Chemosphere 158 (2016) 171e176

Contents lists available at ScienceDirect

Chemosphere
journal homepage: www.elsevier.com/locate/chemosphere

Lowering detection limits for 1,2,3-trichloropropane in water using

solid phase extraction coupled to purge and trap sample introduction
in an isotope dilution GC-MS method
Wenta Liao, Miriam Ghabour 1, William M. Draper*, Esala Chandrasena
Drinking Water and Radiation Laboratory Branch, Center for Environmental Health, California Department of Public Health, 850 Marina Bay Parkway,
Richmond, CA, 94804, United States

h i g h l i g h t s

Combining SPE with PTI significantly reduces detection limits for 1,2,3-trichloropropane in water.

Tandem SPE-PTI achieves a 500 mL (equivalent) sample introduction using conventional drinking water laboratory instruments.

The method may be applicable to other carcinogenic VOCs for which there is also interest to lower detection limits.

a r t i c l e i n f o a b s t r a c t

Article history: Purge and trap sample introduction (PTI) has been the premier sampling and preconcentration technique
Received 4 January 2016 for gas chromatographic determination of volatile organic compounds (VOCs) in drinking water for
Received in revised form almost 50 years. PTI affords sub parts-per-billion (ppb) detection limits for purgeable VOCs including
9 May 2016
fixed gases and higher boiling hydrocarbons and halocarbons. In this study the coupling of solid phase
Accepted 21 May 2016
Available online 2 June 2016
extraction (SPE) to PTI was investigated as a means to substantially increase enrichment and lower
detection limits for the emerging contaminant, 1,2,3-trichloropropane (TCP). Water samples (500 mL)
Handling Editor: Andreas Sjodin were dechlorinated, preserved with a biocide, and spiked with the isotope labeled internal standard, d5-
TCP. The entire 500 mL sample was extracted with activated carbon or carbon molecular sieve SPE
Keywords: cartridges, and then eluted with dichloromethane – excess solvent was removed in a nitrogen evaporator
Volatile organic compounds and diethylene glycol “keeper” remaining was dispersed in 5 mL of water for PTI GC-MS analysis. The
Ultratrace water analysis experimental Method Detection Limit (MDL) for TCP was 0.11 ng/L (ppt) and accuracy was 95e103% in
Tandem sample preparation and sub-ppt determinations. Groundwater samples including impaired California sources and treated water
introduction
(n ¼ 21) were analyzed with results ranging from below the method reporting limit (0.30 ng/L)
Solid phase extraction
to > 250 ng/L. Coupling of SPE with PTI may provide similar reductions in detection limits for other VOCs
Purge and trap
Isotope dilution GC-MS with appropriate physical-chemical properties.
1,2,3-trichloropropane © 2016 Elsevier Ltd. All rights reserved.

1. Introduction 2003). About 80% of TCP produced originates from epichlorohy-

1,2,3-Trichloropropane (TCP) is a groundwater contaminant in
the U.S. and other industrialized countries. TCP is exclusively man-
made and has been used as an industrial solvent, a cleaning/
degreasing agent and paint remover, and a synthetic intermediate,
for example, a crosslinking agent in polymer production (WHO,
* Corresponding author.
E-mail address: william.draper@cdph.ca.gov (W.M. Draper).
1
Current address: Metropolitan Water District of Southern California, Water
Quality Laboratory, 700 Moreno Avenue, La Verne, CA, 91750.
drin manufacture. TCP also is an impurity in soil fumigants
including dichloropropane/dichloropropene mixtures (DD; telone),
some of which have been used extensively in California.
Due to its physical-chemical properties TCP is unlikely to sorb to
soil and leaches readily into groundwater. TCP also is persistent in
natural waters with a hydrolysis half-life of 44 years at pHs be-
tween 5 and 9 (ATSDR, 1992). Regulatory monitoring in California
has identified TCP contamination in 372 active and standby
drinking water sources (SWRCB, 2015), principally in Kern, Fresno,
Los Angeles and Merced Counties. TCP is considered a contaminant
of emerging concern due to mounting evidence of carcinogenicity

http://dx.doi.org/10.1016/j.chemosphere.2016.05.061
0045-6535/© 2016 Elsevier Ltd. All rights reserved.
172 W. Liao et al. / Chemosphere 158 (2016) 171e176
in experimental animals (IARC, 1995; ATSDR, 2011) and anticipated
carcinogenicity in humans (NTP , 2014).
TCP is a purgeable VOC which can be determined by purge and
trap (PTI) gas chromatography (GC) and PTI gas chromatography-
mass spectrometry (GC-MS) using methods such as USEPA 502.2
and USEPA 524.2, respectively e both of which are approved for
drinking water compliance monitoring of regulated VOC. While TCP
is not regulated by the federal government in the U. S. per se, Cal-
ifornia established a drinking water notification level (NL) of 5 ng/L
(ppt) for TCP in 1999. In 2002, our laboratory described two
analytical methods capable of achieving the state’s required
compliance monitoring reporting limit, also 5 ng/L (CDPH, 2002a;
2002b). More recently a revised public health goal (PHG) of
0.7 ng/L was established for TCP (OEHHA, 2009).
This paper describes further research on trace determination of
TCP in finished drinking water and drinking water sources. Spe-
cifically, we investigated reducing detection limits by coupling SPE
with PTI, a novel approach to substantially increase VOC enrich-
ment and preconcentration. Preliminary experiments examined
feasibility, and further experiments established the basic operating
conditions for determination of TCP using conventional water
laboratory instrumentation. Isotope dilution quantitation using the
internal standard, d5-TCP, provided optimal precision and accuracy.
Finally, method performance was evaluated by analyzing California
groundwater samples from sites with historic TCP contamination.
2. Materials & methods
2.1. Chemicals/supplies
All solvents were obtained from Fisher Scientific (Pittsburgh, PA,
USA) and were glass distilled and suitable for GC or HPLC applica-
tions. A Millipore Milli-Q-Plus water purification system (Billerica,
MA, USA) supplied laboratory reagent water for all experiments.
Analytical standards of TCP (5000 mg/mL in methanol) and Supelco
Supelclean ENVI-Carb Plus cartridges (400 mg, P/N 54812-U) were
obtained from Supelco (Bellefonte, PA, USA). The deuterium-
labeled internal standard (IS), d5-TCP, was obtained from Cam-
bridge Isotope Laboratories (Tewksbury, MA, USA). Waters AC-2
Sep-Pak cartridges (400 mg, P/N JJAN20229) were obtained from
Waters (Milford, MA, USA).
2.2. Instruments
An Agilent 5973 MSD mass spectrometer interfaced to an Agi-
lent 6890 gas chromatograph (Agilent Instruments, Wilmington,
DE, USA) was used. The GC-MS was equipped with a Tekmar 3000
automated purge and trap sampler (Teledyne Tekmar, Mason, OH,
USA). The mass spectrometer was operated in electron ionization
(EI) mode with an ionization energy of 70 eV. The instrument was
autotuned and electron multiplier (EM) voltage was increased by
300 V to increase signal. In addition, selected ion monitoring (SIM)
was used to lower detection limits. Two ions were monitored:
C3H4Clþ (m/z 75, 10 ms dwell time) and C3D4Clþ (m/z 79, 5 ms dwell
time). A J & W DB-5.625 column (30 m  0.25 mm, 1 mm)(J & W
Scientific, Folsom, CA, USA) was used with the following GC oper-
ating conditions: split interface split ratio/temp. (20:1/200  C);
helium carrier gas (0.90 mL min1); oven temp. program (35  C,
5 min; 4  C min1 to 100  C; 30  C min1 to 170  C); and transfer
line (220  C). The purge and trap sampler operating conditions
were: sample volume (5 mL); purge vessel (25 mL fritted sparger);
purge time/temp./He flow rate (11 min/40  C/20 mL min1); dry
purge (2 min). A Tenax trap was used: desorb preheat (245  C);
desorb (4 min at 250  C); desorb gas (20 mL min1 He); bake out
(8 min at 270  C). The retention times were: d5-TCP (20.35 min) and
TCP (20.56 min) and a filament delay of 19 min was used.
2.3. Analytical procedure
After preliminary experiments to identify the optimal sample
preparation conditions, the following procedure was adopted.
Samples were extracted manually using ENVI-Carb Plus SPE car-
tridges mounted on an extraction manifold. The SPE cartridges
were conditioned by washing with 3 mL of dichloromethane, 2 mL
of methanol, and a second aliquot of 2 mL of methanol followed by
2 mL of laboratory reagent water. The sorbent was not allowed to
run dry during conditioning. Water samples (500 mL) were forti-
fied with 10 ng/L of the IS, d5-TCP, by transferring 10 mL of a 500 ng/
mL methanolic solution. Water samples were then loaded onto the
cartridge with a flow rate of ~10 mL/min. The receiver was removed
and air was drawn through the cartridge for 2 min. To elute
retained analytes, the cartridges were inverted, the cartridge
reservoir was filled with methylene chloride, and the sorbent was
soaked in the solvent for ~1 min without vacuum. Methylene
chloride was then eluted dropwise under a low vacuum. Methylene
chloride was dried by percolation through 0.4 g of anhydrous so-
dium sulfate in a small column and the extract was combined with
50 mL of diethylene glycol-methylene chloride (1:1, v/v). The sample
was concentrated under a gentle stream of nitrogen in a room
temperature water bath to a final volume of 50e75 mL prior to
addition of 5 mL of laboratory reagent water. The sample was
vortexed to disperse the solvent “keeper” thoroughly prior to
transfer of the sample to the purge vessel.
3. Results
3.1. Performance of existing methods
A number of instrumental methods have been used to deter-
mine TCP in drinking water including PTI GC (e.g., USEPA 502.2
with electron capture detector (ECD)) and PTI GC-MS (e.g., USEPA
524.2 and USEPA 524.3). In addition, enrichment and preconcen-
tration of TCP has been accomplished by solvent extraction with
either hexane (USEPA 504.1 with ECD-GC analysis) or MTBE (USEPA
551.1). MDLs for TCP vary considerably among these procedures
(Table 1). The MDL is defined as the minimum concentration of a
substance that can be measured and reported with 99% confidence
that the analyte concentration is greater than zero and is deter-
mined by analysis of samples in a given matrix containing the an-
alyte (CFR, 1984). MDLs vary according to many factors both within
and between laboratories (Draper et al., 1998). Because reliable
quantitation is possible only at concentrations above the MDL,
method reporting limits are ideally 3- to 5- times the MDL.
Accordingly, an MDL of 1.7 ng/L has been required to achieve the
current, mandatory TCP reporting limit of 5 ng/L.
Since 2002, CDPH has recommended modified or alternative
methods (CDPH, 2002a; 2002b) with increased detection capability
for TCP regulatory monitoring. USEPA Method 524.2, for example,
was modified by use of SIM in quadrupole instruments and selected
ion storage (SIS) in quadrupole ion trap instruments increasing
sensitivity relative to mass scanning. Other modifications such as
the 25 mL sample volume and detector adjustments further boos-
ted sensitivity. Yoo et al., (2005) lowered reporting limits for TCP
(and 1,4-dioxane) by improving purge efficiency 2-fold with a
20 min purge time, and by use of MS-MS with a 50-fold increase in
detector response. Similar approaches have been used in other
monitoring programs. Miermans et al. (2000) determined TCP with
an MDL of 0.4 ng/L using PTI with a quadrupole ion trap instrument
and a modified sample introduction device, a condenser followed
by liquid nitrogen-cooled, fused silica trap.
 W. Liao et al. / Chemosphere 158 (2016) 171e176
Table 1
TCP method detection limits for various drinking water methods.
Method Spike conc. (ng/L)
524.2 500
524.2 200
504.1 150
504.1 5 5.31 ± 0.31
504.1 2 2.13 ± 0.25
Modified 524.2b 10
Modified 524.2b 5 5.21 ± 0.62
Modified 524.2b 2 2.29 ± 0.53
SPE-PTI (this work) 0.50
SPE-PTI (this work) 0.25
a
Mean ± standard deviation.
b
CDHS, 2002a.
USEPA Method 525.2, a GC-MS method for semivolatile organic
compounds, was modified by employing a specialized, continuous
liquid-liquid extractor (CDPH, 2002b). In part, the reduction in TCP
detection limit was achieved by the 1000-fold concentration factor
e the extract injected onto the instrument was equivalent to 20 mL
of water. The PTI method also used the largest practical sample
volume (25 mL) and both methods employed isotope dilution
quantitation improving accuracy and precision and lowering
detection limits. No method was known that could reliably detect
and quantify TCP at the revised public health goal (PHG) of 0.7 ng/L
established in 2009 using widely available instruments.
A great deal of research has been conducted to develop new
approaches to VOC sample preparation and sample introduction
(Demeestere et al., 2007). While there have been many innovations
since the first studies of chlorination by-products in water treat-
ment (Bellar et al., 1974; Bellar and Lichtenberg, 1974), PTI remains
the most widely used sample preparation method. Solid phase
microextraction (SPME), developed in the early 1990’s (Pawliszyn,
1997), is increasingly important but has not been approved for
drinking water compliance monitoring in the U.S. Initially, we
considered use of large volume sample inlets (LVI) to lower TCP
detection limits, and limited comparative studies with the LVI
demonstrating promise for ultratrace TCP determination (unpub-
lished results). The hypothesis investigated here, however, is that
combining two established sample preparation methods, SPE and
PTI, can greatly increase sample enrichment and preconcentration.
Liter-sized water samples can thus be analyzed by coupling SPE
with PTI, and the 20- to 25-fold increase in enrichment/concen-
tration factor practically achieves a proportionate reduction in
detection/reporting limits.
3.2. TCP mass spectra
The 70 eV EI spectrum of TCP includes a base peak at m/z 75
(C3H4Cl) and other fragment ions including m/z 110 (C3H4Cl2, 35%),
m/z 97 (C2H3Cl2, 22%), m/z 61 (C2H2Cl, 25%) and m/z 49 (CH2Cl, 22%)
but no molecular ion. The corresponding deuterated fragments
appear in the 70 eV EI spectrum of d5-TCP with fragments at masses
m/z 79 (C3D4Cl), m/z 114 (C3D4Cl2), m/z 100 (C2D3Cl2), m/z 63
(C2D2Cl) and m/z 51 (CD2Cl). Chemical ionization (CI) MS also
yielded no molecular ion. Instead, the base peak is m/z 111 ([MH e
HCl]þ or C3H5Clþ 2 ) and 2 minor fragment ions including m/z 75
([MH e 2HCl]þ or C3H4Clþ) and m/z 53 are found (Fig. 1). The SIM CI
MS signal was 5 times greater than in EI MS and calibration curves
were equally acceptable e however, CI MS chromatograms in our
hands were relatively noisy and subject to chemical interferences.
3.3. Instrument calibration
The instrument was calibrated with standards over the
173
TCP conc. found (ng/L)a Accuracy (%) MDL (ng/L)
497 ± 23 99 72
193 ± 18 97 57
124 ± 14 83 45
106 0.89
107 0.72
9.22 ± 0.63 92 1.8
104 1.8
115 1.6
0.51 ± 0.062 102 0.19
0.24 ± 0.034 120 0.11
concentration range of 0.25 ng/L to 500 ng/L. The TCP response
factor was approximately constant across this range with a corre-
lation coefficient of 0.9998 (r2). Calibration curves at the low end of
the overall range are depicted in Fig. 2. CI-MS TCP calibration for the
range of 0.1e25 ng/L was also highly linear (r2 ¼ 0.9998).
3.4. Preliminary experiments
The carbon molecular sieve polymer carbon cartridges extract
relatively high polarity compounds from water, e.g., those with log
P from 1.2 to 1.5. TCP is slightly less polar with log P value of 1.98
(ATSDR, 1992). Because the methylene chloride eluent is not
miscible with water, it was desirable to remove excess solvent.
Thus, diethylene glycol glycol was used as a “keeper” and the
sample concentrated under a gentle stream of nitrogen in a room
temperature water bath. Residual water was previously removed by
drying with anhydrous sodium sulfate.
In a solvent exchange experiment a one mL methylene chloride
extract was combined with 25 mL of diethylene glycol and methy-
lene chloride was removed as described. After concentration, the
30 mL sample was dispersed in 5 mL of laboratory reagent water and
transferred to the purge vessel. The absolute recovery of TCP was
over 80% using the ENVI-Carb Plus SPE cartridges. In contrast, the
Sep Pak AC-2 cartridges gave poor to very poor recoveries of TCP
(e.g., < 20%) in a limited comparative study and, as such, these
cartridges were not investigated further.
Another consideration was whether residual methylene chlo-
ride would exceed the capacity of the Tenax PTI trap or the GC
column. Capillary GC columns with conventional phase thicknesses
are subject to overloading which limits splitless injection to 1- or 2-
mL. An experiment was conducted to address GC column capacity as
follows: 200 mL of dichloromethane was added to 5 mL of labora-
tory reagent water in the purge vessel – the MS filament delay was
adjusted so that the methylene chloride peak eluted prior to data
acquisition. In spite of the large quantities of dichloromethane, the
signal and peak shape for both TCP and the IS were not degraded,
and there was no reduction in TCP recovery. Moreover, there was no
detectable carryover of methylene chloride (or TCP) in subsequent
analytical cycles.
3.5. Method detection limit
The MDL was determined in laboratory reagent water with
spike levels of 0.25 ng/L and 0.5 ng/L (Table 2) according to the
approved procedure (CFR, 1984). TCP recoveries in the MDL study
ranged from 80 to 124% with a mean of 99± 13%. The experimental
MDLs were 0.11 ng/L and 0.19 ng/L for 0.25 and 0.50 ng/L spike
concentrations, respectively (Table 1).
174 W. Liao et al. / Chemosphere 158 (2016) 171e176

Fig. 1. Electron ionization (EI) mass spectrum (left) and chemical ionization (CI) mass spectrum (right) of 1,2,3-trichloropropane.

8.0×10 4 0.8

Response ratio
6.0×10 4 0.6
Response

0.4
4.0×10 4

0.2
2.0×10 4

0.0
0 0.0 0.5 1.0
0 5 10 Concentration ratio
Concentration

Fig. 2. 2A) TCP calibration curve (response vs. concentration) with calibrators of 0.25, 0.5, 1, 5, and 10 ng/L; 2B) isotope dilution calibration using 10 ng/L internal standard
(r2 ¼ 0.9998).

Table 2
Isotope Dilution SPE-PTI eGC-MS Data used to Determine the TCP MDL.

Replicate 0.25 ng/L TCP formulation 0.5 ng/L TCP formulation

TCP conc. found (ng/L) Recovery (%) TCP conc. found (ng/L) Recovery (%)

1 0.29 116 0.46 92

2 0.27 108 0.55 110
3 0.23 92 0.53 106
4 0.21 84 0.62 124
5 0.25 100 0.53 106
6 0.21 84 0.45 90
7 0.20 80 0.46 92
Blank ND N/A ND N/A

3.6. Determination of TCP in groundwater samples
The SPE-PTI technique was further evaluated by analyzing water
samples from sites with historic TCP contamination. In all, 21
groundwater samples from Fresno, Los Angeles, Merced, San Ber-
nardino and San Diego counties were analyzed. Some locations had
treatment systems removing TCP with granular activated carbon
(GAC).
Measured TCP concentrations ranged from below the method
reporting limit (0.30 ng/L) to greater than 250 ng/L (Table 3), and all
results were consistent with historic data. Five samples had TCP
concentrations below the current mandated reporting limit (5 ng/
L), but exceeded the 0.7 ng/L PHG. While there is potential for GC-
MS interference from other VOCs with SIM, for example, iso-
propylbenzene’s m/z 75 fragment ion, which has a similar retention
time to TCP using the DB-5.625 stationary phase (CDPH, 2002a),
these chromatograms were free of interferences or nearby
contaminant peaks. The TCP peak shape in groundwater samples
was satisfactory (Fig. 3) and comparable to that in calibration
standard chromatograms. A 1 ng/L matrix spike demonstrated
acceptable recovery (87%) from the Chino Desalter facility sample,
and analytical repeatability was excellent (e.g., 22.36 ng/L vs.
22.52 ng/L) for two samples from City of Livingston Well #15 cor-
responding to a repeatability of 0.71% RPD. Detectability was suf-
ficient for process monitoring as demonstrated at the Pollock Wells
Treatment Plant where the raw water TCP concentrations are
reduced from 5.82 ng/L to below detection.
4. Discussion
4.1. Why Hasn’t SPE been coupled to PTI previously to lower
detection limits?
While SPE and PTI have each been used to analyze drinking
water for over 40 years (Bellar and Lichtenberg, 1974; Liska, 2000),
the two sample preparation methods have never been used in
 W. Liao et al. / Chemosphere 158 (2016) 171e176 175

Table 3 4.2. Which VOCs are compatible with SPE-PTI?

TCP in California groundwater from sites with historic contamination.

Sample no. Location/sample point TCP conc. (ng/L) Successful application of SPE-PTI depends on the sum total of
Los Angeles Co./Monterey Park
performance at each preparative stage. Thus, it depends upon the
11-3021-01 Well 12 3.06 physical-chemical properties of the analyte, performance of the
11-3021-02 Well 15 3.02 adsorbent, recovery through solvent exchange/concentration and,
11-3021-03 LPGAC com effluent 2.93 finally, efficiency of trapping and release in PTI.
Los Angeles DWP/Pollock Wells Treatment Plant
TCP is moderately polar with a Henry’s law constant of
11-3030-01 Granular activated carbon (GAC) 50% port <0.30
11-3030-02 Granular activated carbon (GAC) effluent <0.30 3.43  104 atm-m3/mol at 25  C. The carbon molecular sieve
11-3030-03 4 ANDG combined raw 5.82 carbon polymer cartridge is widely used for highly polar solutes,
11-3030-04 Well 4 raw 1.99 but also extracts TCP adequately. The TCP Henry’s law constant is
11-3030-05 Well 6 raw 12.1
sufficiently low that water samples can be handled without
San Bernardino Co./Chino Desalting Authority
11-3035-01 Well 9 0.47
appreciable volatilization loss – even when drying SPE cartridges by
San Diego Co./City of Oceanside drawing air through them. For example, the half-life of TCP in water
11-3037-01 Well 2 106 is about an hour at room temperature in still air with bulk mixing
11-3037-02 Well 9 79.7 and water depth of 6.5 cm (Dilling, 1977). VOCs with higher Henry’s
11-3037-03 MBDF 2.53
law constants, such as fixed gases, may not have satisfactory re-
Fresno Co./City of Fresno
11-3041-01 Well head 68.0 coveries using standard SPE extraction devices.
11-3041-02 Well head 45.3 The concentration step, in which methylene chloride is removed
11-3041-03 Well head 45.6 in a nitrogen evaporator, depends upon selective volatilization of
11-3041-04 Well Head 38.4 methylene chloride (vp: 400 mm Hg at 24  C) while preserving TCP
Merced Co./City of Livingston
11-3043-01 Well 15 #1 22.4
(3.1 mm Hg at 25  C). This step must be carried out at room tem-
11-3043-02 Well 15 #2 22.5 perature with a high boiling keeper solvent – some methylene
11-3043-03 Well 14 #1 >250 chloride is left in the sample to improve TCP recovery. Throughout
11-3043-04 Well 14 #2 >250 the procedure the use of the isotope-labeled internal standard is
Merced Co./City of Delhi
critical for accuracy and provides essential quality control for each
11-3042-01 Well 6 #1 40.3
individual sample.
Lastly, PTI efficiency depends on several factors. PTI is optimal
for analytes with boiling points below 200  C which are also
combination. SPE is generally not used for VOC preconcentration
insoluble or poorly soluble in water, e.g., those with moderately
due to the risk of volatilization losses (Demeestere et al., 2007;
high Henry’s law constants (Lee et al., 2013; Staudinger and
Miermans et al., 2000). SPE is, however, used to determine
Roberts, 2001). TCP’s efficient stripping from water is consistent
poorly-purged VOCs such as 1,4-dioxane and tetrahydrofuran
with its Henry’s law constant.
(Isaacson et al., 2006). And SPE is used in the food industry to
Analytical methods with lower detection limits for VOCs are of
determine volatile food constituents, for example, in the analysis of
interest and SPE-PTI may be applicable to those with appropriate
volatile flavor compounds in wine (Wada and Shibamoto, 1997;
physical-chemical properties. Adjustments in SPE-PTI operating
Lopez et al., 2002).
conditions may expand applicability to other VOCs and warrants
Solvent extraction, in contrast, is widely used as a VOC sample
continued investigation.
preparation method (Demeestere et al., 2007). Solvent extraction
has been coupled with PTI for determination of VOCs in environ-
mental analytical chemistry. For example, methanol extraction, 5. Conclusions
dilution into water and PTI is used in hazardous waste testing (e.g.,
USEPA 5030B) with VOC detection limits of <50 mg/kg in waste Coupling SPE with PTI greatly increases sample enrichment and
samples. Soil samples can be combined with water and added preconcentration in instrumental analysis of TCP – here a 500 mL
directly to the purge vessel in a technique known as direct soil water sample (equivalent) introduction was achieved. The absolute
sparging, or methanol extracts of soil can be diluted in water in the recovery of TCP through the preparation method depends upon
purge vessel. good recoveries at each step including SPE, solvent concentration/
exchange and PTI. The same approach is expected to be applicable

Fig. 3. Chromatograms of TCP (ion m/z 75) and deuterated TCP (ion m/z 79) in groundwater samples. 3A) Low concentration groundwater sample (2.2 ng/L). 3B) Medium con-
centration groundwater sample (79.7 ng/L).
176 W. Liao et al. / Chemosphere 158 (2016) 171e176
to many other VOCs with appropriate physical-chemical properties
including polarity, vapor pressure and Henry’s law constant.
Application of tandem SPE-PTI to the isotope dilution quantitation
GC-MS determination of TCP resulted in a greater than 10-fold
reduction in the MDL and provided quantitative data that were
both accurate and precise.
Acknowledgements
We thank Jagdev Dhoot, Rachael Dyda, Mario Estrada, John
Remoy, Donald Wijekoon, and Dadong Xu of the California
Department of Public Health for technical assistance and contri-
butions to this work. David Chang, Sean McCarthy, Jeff O’Keefe, and
Oliver Pacifico of the Division of Drinking Water of the California
State Water Resources Control Board provided groundwater sam-
ples for study. Mention of trade names is incidental and does not
constitute endorsement.
References
ATSDR, 1992. Toxicological Profile for 1,2,3-Trichloropropane. Agency for Toxic
Substances and Disease Registry. http://www.atsdr.cdc.gov/toxprofiles/tp57.
pdf.
ATSDR, 2011. Addendum to the Toxicological Profile for 1,2,3-Trichloropropane.
Agency for Toxic Substances and Disease Registry. http://www.atsdr.cdc.gov/
toxprofiles/1_2_3_trichloropropane_addendum.pdf.
Bellar, T.A., Lichtenberg, J.J., 1974. Determining volatile organics at microgram-per-
litre levels by gas chromatographs. AWWA J. 66, 739e744.
Bellar, T.A., Lichtenberg, J.J., Kroner, R.C., 1974. The occurrence of organohalides in
chlorinated drinking waters. AWWA J. 66, 703e706.
California Department of Public Health, 2002a. Determination of 1,2,3-
Trichloropropane in Drinking Water by Purge and Trap Gas Chromatography/
Mass Spectrometry. http://www.cdph.ca.gov/certlic/drinkingwater/Documents/
Drinkingwaterlabs/TCPbyPT-GCMS.pdf.
California Department of Public Health, 2002b. Determination of 1,2,3-
Trichloropropane in Drinking Water by Continuous Liquid-liquid Extraction
and Gas Chromatography/Mass Spectrometry. http://www.cdph.ca.gov/certlic/
drinkingwater/Documents/Drinkingwaterlabs/TCPbyLLE-GCMS.pdf.
California Office of Environmental Health Hazard Assessment (OEHHA), 2009.
Announcement of Publication of the Final Public Health Goal for 1,2,3-
Trichloropropane in Drinking Water. http://oehha.ca.gov/water/phg/
123tcp082009.html.
California State Water Resources Control Board (SWRCB), Division of Drinking
Water, 2015. 1,2,3-Trichloropropane. http://www.waterboards.ca.gov/drinking_
water/certlic/drinkingwater/123TCP.shtml.
CFR (Code of Federal Regulations) 1984, 40 CFR Part 136, Appendix B to Part 136-
Definition and Procedure for the Determination of the Method Detection
Limit-Revision 1.11, Fed. Reg., 49 FR 43430 (10/26/84), 50 FR 694 (1/4/85), and
51 FR (6/30/86).
Demeestere, K., Dewulf, J., De Witte, B., Van Langenhove, H., 2007. Sample prepa-
ration for the analysis of volatile organic compounds in air and water matrices
(Review). J. Chromatogr. A 1153, 130e144.
Dilling, W.L., 1977. Interphase transfer processes. II. Evaporation rates of chloro-
methanes, ethanes, ethylenes, propanes, and propylenes from dilute aqueous
solutions. Comparison with theoretical predictions. Environ. Sci. Technol. 11,
405e409.
Draper, W.M., Dhoot, J.S., Dhaliwal, J.S., Remoy, J.W., Perera, S.K., Baumann, F.J., 1998.
Detection limits of organic contaminants in drinking water. AWWA J. 90,
82e90.
IARC, 1995. 1,2,3-Trichloropropane. In: dry cleaning, some chlorinated solvents and
other industrial chemicals. In: IARC Monographs on the Evaluation of Carci-
nogenic Risk of Chemicals to Humans, vol. 63. International Agency for
Research on Cancer, Lyon, France.
Isaacson, C., Mohr, T.K., Field, J.A., 2006. Quantitative determination of 1,4-dioxane
and tetrahydrofuran in groundwater by solid phase extraction GC/MS/MS. En-
viron. Sci. Technol. 40, 7305e7311.
Lee, S.-H., Murkherjee, S., Brewer, B., Ryan, R., Yu, H., Gangoda, M., 2013.
A laboratory experiment to measure Henry’s law constants of volatile organic
compounds with a bubble column and a gas chromatography flame ionization
detector (GC-FID). J. Chem. Educ. 90, 495e499.
Liska, I., 2000. Fifty years of solid-phase extraction in water analysis e historical
development and overview (Review). J. Chromatogr. A 885, 3e16.
Lopez, R., Aznar, M., Cacho, J., Ferreira, V., 2002. Determination of minor and trace
volatile compounds in wine by solid-phase extraction and gas chromatography
with mass spectrometric detection. J. Chromatogr. A 966, 167e177.
Miermans, C.J.H., van der Velde, L.E., Frintrop, P.C.M., 2000. Analysis of volatile
organic compounds, using the purge and trap injector coupled to a gas
chromatograph/ion-trap mass spectrometer: review of the results in Dutch
surface water of the Rhine, Meuse, Northern Delta Area and Westerscheldt, over
the period 1992-1997. Chemosphere 40, 39e48.
NTP (National Toxicology Program), 2014. Report on Carcinogens, thirteenth ed.
Research Triangle Park, NC: U.S. Department of Health and Human Services,
Public Health Service.
Pawliszyn, J., 1997. Solid Phase Microextraction: Theory and Practice. Wiley-VCH,
New York.
Staudinger, J., Roberts, P.V., 2001. A critical compilation of Henry’s law constant
temperature dependence relations for organic compounds in dilute aqueous
solutions. Chemosphere 44, 561e576.
Wada, K., Shibamoto, T., 1997. Isolation and identification of volatile compounds
from a wine using solid phase extraction, gas chromatography, and gas chro-
matography/mass spectrometry. J. Agric. Food Chem. 45, 4362e4366.
WHO, 2003. Concise international chemical assessment. In: 1,2,3-Trichloropropane,
vol. 56. World Health Organization, Geneva.
Yoo, L.J., Fitzsimmons, S., Wehner, M., 2005. Analysis of less volatile compounds
using ion trap mass spectrometry coupled with modified purge-trap technique.
Proc. Water Qual. Technol. Conf. 2185e2200.