method detection limit is highest for the highly
halogenated compounds - 1 ug/L decending down to the alkyl substituted ben-
zenes at - 0.1 ug/L. Although no significant peaks were noted in the blank
analysis, high recoveries are attributable to an accumulation of errors
associated with the failure to bracket the spike with two standards, and the
additive effects of background and system memory.
SYSTEM MEMORY
Throughout this method evaluation, indications of analyte carry-over or
system memory appeared whenever low level analyses followed high level
samples. This problem existed even though the purging device was flushed
with reagent water two or three times between analyses and after exchanging
purging devices.
In an effort to document the extent of the carry-over problem, the
following experiment was performed. A moderate level standard solution was
prepared by diluting 20.0 wL of the spiking mixture described in Table 9 to
100.0 ml with reagent water. The purging device was flushed with reagent
water followed by a reagent water analysis at the most sensitive FID setting
in order to establish normal system background values. A 5-mL aliquot of
the moderate level standard was analyzed followed by three reagent water
-35-
-------
flushes of the purging device. After the moderate level standard chromato-
gram was completed, a 5-mL aliquot of reagent water was analyzed at the most
sensitive settings. The purging device was again flushed with three reagent
water flushes followed by a second reagent water analysis.
Finally the purging device was exchanged with a new one followed by a
final reagent water analysis. All of these analyses were performed with the
CDS-320 valve and internal plumbing oven set at 125*C. The valve and
internal plumbing oven temperature was then raised to 200*C and a similar
sequence of analyses were performed with the exception that the purging
device was not exchanged for the final analysis. The system memory was then
determined by dividing the peak area of the moderate standard into the blank
corrected peak areas obtained from each of the reagent water analyses times
100. These values appear in Table 23 (system memory).
Based upon these tests it is evident that much of the system memory is
due to sorption of the high boiling compounds within the CDS-320 plumbing
and not the purging device as one would initially believe. Based upon these
observations, system memory and not purging efficiencies and column perform-
ance appears to limit the compounds that can be accurately determined by
sequential trapping capillary column gas chromatography. With the valve
oven operated at 200*C it appears that compounds that exhibit less than a 2%
carry-over can be successfully analyzed. Operating the valve oven at tem-
peratures in excess of 200*C is not practical for the system evaluated as
excessive background occurred due to system bleed.
-36-
-------
CONCLUSIONS
The accuracy and precision data gathered during this study for over 40
compounds clearly demonstrate that a properly optimized automated purge and
sequential trapping capillary column gas chromatograph can generate accurate
and precise data for a wide variety of synthetic organic compounds contained
in drinking water and related matrices. Each critical parameter was
identified and optimized in the study. Once the system was optimized, the
automatable features and the inherent ruggedness of the capillary FID-data
system allow the system to generate dependable data with minimal operator
skills. Using a flame ionization detector, the method detection limits vary
between 0.1 and 1 ug/L for reagent water spikes.
The holding time data show that preservation is necessary to guarantee
integrity of certain compounds. Sample storage at 4*C is far superior to
storage at 22*C and the addition of HC1 (pH adjustment to 2) effectively
halts biological degradation and stops chemical decomposition of penta-
chloroethane and tetrachloroethane which form tetrachloroethylene and
trichloroethylene, respectively. Other biological controls show no
advantages over pH adjustment. System memory to high boiling compounds,
which in turn affects accuracy and precision, appears to be the compound
limiting factor for the method.
-37-
-------
REFERENCES
1. "The Determination of Halogenated Chemicals 1n Water by the Purge and
Trap Method," Method 502.1, EPA 600/4-81-059, Environmental Monitoring
and Support Laboratory, U. S. Environmental Protection Agency,
Cincinnati, Ohio 45268. (1981).
2. "Handbook for Analytical Quality Control in Water and Wastewater
Laboratories," EPA 600/4-79-019, Environmental Monitoring and Support
Laboratory, U. S. Environmental Protection Agency, Cincinnati, Ohio
45268. (1979).
3. "The Determination of Volatile Aromatic Compounds in Drinking Water and
Raw Source Water," Thomas A, Bellar and James 0, Liehtenberg,
Environmental Monitoring and Support Laboratory, U. S. Environmental
Protection Agency, Cincinnati, Ohio 45268. (1981).
4. Report on "Preservation and Sample Storage of Volatile Organics in
Drinking and Raw Source Waters," USEPA Contract No. 68-03-3103. In
preparation. Environmental Monitoring and Support Laboratory, U. S.
Environmental Protection Agency, Cincinnati, Ohio 45268.
5. Glaser, J.A., Foerst, D.L., McKee, 6.D., Quave, S.A., Budde, W.L.,
"Traca Analysis for Wastewaters," Environmental Science and Technology,
15, 1426 (1981).
-38-
-------
Table 1. Initial Evaluation of the Sequential Trapping Capillary System
Percent Response Values for Test Mixture
Compound ^^__^___
2,6-dlinethyl 2,6-dlmethyl naphthalene
2-octanone 1-octanol phenol aniline
Primary chroma to gram
Initial Installation
Chr oma to gr am
CDS transfer line
chr OTO to gram
Modified transfer
line chromatogram
Sequential Trapping
Chromatogram
82
75
80
38
71
74
28
67
32
78
85
29
86
16
79
85
51
89
49
100
112
84
111
33
-39-
-------
Table 2. Effect of Various Traps Upon Chroma to graphic Data
Retention Times (minutes)
n-Cs n-Cg n-Cy n-Cs
Direct injection
2.7 mm stainless steel trap
1.8 mm glass trap
1 .7 mm copper trap
Retention time difference4
4.03
4.47
4.47
4.46
.44
4.52
4.94
4.94
4.93
.42
5.58
5.96
5.96
5.98
.38
7.88
8.25
8.25
8.25
.37
n-Cg
11.35
11.55
11.55
11.54
.20
n-Cio
15.18
15.57
15.57
15.44
.39
Peak Area Comparisons
(Integrator Units)
Direct injection
2.7 mm stainless steel trap
1.8 mn glass trap
1.7 mm copper trap
Direct Injection
2.7 mm stainless steel trap
1.8 mm glass trap
1 .7 mm copper trap
Direct injection
2.7 mm stainless steel
1.8 mm glass trap
1.7 rim copper trap
460
376
396
380
Peak
133
52.0
78.2
83.3
Peak
.03
.069
.049
.044
437 400
434 387
424 376
446 369
Heigh* Comparisons
(run)
115
50.7
84.3
89.2
Width Cooipar
(seconds)
•°$
86.0
52.8
62.0
63.7
isons
.045
— b .070
.049
.047
.060
.056
365
352
320
308
52.3
38.5
39.8
39.7
.067
.087
.077
.074
337
333
298
283
54.3
42.5
42.1
40.5
.060
.074
.068
.067
310
309
278
265
34.5
27.3
25.8
23.0
.086
.107
.10
.11
Chroma to graphic conditions: 70*C 8 minutes - 8*/minute to 100*C
aAverage difference in retention time between direct injection and thermal
desorption
bData system malfunction
-40-
-------
Table 3. Trap Performance Non-Cold Trapping Chromatography
2 -oc tan one 1-octanol
Percent Response
Area Ratio3
Theor. Plates/Meter
Percent Response
Area Ratio3
Theor. Plates/Meter
Percent Response
Area Ratio3
Theor. Plates/Meter
Percent Recovery
Area Ratio3
Theor. Plates/Meter
1.0 i
77
77
—
1.0
53
67
—
1.0
69
66
—
1.0
73
74
—
2,6-d1methyl-
phenol
n-Cn
2, 6-di methyl
anli ine
naphthalene
n-C12 n-Ci3
jL Injection into column
71
70
—
pL injection 2
29
57
—
pL Injection 1
44
48
—
pL Injection 1
46
64
—
86
78
—
95
—
86
74
—
115
92
—
90 83
— 1030
.7 mm ID "B" Trap
NDb
NDb
—
147b
—
.8 mm ID glass-lined "B"
60
56
~
.7 mm ID copper
65
64
—
95
—
"B" Trap
116
— —
22
48
—
Trap
55
60
—
60
68
— —
38
68
—
55
77
—
83
90
— —
101 72
— 175
80 76
— 452
90 87
— 492
3Relative to n-decane
for 2,6-dimethyl phenol fused with n-undecane
-------
Table 4. Trap Performance Cold Trapping Chromatography
2,6-dlmethyl 2, 6-di methyl
2-o ct an one 1-octanol phenol n-Cn aniline
Percent Response
Area Ratio3
Theor. Plates /Meter
Percent Response
Area Ratio3
Theor. Plates/Meter
Percent Response
Area Ratio3
£ Theor. Plates/Meter
i
Percent Response
Area Ratio3
Theor. Plates/Meter
1.0 pL
86
83
—
1.0 til
no data
82
—
1.0 ML
78
81
1.0 ML
77
84
—
injections
76
77
—
injections
70
74
—
Injections
62
76
Injections
65
78
—
into column
95
87 101
—
2.7 mm "B" trap
86 —
79 97
—
1.8 mm glass-lined "B" Trap
87 —
87 104
1.7 mm ID Copper "B" Trap
88 —
86 104
— —
97
90
—
93
81
—
82
83
91
82
— —
Naphthalene n-Cjp "-£13
123
110
—
124
102
—
113
107
120
107
— —
105 107
— 1230
101 98
— 1210
103 102
— 1220
102 104
— 1180
aRelative to n-decane
-------
Table 5. Sequential Trapping 200* C at 15 mL/nrlnute
Direct Desorptlon Time (seconds)
Injection 200 150 120 100 80 60 40 20
2-octanone .80 .83 .81 .85 .81 .81 .80 .70 NO
1-octanol .85 .76 .76 .77 .75 .76 .74 .46 NO
2,6-dimethyl phenol .89 .86 .84 .88 .85 .85 .62 NO NO
n-undecane .98 1.00 .99 1.03 1.02 1.19 1.13 .85 Trace
2,6-dimethyl aniline .91 .95 .90 .97 .87 .73 .30 NO NO
Naphthalene 1.09 1.15 1.08 1.15 1.03 .83 .29 NO NO
n-dodecane .98 1.00 .98 1.07 .98 .98 .97 .70 NO
n-tridecane 1.02 1.01 .99 1.07 .98 .92 .84 .42 NO
Areas of the Resulting Peaks Relative to n-decane
-43-
-------
Table 6. AnaUylcal Conditions
1. Purging conditions
Sample volume: 5.0 mL
Purge gas: Helium
Purge gas flow rate: 40 ml_/m1nute
Sample temperature: Room Temperature (22 * 2*C)
Trap "A": 0.105" ID stainless steel packed with
23 cm of Tenax GC 60/80 mesh sorption
temperature < 29*C
2. Sequential Trapping
Backflush Trap "A" at 180*C * 10*C
For 120 sec. flow rate
15 mL/minute
Trap "A" heating rate (outside surface) 10*/sec
Trap "B" 1.5 to 1.8 mm ID
Copper or glass lined stainless steel
packed with 23 cm of Tenax GC 60/80 mesh
operated at room temperature 22*C * 2*C
3. Desorb "BM Trap to Column
Backflush trap "B' at 180*C * 10*C
for 120 seconds at a flow rate between
8 and 12 mL/minute (column flow rate)
Trap "B" heating rate (outside surface)
10*/sec.
4. Column
0.75 mmID x 60 m long coated with SE-30 (Bonded)
1 um film thickness with a reported 101 percent coating
efficiency carrier gas helium flowing at 27 cm/sec
measured at 115*C (10 mL/minute).
5. Program
10*C Isothermal for 4 minutes, then program at 4*/minute to 210*C
6. Miscellaneous
With the exception of the purging device, all transfer
lines and valves were maintained at^OO C. Traps were
conditioned between analyses at 200*C for various periods
of time to minimize carry-over.
The purging device was flushed out twice with - 7 mL of
reagent water between each analysis.
-44-
-------
Table 7. Analysis of Performance Evaluation Sample - PE-1
Concentration (uq/L)
Order
Sample of Brorodlchloro- Oibromochloro-
Ident1f1cat1on Analysis Chloroform methane methane Bromoform
PE-1-1
PE-1-2
PE-1-3
PE-1-4
1
2
5
6
85.6
83.0
81.6
81.0
82.2
79.9
82.7
78.3
103
105
101
101
52.5
55.1
53.1
53.1
Average 82.8 80.8 103 53.5
Std. Deviation 2.05 2.05 1.91 1.14
99 Percent Confidence
Interval (wg/L) 76.7 to 88.9 74.7 to 87.0 96.8 to 108 50.1 to 56.9
True Value (u9/L) 86.8 82.1 107 54.9
20 Percent Acceptance
Interval Around True 69.4 to 104 65.7 to 98.5 85.6 to 128 43.9 to 65.1
Value (ug/L)
Percent Recovery 95.4 98.4 96.3 97.4
-45-
-------
Table 8. Analysis of Performance Evaluation Sample - PE-2
Concentration (ug/L)
Sample
Identification
2-1
2-2
2-3
2-4
Average
Std. Deviation
Order
of
Analysis
3
4
7
8
Chloroform
19.7*
17.9
14.7
15.5
16.0
1.67
Bromodlchloro-
me thane
7.91
8.43
8.30
8.36
8.25
0.23
01 br one chl or o-
me thane Bromoform
17.8
18.0
16.7
17.2
17.4
0.59
16.0
16.1
15.9
15.9
16.0
0.095
99 Percent Confidence
Interval (Mg/L) 11.0 to 21.0 7.55 to 8.95 15.7 to 19.2 15.7 to 16.3
True Value (ug/L) 15.3 9.12 17.8 16.5
20 Percent Acceptance
Interval Around True 12.2 to 18.4 7.30 to 10.9 14.2 to 21.4 13.2 to 19.8
Value (ug/L)
Percent Recovery 105 90.5 97.8 97.0
*Value deleted (see text)
-46-
-------
Table 9. Spiking Mixture Concentrations and Retention Data
Compound
Pen tan e°
1,1-Dichloroethylene
Methyl ene chloride
Al 1 yl ch 1 or i de
t-1 , 2-0 1 ch lor oeth yl ene
c-1 , 2-01 ch 1 or oethyl ene
Allyl bromide
Chloroform
1,2-Dichloroethane
1, 1 , 1-Tr ichl or oe thane
Benzene
Carbon tetrachloride
1,2-Oichloropropane
Br omo di ch 1 or ome th an e
1,1,2-Trichloroethylene
2-Chloroethylvinyl ether
Heptane13
1, 3-01 chloro pro pen e
1 , 3-D1 ch 1 oro propene
1,1,2-Trichloroethane
Toluene
Di br omo ch lor ome thane
1,1,2,2,-Tetrachloroethylene
Octane11
Chlorobenzene
Ethylbenzene
Br omo form
p-Xylene
Styrene
1 ,1 , 2 , 2-Tetr ach 1 or oeth an e
Nonaneb
Bromobenzene
n-Propyl benzene
Pentach 1 or oeth ane
m-Di Chlorobenzene
p-Di Chlorobenzene
Decane0
o-Di Chlorobenzene
1 ,2-Dibromo-3-chloropropane
He xach lor oeth ane
1,3,5-Trichlorobenzene
Cone.
Spiking
Solution
(ng/ML)
99.8
375
100
100
99.7
99.5
299
100
175
25
303
75
350
125
125
—
151 a
150
25
450
175
—
50
25
498
25.0
25.0
250
_
49.9
25
400
50
50
50
250
250
99.9
Cone.
Aqueous
Dilution
(ug/U)
10.0
37.5
10.0
10.0
10.0
10.0
29.8
10.0
17.5
2.5
30.3
7.5
35.0
12.5
12.5
—
15.1
15.0
2.5
45.0
17.5
— —
5.0
2.5
49.8
2.5
2.5
25.0
_
5.0
2.5
40.0
5.0
5.0
—
5.0
25.0
25.0
10.0
Retention
Time
(Min.)
6.24
6.50
6.69
6.81
7.86
9.27
9.54
9.71
10.78
11.19
11.93
12.18
13.34
13.70
13.80
14.60
15.41
16.43
16.60
17.08
17.77
19.23
19.43
20.60
21.55
21.81
22.00
22.73
22.95
23.99
24.33
25.70
26.25
27.63
27.86
28.35
28.81
30.20
30.90
33.14
-47-
-------
Table 9. (Continued)
Compound
1,2,4-Trichlorobenzene
Naphthalene
1,2,3-Trlchlorobenzene
Hexachlorobutadiene-1,3
Oodecaneb
1,2,4,5-Tetrachlorobenzene
1 ,2,3,4-Tetrachlorobenzene
1-Chloronaphthalene
2-Chlorobiphenyl
Cone,
Spiking
Solution
(ng/wl)
100
50
99.9
250
151
150
150
150
Cone.
Aqueous
Dilution
(ug/L)
10.0
5.0
10.0
25.0
15.1
15.0
15.0
15.0
Retention
Time
(M1n.)
34.77
35.02
36.03
36.76
40.03
40.33
41.76
41.84
45.51
aMixture of els and trans isomers assumed to be 50/50 mixture.
bn-alkanes used as Internal standard.
-48-
-------
Table 10. Spike Day Accuracy and Precision Quenched Tap Water
Compound
1 , 1-01 ch 1 or oethyl ene
Methylene chloride
Allyl chloride
t-1 ,2-01 ch lor oethyl ene
c-1 , 2-01 ch 1 or oethyl ene
Allyl bromide
Chloroform
1,2-Dichloroethane
1 ,1 ,1-Tri chl or oeth ane
Benzene
Carbon Tetrachloride
1 , 2-01 chl or opr open e
Bromodi ch 1 orometh ane
1,1,2-Trich lor oethyl ene
2-Chloroethyl vinyl ether
1 , 3-01 ch 1 or o pro pen e
1 ,3-01 chl or opr open e
1,1,2-Trich lor oeth ane
Toluene
Di br omo ch 1 or ome th an e
1,1,2,2-Tetrachloroethylene
Chlorobenzene
Ethyl benzene
Bromoform
p-xylene
Styrene
1,1,2,2-Tetrachloroethane
Bromobenzene
n-propyl benzene
Pentachloroethane
m-Oi ch 1 or ob enz en e
p-Oi Chlorobenzene
o-Di ch 1 or ob enz en e
1 ,2-Dibromo-3-chloropropane
Hexa chl or oeth ane
1,3,5-Trichlorobenzene
1, 2, 4-Tri Chlorobenzene
Naphthalene
1, 2, 3-Tri Chlorobenzene
N
5
5
5
5
5
5
5
5
5
5
5
5
5
5.
2b
5
5
5
5
5
2C
5
5
5
5
5
5
5
5.
2d
5
5
5
5
5
5
5
5
5
Average
ug/l
9.05
35.6
8.66
8.51
9.26
3.99
48.6
9.60
15.8
2.54
27.2
6.59
57.2
12.2
12.2
6.74
6.57
14.0
2.36
65.7
16.1
4.87
2.29
53.3
2.39
2.27
24.6
4.78
2.26
38.1
4.63
4.59
4.71
24.7
22.8
9.67
9.23
4.93
10.1
SO
0.82
1.9
0.18
0.72
0.40
1.9
1.4
0.18
0.55
0.08
1.5
0.17
1.2
0.75
—
0.44
0.40
0.95
0.11
1.8
—
0.33
0,04
0.03
0.03
0.08
0.40
0.09
0.06
—
0.10
0.15
0.12
0.34
0.50
0.80
0.19
0.12
0.51
RSO
(%)
9.1
5.3
2.1
8.4
4.4
47
3.0
1.9
3.5
3.0
5.3
2.6
2.0
6.1
—
6.6
6.1
6.8
4.8
2.8
—
6.8
2.0
0.5
1.3
3.4
1.4
2.0
2.8
—
2.2
3.3
2.5
1.4
2.0
8.3
2.0
2.4
5.1
Recovery
(%)
91
95
87
85
93
40
87
96
90
101
90
88
92
97
97
45
44
93
95
98
92
97
92
96
96
91
98
96
92
95
93
92
94
99
91
97
92
99
101
-49-
-------
Table 10. (Continued)
Compound
Hexachlorobutadiene-1,3
1,2,4 ,5-Tetrach lor ob enzene
1,2,3,4-Tetrachlorobenzene6
l-Chloronaphthalenee
2-Chlorobiphenyle
N
5
5
5
Average
wg/'l
22.4
14.7
SO
0.52
1.2
RSO Recovery
(%) (%)
2.3 90
7.8 98
aAverage of 5 spike day analyses, two each non-preserved, 2 each preserved
with HC1 and one preserved with mercury.
b Aver age of two non-preserved samples 100% loss in HC1 preserved and 60%
recovery in Hg preserved.
cAverage of two HC1 preserved analyses 179% recovery for mercury preserved
sample and 189% recovery for 22 *C non-preserved.
^Average of two HC1 preserved analyses - 39% recovery for non-preserved
samples and 8.5% recovery for Hg preserved sample.
compounds were deleted from the study because variable retention
times caused the data system errors. The error was traced to a faulty oven
temperature controller.
-50-
-------
Table 11. Spiked Quenched Cincinnati Tap Water Stored at 22*C
24 day
Average
Compound Cone. (uq/L)
1,1-Dichloroethylene
Methylene Chloride
Ally! Chloride
Tr an s-1 , 2-Oi ch 1 oro-
eth yl ene
ci s-1, 2-01 ch lor o-
ethylene
Allyl Bromide
Chloroform
1 , 2-ni ch 1 or oeth ane
1,1,.1-trich lor oeth ane
Benzene
Carbon tetrachloride
1 , 2-di ch 1 oro pro p an e
Br omo di ch 1 or one th an e
1,1,2-trichloro-
ethylene
2-Chl or oethyl vinyl
ether
1 , 3-di ch 1 oro pr o pen e
1,3-dichloropropene
1 ,1 , 2-tr i chl oroeth ane
Toluene
01 br omo chl or o-
me thane
1, 1,2, 2-tetrachl oro-
eth yl en e
Chlorobenzene
Ethyl benzene
Br omo form
p-xylene
Styrene
1,1,2,2-tetrachloro-
ethane
Bromobenzene
n-Propylbenzene
P en tach 1 oroeth ane
m-di ch 1 orob enz ene
p-Di ch 1 or ob enz ene
o-Oi Chlorobenzene
Oibr omo chl oro propane
Hexachlor oeth ane
1, 3, 5-tri Chlorobenzene
8.32
34.6
3.82
9.11
9.59
—
51.0
10.1
16.8
2.55
28.2
6.94
58.2
19.4
12.0
1.9
2.6
14.1
2.53
65.6
41.8
4.74
53.9
2.41
2.16
15.6
4.81
2.22
3.36
4.56
4.55
4.73
24.7
20.9
8.37
S.D.
1.20
2.4
2.7
1.44
0.65
—
2.5
0.7
1.3
0.12
2.1
0.47
2.4
3.9
0.3
2.4
2.7
0.5
0.17
2.0
4.7
0.33
1.7
0.13
0.14
4.2
0.51
0.09
5.9
0.15
0.17
0.17
0.8
2.4
0.41
RSD
15
6.8
71
16
6.8
—
4.8
6.7
7.7
4.8
7.3
6.8
4.0
20
2.5
123
106
3.6
6.6
3.0
11.2
6.9
3.1
5.2
6.3
27
3.1
3.9
176
3.3
3.8
3.7
3.3
11
4.8
Study Ave.
Recover (%)
83
92
38
91
96
—
91
101
96
102
93
93
94
155
96
25
34
94
101
97
239
95
97
96
86
62
96
89
8.4
91
91
95
99
83
84
-51-
-------
Table 11. (Con t1nued}
24 day
Average
Compound Cone. (yg/L)
1,2,4-TMchlorobenzene
Naphthalene
1,2,3-Trichlorobenzene
Hexachlorobuta-
diene,-l,3
1,2,4,5-Tetrachloro-
benzene
l,2,3,4-Tetrad'.'1,oro-
benzene
1-Ch 1 or on a ph th al en e
2-Chlorobiphenyl
8.61
4.70
9.16
20.1
12.8
22.7
S.O.
0.5
0.24
0.69
1.5
2.4
4.6
RSO
5.8
5.2
7.5
7.6
19
20
Study Ave.
Recover (%)
86
94
92
81
76
-52-
-------
Table 12. Spiked Quenched Cincinnati Tap Hater Stored at 4*C
Compound
1,1-Oichloroethylene
Methylene Chloide
Ally! Chloride
t-l,2-0ichloroethylene
c-l,2-Dichloroethylene
Ally! Bromide
Chloroform
1,2-Dichloroethane
1,1,1-trichloroethane
Benzene
Carbon Tetrachloride
l,2-d1chloropropane
Bromodichloromethaiie
1,1,2-trichloroethylene
2-chloroethyl vinyl
ether
1 , 3-di ch 1 oropropene
1,3-dichloropropene
1,1,2-trichloroethane
Toluene
01 br omoch 1 or ometh ane
1,1,2,2-tetrachloro-
ethylene
Chlorobenzene
Ethylbenzene
Bromoform
p-Xylene
Styrene
1,1,2,2-tetrachloro-
ethane
Bromobenzene
n-Propyl benzene
Pentachloroethane
m-Dichlorobenzene
p-Oi Chlorobenzene
o-Dichlorobenzene
l,2-Dibromo-3-chloro-
propane
Hexachl oroeth ane
1,3,5-Trichlorobenzene
1,2,4-Trichlorobenzene
Naphthalene
1, 2, 3-Tri Chlorobenzene
Hexachlorobutadiene-1,3
Average
(uq/L)
7.15
32.1
6.75
7.79
8.55
—
44.4
9.21
14.0
2.27
22.7
6.26
54.8
12.0
12.0
4.33
5.53
14.2
2.24
62,8
37.0
4.27
2.08
54.3
2.16
2.15
22.8
4.58
1.97
3.24
4.27
4.31
4.53
25.3
20.7
8.04
8.54
4.86
9.24
18.9
24 Day
SO
1.3
2.5
0.44
1.20
0.43
__
1.9
0.26
0.59
0.07
0.98
0.33
1.5
1.2
0.39
1.3
0.70
0.70
0.13
2.3
2.6
0.35
0.06
1.1
0.05
0.06
1.1
0.09
0.07
4.4
0.11
0.11
0.15
1.3
0.90
0.35
0.33
0.18
0.24
1.1
Study
RSD
%
19
7.7
6.5
15
5.0
—
4.2
2.8
4.2
3.0
4.3
5.2
2.8
10
3.3
29
13
4.7
5.8
3.6
7.0
8.2
2.9
2.0
2.4
2.7
4.6
2.0
3.4
134
2.6
2.5
3.4
5.0
4.4
4.4
3.9
3.7
4.5
5.6
Recovery
%
72
86
68
78
86
__
80
92
80
91
75
83
88
96
96
58
74
95
90
93
211
85
83
96
86
86
91
92
79
8.1
85
86
91
101
83
80
85
97
92
76
Day 18
Recovery %
58
76
61
93
87
. 0
80
94
84
92
78
83
90
109
97
21
34
97
94
99
230
79
81
100
88
88
92
92
86
4
86
86
90
111
86
80
86
103
92
76
Spike Day
Recovery
X
83
91
88
86
91
50
86
95
91
100
88
88
94
98
97
96
92
99
99
100
189
100
96
96
96
93
97
95
91
39
93
93
95
99
91
90
92
98
97
89
Remaining compounds deleted from study because of memory effects.
-53-
-------
Table 13. Method Accuracy and Precision
Spiked Quenched Tap Water * HgCle Stored at 22"c
25 Day
Average
Cone.
Compound ug/L
1,1-Dichloroethylene
Methyl ene Chloride
Ally! Chloride
tr ans-1 , 2-Oi ch 1 oro-
ethylene
cis-l,2-Dichloro-
ethylene
Allyl Bromide
Chloroform
1,2-Oichloroethane
1,1,1-Trichloroethane
Benzene
Carbon Tetrachloride
1,2-Dichloropropane
Bromodichloromethane
1,1,2-Trichloroethylene
2-Chloroethyl vinyl ether
1,3-Oichloropropene
1 ,3-Dichloropropene
1,1,2-Trichloroethane
Toluene
Di br omoch 1 or ometh ane
1 ,1 ,2 ,2-Tetrachloro-
ethylene
Chlorobenzene
Ethylbenzene
Bromoform
p-Xylene
Styrene
1,1,2,2-tetrachloro-
ethane
Bromobenzene
n-Propylbenzene
Pentachloroethane
m-Oi Chlorobenzene
p-Di Chlorobenzene
o-Oichlorobenzene
1 , 2-Oi br omo-3-ch 1 oro-
propane
Hexachloroethane
1, 3, 5-Tri Chlorobenzene
1, 2, 4-Tri Chlorobenzene
Naphthalene
1, 2, 3-Tri Chlorobenzene
Hexachlorobutadiene-1,3
9.06
35.4
3.01
8.28
9.56
—
51.4
17.5
16.9
2.5
27.1
6.93
57.5
20.0
1.13
1.43
2.23
14.1
2.48
64.2
41.0
4.74
2.35
51.9
2.39
2.20
14.5
4.79
2.19
__
4.53
4.52
4.70
25.4
23.4
8.22
8.67
4.89
9.39
20.1
SD
1.29
2.3
3.1
0.81
0.63
—
2.7
4.77
1.34
0.2
1.5
0.58
2.7
3.2
2.3
2.19
2.25
0.6
0.16
1.1
2.0
0.30
0.13
1.7
0.11
0.21
3.4
0.20
0.15
__
0.25
0.27
0.26
0.7
1.5
0.77
0.73
0.33
0.77
?.5
Spike
Day
Recovery
RSD I
14
6.4
104
9.8
6.6
—
5.3
27
7.9
6.6
5.5
8.4
4.7
16
201
153
100
4.4
6.3
1.7
4.9
6.3
5.6
3.2
4.6
9.4
23.1
4.1
6.9
—
5.6
5.9
5.2
2.6
6.5
9.4
8.4
6.8
8.1
12.9
96
97
75
79
96
46
91
110
98
100
91
88
92
133
39
68
73
93
97
97
240
101
95
96
96
95
76
97
95
8.5
97
93
93
102
96
96
96
103
104
95
Study
Recovery
X
91
94
30
83
96
—
92
180
97
100
90
92
93
160
9
19
30
94
99
95
234
95
94
94
96
88
58
96
85
—
91
92
94
102
94
82
87
98
99
80
X
4 Day Recovery
Period Day 18
-25
-100
+29.9
+58
-72
-43
-34
+10.4
+4
-11
-23
89
99
11
75
91
0
97
221
88
97
86
89
90
154
0
0
0
92
98
95
234
91
92
93
95
decay
58
93
82
0
87
87
93
103
88
77
85
103
93
72
Remaining compounds deleted from study because of memory effects,
-54-
-------
Table 14. Method Accuracy and Precision
Spiked Quenched Tap Water and HC1 Stored at 22 *C
25 Day Study
Average
Compound Cone. (u3/D
1 , 1-Di chl oroethyl ene
Methyl ene Chloride
Allyl Chloride
trans-l,2-0ichloro-
ethylene
cis-l,2-di chl oro-
ethyl ene
Allyl Bromide
Chloroform
1,2-Di chl or oe thane
1,1,1-Trichloroe thane
Benzene
Carbon tetrachloride
1,2-01 chl oro pro pane
Br omo d1 ch 1 or one th an e
1,1,2-Tri chl oro-
ethyl ene
2-Chl oroethyl vinyl
ether
1 , 3-01 ch 1 oro pro pen e
1 ,3-01 chl or opr open e
1 ,1 , 2-Tr 1 ch 1 or oeth ane
Toluene
Dibromochloro-
me thane
1 ,1 , 2, 2-Tetrachl oro-
ethyl ene
Chlorobenzene
Ethylbenzene
Br omo form
p-xylene
Styrene
1,1,2,2-Tetrachloro-
e thane
Bromobenzene
n-Propylbenzene
P en tach 1 oroeth ane
nv-Oi Chlorobenzene
p -Di ch 1 or ob enz en e
o-O i ch 1 or ob enz en e
1 , 2-Oi br omo- 3-ch 1 oro-
propane
Hexa chl or oeth ane
1,3,5-Trichlorobenzene
8.29
34.3
3.39
8.47
9.67
—
50.8
9.64
15.2
2.68
27.9
6.74
58.8
11.7
NO
1.63
2.04
14.1
2.37
66.4
14.8
4.76
2.30
54.7
2.39
1.09
25.2
4.76
2.14
39.7
4.44
4.39
4.64
25.4
23.3
7.37
SO
1.25
2.2
2.76
0.98
0.66
—
7.1
0.28
0.63
0.17
2.1
0.25
2.2
0.6
—
2.42
2.33
0.77
0.11
2.0
0.72
0.25
0.91
1.3
0.84
0.66
0.79
0.14
0.10
2.5
0.14
0.13
0.14
1.2
1.2
0.77
RSD
15.1
6.5
81
11.6
6.8
—
14
2.8
4.1
6.3
7.6
3.6
3.7
5.2
—
148
113
5.5
4.7
3.1
4.8
5.3
4.0
2.4
3.5
60
3.1
2.9
4.5
6.3
3.2
3.9
3.1
4.7
5.0
9.7
Study
Recovery (%)
83
92
34
85
97
—
91
96
89
107
92
90
95
94
0.0
22
15
94
95
99
85
95
92
99
95
43
101
99
86
99
89
88
93
102
93
74
-55-
-------
Table 14. (Con tinued)
Compound
25 Day
Average
Cone. (uQ/L)
1,2,4-Trichlorobenzene
Naphthalene
1,2,3-Trlchlorobenzene
8.38
4.82
9.7
Study
SO
0.64
0.27
0.75
RSO
7.7
5.6
7.7
Study
Recovery
84
96
97
(%}
Hexachlorobuta-
d1ene,-l,3 19.2 1.9 9.8 77
1,2,4,5-Tetrachloro-
benzene
1,2,3,4-Tetrachloro-
benzene
1-Chloronaphthal ene
2-Chlorob1phenyl 29.9 3.9 12.9 100
-56-
-------
Table 15. Spiked Cincinnati Tap Water
Recovery
1,1-Dichloroethylene
Methylene Chloride
Allyl Chloride
tr an s-1 , 2-Oi ch 1 oro-
ethylene
cis-l,2-Dichloro-
ethylene
Allyl Bromide
Chloroform
1, 2-Oichl or oe thane
1 ,1 ,1-Trich lor oe thane
Benzene
Carbon tetrachloride
1 , 2-Oi chl oropropane
8r omo di ch 1 or ome th an e
1,1,2-Trichloro-
ethyl en e
2-Chl oroethylv inyl
ether
1 , 3-Di chl oropropene
1 ,3-Oichloropropene
1, 1, 2-Tr i chl or oe thane
Toluene
01 br omo chl or o-
me thane
1,1,2,2-Tetrachloro-
ethylene
Chlorobenzene
Ethyl benzene
Br omo form
p-xylene
Styrene
1,1,2,2-Tetrachloro-
ethane
Bromobenzene
n-Propylbenzene
Pentachloroethane
m-Oi ch 1 or ob enz en e
p-Di Chlorobenzene
o-Di Chlorobenzene
1 , 2-Oi bromo-3-ch 1 oro-
propane
He xachloroe thane
1,3,5-Trichlorobenzene
1, 2, 4-Tri Chlorobenzene
Naphthalene
22*
76
85
19
107
95
0
95
103
94
99
90
90
94
176
94
0
3
91
100
102
255
91
0
99
93
88
50
93
86
0
89
89
95
103
72
79
82
92
4'
58
76
61
93
87
0
80
94
84
92
78
83
90
109
97
21
34
97
94
94
230
79
81
100
88
88
92
92
80
4
86
86
90
111
86
80
86
103
Day 18a
Hg
89
99
11
75
91
0
97
221
88
97
86
89
90
154
0
0
0
92
98
95
234
91
92
93
95
76
58
93
82
0
87
87
93
103
88
77
85
103
Average Study Recovery^
HC1
98
112
13
91
106
0
118
100
92
110
95
94
98
97
0
0
0
95
98
102
81
92
93
102
94
60
106
95
83
102
88
86
94
110
97
70
78
101
22'
83
92
38
91
96
_
91
101
96
102
93
93
94
155
96
25
34
94
101
97
239
95
0
97
96
86
62
96
89
8.4
91
91
95
99
83
84
86
94
4*
72
86
68
78
86
_
80
92
80
91
75
83
88
96
%
58
74
95
90
93
211
85
83
96
86
93
91
92
7-9
8.1
85
86
91
101
83
80
85
97
Hg
91
94
30
83
96
—
92
180
97
100
90
92
93
160
9
19
30
94
99
95
234
95
94
94
96
98
58
96
85
-
91
92
94
102
94
82
87
98
HCl
83
92
34
85
97
—
91
96
89
107
92
90
95
94
_
22
15
94
95
99
85
95
92
99
95
43
101
99
86
99
89
88
93
102
93
74
84
96
-57-
-------
Table 15. (Continued)
Recovery
1,2,3-Trichlorobenzene
Hexachlorobuta-
d1ene,-l,3
22'
87
75
4*
92
76
Day 18A
Hg
93
72
HC1
92
68
Average Study
22*
92
81
4'
92
76
Recovery8
Hg
99
80
HC1
97
77
1,2,4,5-Tetrachloro-
benzene
1,2,3,4-Tetrachloro-
benzene
1-Chloronaphthalene
2-Chlorobiphenyl
aaverage of two analyses performed 18 days after spiking for the 22* sample 17
days after spiking
''average of all analyses performed from spike day to end of study
n - 11 ea 22*C, 8 ea 4 C, 11 ea Hg and 11 ea HC1
-58-
-------
Table 16. Suirrary of Method Recovery Spiked Ohio River Water
1,1-Dichlor oethylene
Methylenechloride
Allyl Chloride
t-1, 2-01 ch lor oethylene
c-1 , 2-01 ch 1 oroe thyl ene
Allyl Bromide
Chloroform
1 , 2-01 ch lor oe thane
1 ,1 , 1-Tr i ch 1 or oeth ane
Benzene
Carbon Tetrachlorlde
1,2-01 chloropropane
Br omo di ch 1 or ome th an e
1,1,2-Trichlor oethylene
1 , 3-01 ch 1 oro propene
1,3-Dichloropropene
1 ,1,2-Tri chl or oeth ane
Toluene
Dibromo chl or ome thane
1,1,2,2-Tetrachloro-
ethylene
Chlorobenzene
Ethylbenzene
Bromoform
p-Xylene
Styrene
1,1,2,2-Tetrachloroethane
Bromobenzene
n-Propyl Benzene
Pentach 1 or oeth ane
m-Oi chlorobenzene
p-Di Chlorobenzene
o-Oi ch 1 or ob enz en e
l,2-Dibromo-3-
chloropropane
He xa chl or oeth ane
1,3,5-Trichlorobenzene
1 , 2 ,4-Tri chl orobenzene
Naphthalene
1, 2, 3-Tri chl orobenzene
Hexachlorobutadiene-1,3
1, 2,4, 5-Tetrachl oro-
benzene
22*
Spike
Day
(*)
90
91
89
92
95
48
87
96
88
95
90
94
94
90
89
94
100
95
99
107
95
91
102
91
94
107
95
92
85
95
96
97
114
93
91
96
107
98
83
93
Ave. Recovery Day 2 Through Day 26
22*
(%)
93
98
56
92
93
—
91
98
91
89
73
98
92
94
46
46
94
79
93
187
91
89
97
84
82
%
85
86
38
92
92
98
104
49
83
91
98
96
76
84
4*
(%)
92
104
84
91
91
15
89
97
97
87
91
97
93
88
77
83
97
80
94
112
90
80
99
79
77
99
83
68
84
90
91
95
98
79
87
93
85
95
79
86
Hg
(%)
95
103
39
87
96
—
85
94
91
92
89
93
89
103
30
28
110
87
93
193
92
87
97
80
13
78
87
83
7
87
87
90
99
88
77
84
98
91
73
79
Sllme-
Trol
(X)
84
97
61
90
95
—
86
97
96
97
91
97
Fused Peaks
Fused Peaks
0
0
88
89
95
210
95
89
100
84
95
51
89
86
0
90
90
94
100
92
81
88
100
92
75
78
HC1
(*)
90
98
44
89
38
—
84
95
91
94
90
95
93
85
31
29
92
89
94
82
92
87
97
80
82
95
87
82
96
86
86
92
98
86
78
85
99
91
66
73
-59-
-------
Table 16. (Continued)
22*
Spike
Day
(%)
Ave.
22*
(t)
Recovery
4'
(X)
Day 2 Throutfi Day 26
Sllme-
Hg Trol
(X) (X)
HC1
(X)
1,2,3,4-Tetrachloro-
benzene 98 93 91 84 88
1-Chloronaphthalene 109 105 97 96 100
2-Chlorobiphenyl 117 109 87 98 102
83
96
96
-60-
-------
Table 17. Spiked Ohio River Water Stored at 22*C
Average
Concentration
ug/L
S.O.
RSO
Recovery
1 ,1-Oichloroethylene
Methylenechloride
Allyl Chloride
t-l,2-0ichloroethylene
c-1 , 2-Oi ch 1 oroethyl ene
Allyl Bromide
Chloroform
1,2-Dichloroethane
1,1,1-Trichloroe thane
Benzene
Carbon Tetrachloride
1,2-01 chl or opropane
Bromodi chl orometh ane
1,1,2-Trichloroethylene
1 , 3-01 chl or o pro pen e
1,3-Dichloropropene
1 ,1 , 2-Tri chl oroeth ane
Toluene
01 bromo chl orometh ane
1,1,2,2-Tetrachloro-
ethylene
Chlorobenzene
Ethylbenzene
Bromoform
p-Xylene
Styrene
1,1,2,2-Tetrachloroethane
Bromobenzene
n-Propyl Benzene
P en tachl oroeth ane
m-Oichlorobenzene
p-Di Chlorobenzene
o-Oi ch 1 or obenz en e
l,2-Dibromo-3-
ch lor opropane
Hexachloroe thane
1,3,5-Trichlorobenzene
1,2,4-Trichlorobenzene
Naphthalene
1,2,3-Trichlorobenzene
Hexachlorobutadiene-1,3
1 , 2 ,4 ,5-Tetrachl oro-
b enz ene
1,2,3,4-Tetrachloro-
b enz ene
1-Ch 1 or on a ph th al en e
2-Chlorobiphenyl
18.6
73.3
11.1
18.4
18.7
-
54.6
19.6
32.0 ,
4.45
44.5
14.7
64.7
23.6
6.9
6.9
28.2
3.93
83.9
65.4
9.04
4.46
96.7
4.19
4.08
48.0
8.45
4.3
30.7
9.19
9.19
9.79
51.8
24.3
18.0
18.2
9.78
19.1
38.1
25.1
27.9
31.4
32.6
3.2
11.
5.4
1.2
1.3
-
2.8
0.7
2.7
0.86
16.8
0.4
5.8
1.4
5.1
5.5
2.0
1.29
7.3
17.2
1.2
0.'35
9.4
0.41
1.04
4.93
1.7
0.44
25
0.54
0.57
0.57
5.9
19
1.7
1.6
1.4
1.2
6.0
4.0
3.46
5.9
4.6
17
15
49
6.5
6.8
-
5.1
3.7
8.4
19.4
38
2.9
9.0
5.9
74
79
7.2
33
8.7
26
14
7.9
8.9
9.8
26
10
20
10.3
80
5.9
6.2
5.8
12
77
9.6
8.7
14
6.3
16
16
12.4
18.6
14.1
93
98
56
92
93
-
91
98
91
89
73
98
92
94
46
46
94
79
93
187
91
89
97
84
82
96
85
86
38
92
92
98
104
49
90
91
98
96
76
84
93
105
109
-61-
-------
Table 18. Spiked Ohio River Water Stored at 4*CA
Method
Ave.
Concentration
1,1-Oi chl or oethyl ene
Methylenechloride
Allyl Chloride
t-1 ,2-01 ch lor oethyl ene
c-1 , 2-Di ch 1 or oethyl ene
Allyl Bromide
Chloroform
1,2-Oich lor oeth ane
1 ,1 ,1-Tri chl oroeth ane
Benzene
Carbon Tetrachloride
1,2-Dichloropropane
Br omo di ch 1 or ome thane
1,1,2-Trichloroethylene
1 , 3-01 ch 1 or o pro pen e
1 ,3-Dichloropropene
1 ,1 , 2-Tr i ch 1 or oeth ane
Toluene
01 br omo chl or ome th ane
1,1 , 2, 2=Tetrach1 oro-
eth yl ene
Chlorobenzene
Ethyl benzene
Bromoform
p-Xylene
Styrene
1,1,2,2-Tetrachloroethane
Bromobenzene
n-Propyl Benzene
Pen tachlor oeth ane
m-Oi ch 1 or ob enz en e
p-Oi Chlorobenzene
o-Oi ch lor obenz ene
l,2-Dibromo-3-
chloropropane
He xa ch 1 or oeth an e
1,3,5-Trichlorobenzene
1, 2, 4-Tri Chlorobenzene
Naphthalene
1,2,3-Trichlorobenzene
Hexachlorobutadiene-1,3
1,2,4,5-Tetrachloro-
benzene
ug/l
18.4
78.1
16.7
18.2
18.2
3.0
53.5
19.3
33.8
4.36
54.9
14.5
64.8
22.0
11.6
12.4
29.2
4.02
84.8
39.0
8.98
4.02
98.8
3.94
3.87
49.3
8.30
3.38
67.3
9.04
9.06
9.49
49.2
39.5
17.5
18.5
8.51
19.0
39.5
25.7
S.O.
2.1
4.7
3.4
0.7
1.3
5.4
3.0
0.54
1.5
0.51
3.4
0.3
2.6
0.92
2.1
2.3
2.2
0.86
4.2
3.2
0.72
0.63
4.4
0.32
1.02
2.5
0.79
1.14
4.3
0.28
0.29
0.35
6.05
9.2
0.94
0.78
1.9
0.56
3.2
2.3
RSO
U)
11
6.0
20
6.7
7.2
177
5.6
2.8
4.5
12
6.2
1.9
4.0
4.2
18.3
18.2
7.5
21
4.9
8.1
8.0
16
4.4
8.2
26
5.1
9.6
34
6.4
3.1
3.2
3.9
12
23
5.4
4.2
22
3.0
8.2
9.0
Average
Recovery
(*)
92
104
84
91
91
15
89
97
97
87
91
97
93
88
77
83
97
80
94
112
90
80
99
79
77
99
83
68
84
90
91
95
98
79
87
93
85
95
79
86
-62-
-------
Table 18. (Continued)
1 , 2,3 ,4-Tetrach 1 oro-
benzene
1-Ch 1 or onaph th al en e
2-Chlorobiphenyl
Method
Ave.
Concentration
ug/L
27.3
29.2
26.1
S.O.
2.2
3.2
7.9
RSD
(X)
8.2
11.1
30
Average
Recovery
(X)
91
97
87
No spike day analysis performed data include spike day +3 through Spike
day +26
Number of analyses » 7
-63-
-------
Table 19. Spiked Ohio River Water Preserved with Mercury
(3/15 to 4/7)
Method Ave.
Concentration S.O.
1,1-01 chl or oethylene
Methylenechloride
Allyl Chloride
t-l,2-Dich lor oethylene
c-1 , 2-01 chl or oethylene
Allyl Bromide
Chloroform
1, 2-01 ch lor oeth an e
1 ,1 ,1-Trichloroe thane
Benzene
Carbon Tetrachloride
1, 2-01 chloro pro pane
Br onodi ch 1 or one th an e
1,1,2-Tri chl or oethylene
1,3-Dichloropropene
1 , 3-01 ch 1 or o pro pen e
1 ,1 ,2-Trichloroe thane
Toluene
01 br one ch 1 or one thane
1,1,2,2-Tetrachloro-
ethylene
Chlorobenzene
Ethyl benzene
Bromoform
p-Xylene
Styrene
1,1,2,2-Tetrachloroethane
Bromobenzene
n-Propyl Benzene
P en tach 1 or oeth ane
m-Oi Chlorobenzene
p-Oi ch 1 or ob enz en e
o-Oi chl orobenzene
l,2-Dibromo-3-
chloropropane
He xa ch 1 or oeth an e
1,3,5-Trichlorobenzene
1 , 2, 4-Tri chl orobenzene
Naphthalene
1, 2, 3-Tri chl orobenzene
Hexachlorobutadiene-1,3
1 , 2 ,4 , 5-Tetr a ch 1 oro-
benzene
1, 2,3, 4-Tetrachl oro-
benzene
1-Ch 1 or onaph th al ene
2-Chlorobiphenyl
ug/L
18.9
77.0
7.7
17.3
19.2
50.8
18.8
31.7
4.58
54.2
14.0
62.2
25.6
4.45
4.16
27.5
4.35
83.6
67.5
9.17
4.33
96.9
4.02
0.66
38.8
8.70
4.13
5.39
8.73
8.65
8.95
49.2
43.9
15.5
16.9
9.80
18.2
35.3
23.6
25.2
23. 7
29.4
2.6
11.2
4.4
1.5
2.2
4.0
0.9
2.4
0.28
4.4
0.8
3.6
2.3
4.0
4.0
1.8
0.34
5.5
3.2
0.88
0.31
4.2
0.24
0.52
7.1
0.45
0.37
6.9
0.52
0.50
0.54
4.6
3.2
1.7
1.4
0.60
1.3
5.2
4.2
2.9
2.4
3.7
RSO
(%)
14
15
57
8.9
12
7.9
4.9
7.5
6.1
8.1
6.0
5.7
8.9
89
96
6.7
7.8
6.5
4.7
9.6
7.1
4.3
6.0
80
18
5.1
8.9
128
5.9
5.8
6.1
9.3
7.4
11
8.5
6.1
6.9
14
18
11
8.5
12
Average
Recovery
(%)
95
103
39
87
96
85
94
91
92
89
93
89
103
36
28
110
87
93
193
92
87
97
80
13
78
87
83
7
87
87
90
99
88
77
84
98
91
73
79
84
96
98
-64-
-------
Table 20. Spiked Ohio River Water Preserved
with Sl1me-Tro1 RX-34
Method Ave.
Concentration S. 0.
1 , 1-01 chl or oethyl ene
Methylenechloride
Ally! Chloride
t-1, 2-01 chl or oethyl ene
c-l,2-Di chl or oethyl ene
Allyl Bromide
Chloroform
1, 2-0 1 ch 1 or oe thane
1 ,1 ,1-Tri chl oroeth ane
Benzene
Carbon Tetrachloride
1,2-Dichloropropane
Br omodi ch 1 orometh ane
1,1,2-Trichloroethylene
1,3-Dichloropropene
1 ,3-01 chl or opr open e
1 ,1 , 2-Tri chl oroeth ane
Toluene
Oi br omo ch 1 or ome th an e
1 ,1 ,2, 2-Tetr ach loro-
ethylene
Chlorobenzene
Ethylbenzene
Bromoform
p-Xylene
Styrene
1,1 , 2, 2-Te tr ach 1 oroeth ane
Bromobenzene
n-Propyl Benzene
P en ta chl oroeth ane
m-Oi ch 1 or ob enz en e
p-Oi Chlorobenzene
o-Oi ch 1 or ob enz en e
l,2-Oibromo-3-
chloropropane
He xa chl oroeth ane
1,3,5-Trichlorobenzene
1, 2, 4-Tri Chlorobenzene
Naphthalene
1 , 2, 3-Tri ch lorobenzene
Hexach lorobutadi ene-1 ,3
1, 2,3, 4-Tetrach loro-
benzene
1,2, 3, 4-Tetrach loro-
benzene
1-Ch 1 or ona ph th al en e
2-Chlorobiphenyl
ug/l
16.8
72.7
12.2
18.0
19.1
0
51.4
19.5
33.5
4.85
55.3
14.5
Fused
Fused
0
0
26.4
4.44
85.6
73.4
9.48
4.46
100
4.18
4.76
25.3
8.9
4.3
0
9.03
9.01
9.42
50.1
46.2
16.2
17.5
10.1
18.5
37.4
23.4
26.3
30.0
30.6
1.1
9.1
9.0
1.4
1.1
_
3.0
0.9
3.4
0.27
3.9
0.7
Peaks
Peaks
-
-
1.4
0.26
4.1
5.7
0.67
0.34
5.4
0.22
0.21
15.1
0.36
0.36
—
0.66
0.66
0.66
7.0
3.9
1.4
1.2
0.61
1.0
4.8
2.9
2.7
3.0
4.3
RSO
1%)
6.7
1.3
73
7.8
5.7
_
5.9
4.7
10
5.5
7.1
4.5
-
—
5.2
5.9
4.8
7.8
7.1
7.7
5.4
5.3
4.59
60
4.0
8.5
-
7.4
7.4
7.0
14
8.4
8.8
6.7
6.1
5.6
12.8
12.5
10.1
9.9
14.0
Average
Recovery
(%)
84
97
61
90
95
0
86
97
96
97
91
97
0
0
88
89
95
210
95
89
100
84
95
51
89
86
0
90
90
94
100
92
81
88
100
92
75
78
88
100
102
-65-
-------
Table 21. Spiked Ohio River Water Preserved with HC1
Method Ave.
Concentration
I,l-01ch1oroethy1ene
Methylenechloride
Allyl Chloride
t-l,2-0ich1oroethylene
c-l,2-01chloroethylene
Allyl Bromide
Chloroform
1, 2-Dichl or oe thane
1 ,1 ,1-Trichloroethane
Benzene
Carbon Tetrachloride
1 ,2-Di chl oropropane
Rrorodi chlorome thane
i ,1 , 2-Tri chl oroethyl ene
1 ,3-Dichloropropene
1 , 3-Di chl oropropene
1 ,1 ,2-Trichloroethane
Toluene
01 br orao ch 1 or ome th an e
1 ,1 , 2, 2-Tetrachl oro-
ethyl ene
Chlorobenzene
Ethylbenzene
Bromoform
p-Xylene
Styrene
1,1,2,2-Tetrachloroethane
Bromobenzene
n-Propyl Benzene
Pen ta chl or oe thane
m-Oi ch 1 or ob enz en e
p-Di Chlorobenzene
o-Oi chlorobenzene
l,2-Dibromo-3-
chloropropane
Hexa chl or oe thane
1,3,5-Trichlorobenzene
1 , 2 ,4-Tri chl or ob enz ene
Naphthalene
1, 2, 3-Tri chlorobenzene
Hexachlorobutadiene-1 ,3
1 , 2,3 ,4-Tetrach 1 oro-
benzene
1 , 2, 3 , 4-Te tr ach 1 oro-
benzene
1-Ch 1 or onaph th al ene
2-Chlorobiphenyl
wg/L
18.1
73.1
8.84
17.8
17.5
0
50.2
19.0
31.8
4.69
54.4
14.2
65.0
21.2
4.7
4.40
27.5
4.47
84.3
28.5
9.19
4.37
97.3
3.99
4.11
47.5
8.73
4.11
78.9
8.64
8.63
9.17
49.0
42.8
15.6
17.0
9.85
18.1
32.8
22.0
25.0
28.9
28.7
S.O.
1.5
8.1
4.1
1.7
1.1
-
2.8
1.0
3.0
0.40
3.4
0.91
5.0
1.6
4.0
3.9
1.1
0.23
2.5
2.7
1.1
0.34
4.0
0.23
0.55
1.2
0.47
0.36
3.9
0.65
0.52
0.51
3.8
3.0
1.6
1.6
0.45
1.34
3.1
2.7
2.7
1.9
4.3
RSO
(%)
8.4
11
47
9.5
6.3
-
5.6
5.3
9.6
8.6
6.2
5.4
7.7
7.5
85
88
4.0
5.2
3.0
9.4
11
7.8
4.1
5.7
13.4
2.5
5.4
8.76
5.1
7.5
6.0
5.6
7.7
7.1
10
9.4
4.57
7.39
9.4
12
11
6.4
15
Average
Recoverv
(%)
90
98
44
89
38
0
84
95
91
94
90
95
93
85
Ji
29
92
89
94
82
92
87
97
80
82
95
87
82
96
86
86
92
98
86
78
85
99
91
66
73
83
96
96
-66-
-------
Table 22. Method Detection Limit Study
Spike Average
Cone. Concentration Sd
wg/L
I,l-01ch1oroethylene
Me thy 1 enech 1 or 1 de
Ally! Chloride
t-1 , 2-Di ch 1 or oethyl ene
c-l,2-01ch1oroethylene
Ally! Bromide
Chloroform
1, 2-01 chl or oe thane
1 ,1 ,l-Tr1chloroethane
Benzene
Carbon Tetrachloride
1 , 2-01 chl or o pro pane
omodi chlorome thane
1,1,2-Tri chl or oethyl ene
2-Chlor oethyl v inyl ether
1 ,3-Dichloropropene
1 , 3-01 chl oropropene
1,1,2-Trichloroethane
Toluene
D 1 br omo ch 1 or ome th an e
1 ,1 , 2, 2-Tetrachl oro-
ethylene
Chlorobenzene
Ethyl benzene
Bromoform
p-Xylene
Styrene
1,1,2,2-Tetrachloroethane
Bromobenzene
n-Propyl Benzene
Pen tachloroe thane
m-Oi Chlorobenzene
p-Oi Chlorobenzene
o-Oi ch 1 or obenzene
1 , 2-01 bromo-3-
chloropropane
Hexachloroe thane
1, 3, 5-Trichlor obenzene
1 , 2 ,4-Tr 1 ch 1 or obenz ene
Naphthalene
1 , 2, 3-Tr 1 ch 1 or obenz ene
Hexachlorobutadiene-1,3
1,2,3,4-Tetrachloro-
benzene
0.5
1.88
0.5
0.5
0.5
0.5
1.50
0.50
0.88
0.125
1.52
0.375
1.75
0.63
0.38
0.75
0.125
2.25
0.88
0.25
0.125
2.49
.125
.125
1.25
0.25
0.125
2.0
0.25
0.25
0.25
1.25
1.25
0.50
0.50
0.25
0.50
1.25
0.76
ug/L
1.01
1.50
0.34
2.02
0.41
2.88
0.56
0.81
0.35
1.43
0.34
1.47
0.55
0.24
0.66
0.21
2.18
1.72
0.14
0.15
2.13
0.18
0.12
1.61
0.38
0.15
1.56
0.27
0.23
0.25
1.40
1.38
0.56
0.52
0.30
1.29
1.23
0.40
0.62
0.099
1.53
0.085
0.26
0.041
0.087
0.029
0.41
0.013
0.081
0.029
0.035
0.035
0.031
0.29
0.37
0.0073
0.022
0.16
0.054
0.011
0.28
0.081
0.012
0.39
0.060
0.039
0.053
0.17
0.11
0.055
0.031
0.021
0.071
0.25
RSO
I
39
41
29
76
21
9.0
7.4
11
8.2
28
3.8
5.5
5.3
15
5.3
15
13
22
5.3
15
7.5
30
9.3
17
21
9.4
25
22
17
21
13
7.7
9.8
6.0
6.9
5.5
20
Percent
MDL Recovery
wg/L
1.2
1.8
0.31
4.8
0.27
0.81
0.13
0.27
0.09
1.3
0.04
0.25
0.09
0.12
0.11
0.096
0.91
1.2
0.03
0.069
0.48
0.17
0.035
0.88
0.26
0.043
1.2
0.19
0.12
0.17
0.59
0.34
0.17
0.097
0.065
0.11
0.78
(*)
248
80
68
404
82
190
112
92
280
94
91
84
87
63
88
168
97
195
56
120
86
144
96
128
152
120
78
108
92
100
112
110
112
104
120
103
162
-67-
-------
Table 22. (Continued)
Spike Average Percent
Cone. Concentration Sd RSO MDL Recovery
ug/L ug/L X ug/L (%)
1,2,3,4-Tetrachloro-
benzene 0.75 0.92 0.12 13 0.39 123
1-Ch 1 or onaph th al ene
2-Chlorobiphenyl 0.75 1.19 0.21 18 0.67 154
-68-
-------
• — „ .,. _ ,J, .
Valve Oven 125*C
,1-Dichloroethylene
tethylene Chloride
illvl Chloride
Analysis
#1 (%)
0.0
0.0
0,0
Analysis
#2 (%)
0.0
0.0
0.0
Valve Oven 200*C
Analysis
#1 (%)
0.0
0.0
0.0
Analysis
#2 (%)
0.0
0.0
0.0
Valve Oven 125 *C
Memory After
Purging Device
Exchange (%)
0.0
0.0
0.0
trans-l,2-Dichloro-
ethylene
cis-l,2-Dichloro-
ethylene
Allyl Bromide
Chloroform
1,2-Oichl or oe thane
1,1,1-Trichloroethane
Benzene
Carbon tetrachloride
1,2-Oichl oro propane
Bromodichlorome thane
1,1,2-Trichloro-
ethylene
2-Chloroethyviny!
ether
1,3-Di ch 1 oropropene
1,3-Oichloropropene
1,1,2-Trichloroethane
Toluene
Dibromochloro-
me thane
1,1,2,2-Tetrachloro-
ethylene
Chlorobenzene
Ethylbenzene
Bromoform
p-xylene
0.0
0.0
0.0
0.0
0.0
0.0
1.1
0.0
1.4
0.0
1.0
1.0
1.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0 .
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
-69-
-------
Table 23. System Memory (Continued)
Valve Oven 125*C Valve Oven 200*C Valve Oven 125*C
Analysis
Analysis
#1 (%)_
Analysis Analysis
Memory After
Purging Device
Exchange (%)
Styrene
1,1,2,2-Tetrachloro-
ethane
Bromobenzene
n-Propylbenzene
Pentachloroethane
m-01chlorobenzene
p-01ch1orobenzene
o-Oichlorobenzene
1,2-Dibromo-3-cloro-
propene
Hexachloroethane
1,3,5-Trlchlorobenzene
1,2,4-Trlchlorobenzene
Naphthalene
1,2,3-Trlchlorobenzene
Hexachlorobuta-
d1ene,-l,3
1,2,4,5-Tetrachloro-
benzene
1,2,3,4-Tetraehloro-
benzene
1-Chloronaphthalene
2-Chlorobiphenyl
1.0
0.0
0.0
0.0
1.76
1.7
2.3
2.3
2.5
2.6
2.9
6.0
2.9
5.3
7.3
8.4
9.4
9.9
27
48
48
92
0.0
0.0
0.0
0.0
0.0
0.0
0.0
2.5
1.2
6.2
4.6
4.9
5.8
4.3
12
30
30
69
0.0
0.0
0.0
0.0
0.0
0.0
2.2
0.0
0.0
1.0
2.0
2.2
1.5
1.3
3.0
8.0
8.0
30
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
r.o
0.9
0.8
0.0
0.0
0.6
2.6
2.6
7.0
0.0
0.0
1.5
1.5
1.4
1.4
1.4
1.4
0.0
0.0
1.3
2.5
3.7
4.6
4.1
12
16
26
38
-70-
------- |