Journal of Chromatography A, 1157 (2007) 391–398

Gas chromatographic–mass spectrometric determination of alkylphosphonic

acids from aqueous samples by ion-pair solid-phase extraction on
activated charcoal and methylation夽
U.V.R. Vijaya Saradhi ∗ , S. Prabhakar, T. Jagadeshwar Reddy, M.R.V.S. Murty
National Centre for Mass Spectrometry, Indian Institute of Chemical Technology, Tarnaka, Hyderabad, India
Received 11 October 2006; received in revised form 26 April 2007; accepted 26 April 2007
Available online 1 May 2007

Abstract
In the present paper, we report an improved ion-pair solid-phase extraction (IP-SPE) method for the analysis of alkylphosphonic acids, namely,
methyl, ethyl and propylphosphonic acids, present in the aqueous sample. The aqueous sample was mixed with an ion-pair reagent, phenyltrimethy-
lammonium hydroxide (PTMAH) and passed through activated charcoal SPE cartridge. The retained chemicals in the cartridge were extracted
with methanol and analysed by gas chromatography–mass spectrometry (GC–MS) under the electron impact ionization (EI) mode. The analytes
were converted to their methyl esters by pyrolytic methylation in the hot GC injection port. The recoveries of alkylphosphonic acids were above
95% and the minimum detection limits were as low as 10 ng/mL. The recovery of the test chemicals was tested with solvents, dichloromethane,
n-hexane, ethyl acetate, acetone, acetonitrile and methanol. The chemicals could be efficiently extracted by the hydrophilic solvents. The method
did not work at the highly acidic pH (when acidified with dilute HCl) but worked well from pH 4.0 to 14.0. The present method was also tested with
other tetra-(methyl, ethyl, propyl and n-butyl)ammonium hydroxides. The test chemicals were not converted to their methyl and ethyl esters with
tetramethyl and tetraethylammonium hydroxides, whereas they were converted to their corresponding propyl and n-butyl esters with tetrapropyl
and tetra(n-butyl)ammonium hydroxides. The method was also applied to two highly cross-linked polymeric sorbents DSC-6S and Oasis HLB. The
recovery of the chemicals on these sorbents was observed to be poor. Methylation using phenyltrimethylammonium hydroxide is non-hazardous
and advantageous over methylation using diazomethane. The method was applied to the analysis of aqueous samples given in one of the official
proficiency tests conducted by the Organization for the Prohibition of Chemical Weapons and all the spiked chemicals were identified as methyl
esters.
© 2007 Elsevier B.V. All rights reserved.

Keywords: Solid-phase extraction; Ion-pair chromatography; Derivatization; Chemical warfare agents; Gas chromatography; Mass spectrometry

1. Introduction
Retrospective identification of chemical warfare agents
(CWA) relies on the determination of the characteristic degra-
dation products of the actual toxic chemicals. The CWA once
exposed to environment degrade to various polar degradation
products. The reliable identification of these chemicals depends
on systematic, effective and simple sample preparation meth-
ods and sensitive detection techniques. Development of simple
identification techniques is always a challenging task and the
夽 In order to detect alkylphosphonic acids in aqueous samples,
IICT communication number 060713.
∗ Corresponding author. Tel.: +91 40 27193482; fax: +91 40 27193156.
E-mail address: saradhi@iict.res.in (U.V.R. Vijaya Saradhi).
methods can be used for the proficiency tests conducted by the
Organization for the Prohibition of Chemical Weapons (OPCW)
for the off-site analysis of samples. Generally, the concentration
of CWA and their degradation products are in the parts per mil-
lion ranges; hence the methods should be efficient to extract and
detect them to maximum possible levels. The extraction methods
and state of art detection and identification methodologies have
been extensively reviewed in the recent review articles [1,2].
Among the compounds related to chemical weapons con-
vention (CWC), nerve agents are the most toxic and volatile
chemicals. They degrade to their corresponding alkylphospho-
nic acids when they come into contact with the aqueous medium.
the samples are generally subjected to direct evaporation fol-
lowed by derivatization (e.g., silylation and methylation) as per

0021-9673/$ – see front matter © 2007 Elsevier B.V. All rights reserved.
doi:10.1016/j.chroma.2007.04.057
392 U.V.R. Vijaya Saradhi et al. / J. Chromatogr. A 1157 (2007) 391–398
the recommended operating procedures [2,3]. The recovery of
alkylphosphonic acids and their derivatization is often hampered
by the dissolved salts present in the matrix. Alternatively, cation
exchange and anion exchange solid-phase extraction (SPE)
methods have been developed to improve the recoveries. Reverse
phase (RP) SPE methods were also developed but they show
very low recoveries when compared to the ion-exchange SPE
methods. Recently, we have reported an ion-pair solid-phase
extraction (IP-SPE) method on C18 RP cartridges to improve
recoveries of alkylphosphonic acids and other acidic degradation
products thiodiglycolic acid and benzilic acid, which include
silylation of extracts and detection by gas chromatography–mass
spectrometry (GC–MS) [4]. IP-SPE method is advantageous
over the ion-exchange methods due to the reduction in num-
ber of steps and simultaneous extraction of other non polar
compounds along with the polar compounds. In our previous
communication, we reported a recovery of 20–35% for methyl
and ethylphosphonic acids on C18 SPE columns. In the present
method, we have selected the activated charcoal adsorbent for
improving the recoveries of alkylphosphonic acids.
Carbon, porous polymers, ion exchangers and metal-loaded
chelating resins are suitable for the extraction of medium and
high-polarity compounds [5]. Of these, carbon sorbents were
found to be advantageous in the trace analysis of various groups
of ionogenic compounds [6–9]. Graphitized carbon black has
been used for the recovery of some organic substances from
water and for sample enrichment traps in air pollution analy-
sis. GCB was also shown to be a good adsorbent and have a
great potential for the trace enrichment of organic acids [10]. A
macroporous carbon sorbent, packed into disposable columns
was used for the preconcentration of short chain fatty acids (FA)
from drinking water in conjunction with their determination by
capillary isotachophoresis [11]. A macroporous carbonaceous
sorbent having high surface area 1600 m2 /gm was evaluated for
the enrichment of C1–C9 FAs from aqueous solutions and its
capabilities for sample preparation in the determination of these
compounds in drinking water [12]. Combination of activated
carbon and C18 RP silica was used for the effective extraction
of less volatile flavour compounds from ultrahigh temperature
processed milk [13]. Activated charcoal was also used for the
efficient extraction of polar organic pesticides like acephate from
human urine [14]. Sorbent activated charcoal is widely used for
plant extract purification as well as for other matrices, especially
for the recovery of more polar organophosphorus pesticides like
methamidophos and acephate [15–19]. The EPA method for
the determination of nitrosamines in drinking water by solid
phase extraction recommends the use of activated charcoal as
the stationary phase [20].
Derivatization is an important step for the analysis of polar
compounds by GC. Various derivatization methods were used
for the identification of alkylphosphonic acids from the envi-
ronmental matrices and have been reviewed in the recent
articles [21–24]. The most important derivatization methods
are silylation and alkylation. Alkylation methods like methyla-
tion, pentafluorobenzylation [25–27] and acetylation [28] were
frequently used. Out of these methods, methylation with dia-
zomethane is widely used during the proficiency tests. Although,
the reaction with diazomethane is fast and the yield of the
products is high, alternative methylation methods are preferred
due to the potential hazards and difficulty in the preparation
of diazomethane. Pyrolytic alkylation reactions are well docu-
mented in the literature for various kinds of organic chemicals
and have been reviewed in some recent reviews [29–31]. The
pyrolytic alkylation reactions are generally conducted with tetra-
alkylammonium salts. The analytes which are acidic in nature
are mixed with the tetra-alkylammonium salts/hydroxides and
injected on column in to the GC injection port operated at
250–300 ◦ C, where the analytes will be derivatized. Among the
tetra-alkylammonium salts, phenyltrimethylammonium hydrox-
ide (PTMAH) is frequently used for the on column methylation
of fatty acids, herbicides, phenols and in several biological appli-
cations [28–30]. PTMAH has also been used for the extraction
of methylphosphonic acids retained on the quaternary ammo-
nium anion exchange cartridges and on column methylation in
the GC injection port [32]. PTMAH also acts as an IP reagent
and hence can be used for the extraction of alkylphosphonic
acids from aqueous samples.
In the present paper, we report a simple, economical,
non-hazardous and efficient method for the extraction of
alkylphosphonic acids from aqueous samples using IP-SPE with
PTMAH on activated charcoal adsorbent and pyrolytic methy-
lation by GC–MS.
2. Experimental
2.1. Materials
Methylphosphonic acid (MPA) (1), ethylphosphonic acid
(EPA) (2), propylphosphonic acid (PPA) (3), 1.0 M solu-
tion of phenyltrimethylammonium hydroxide (PTMAH)
in methanol, 1.0 M tetramethylammonium hydroxide
(TMAH) in water, 1.0 M tetraethylammonium hydroxide
(TEAH) in water, 1.0 M tetrapropylammonium hydroxide
(TPAH) in water, 1.0 M tetra(n-butyl)ammonium hydroxide
(TBAH) in methanol, dimethyl methylphosphonate (DMMP),
ethylphosphonyl dichloride (EPDCL), propylphosphonyl
dichloride (PPDCL), tripropylphosphate (TPP) and N,O-
bis(trimethylsilyl)trifluoroacetamide (BSTFA) were purchased
from Sigma (St. Louis, MO, USA). HPLC grade solvents
methanol, acetone, acetonitrile, dichloromethane, n-hexane,
ethyl acetate, anhydrous magnesium sulfate, calcium chloride,
sodium hydrogencarbonate, sodium acetate, and anhydrous
sodium sulphate were obtained from E. Merck (Mumbai, India).
Phosphorus free activated charcoal was purchased from S.D.
Fine Chemicals (Mumbai, India). The SPE vacuum manifold,
Discovery C18 (DSC-18, 500 mg, 6 mL) cartridges and discov-
ery 6S cartridges (DSC-6S, 60 mg, 1 mL) were purchased from
Supelco (Bellefonte, PA, USA). The Oasis HLB (200 mg, 6
CC) SPE cartridges were obtained from Waters Corporation
(Milford, Massachusetts, USA). Propyl propylphosphonate
(PRPP) (4) and n-butyl propylphosphonate (BUPP) (5) were
prepared by treating propylphosphonyldichloride with the
respective alcohols in equimolar ratios. The monochloridates
were distilled under vacuum and divided into two parts. The
 U.V.R. Vijaya Saradhi et al. / J. Chromatogr. A 1157 (2007) 391–398
first part is hydrolysed to obtain chemicals 4 and 5. The second
part was converted to their methyl esters by the addition
of one mole of methanol dropwise at room temperature.
The samples were purified by vacuum distillation. Dimethyl
ethylphosphonate (DMEP) and dimethyl propylphosphonate
(DMPP) were prepared by the reaction of one mole of ethyl and
propylphosphonyldichlorides with a little excess of 2 mol of
methanol at room temperature and the final products DMEP and
DMPP were distilled under vacuum. The purities were assessed
by NMR and GC–MS methods. Chemicals 4 and 5 were
confirmed by GC–MS analysis after silylation with BSTFA
at 60 ◦ C and comparison of their EI mass spectra with those
in e-OCAD library database (version 6.0). Activated charcoal
SPE cartridges were prepared by packing 200 mg of activated
charcoal in 6 mL SPE tubes by using 0.2 ␮m polyethylene frits.
Similarly, anhydrous sodium sulphate cartridges (1 g, 6 mL)
were prepared in-house.
2.2. Preparation of stock solutions
The test sample was prepared by adding 100 mg each of
anhydrous magnesium sulfate, calcium chloride and sodium
hydrogen carbonate in 50 mL double distilled water. The water
was stirred vigorously and sonicated for 10 min. The sample was
filtered through 0.45 ␮ cellulose filter paper. The test sample was
spiked with 10 mg each of chemicals 1–5. Standard solutions of
DMMP, DMEP, DMPP, PRPP, BUPP and TPP were prepared by
adding 10 mg of sample in 10 mL methanol. The standard solu-
tions were diluted accordingly as and when required. Acetate
buffer was prepared by dissolving 1.36 g of sodium acetate and
0.6 mL of glacial acetic acid in 100 mL water. The pH of the
solution was adjusted with 10% glacial acetic acid towards the
acidic region and 0.1 M sodium hydroxide towards the basic
region, respectively, as and when required.
2.3. Solid-phase extraction
The SPE cartridge (activated charcoal, DSC-18, DSC-6S and
HLB) was conditioned by passing 2 mL methanol followed by
2 mL deionized water. The test sample (5 mL) was mixed with
200 ␮L of 1.0 M PTMAH solution and allowed to stand for
10 min. The sample was then passed through the cartridge and
the cartridge was washed with 5 mL of deionized water. The car-
tridge was dried for 5 min by applying vacuum. The chemicals
retained in the cartridge were eluted with 5 mL methanol. The
methanol fraction was passed through the anhydrous sodium
sulphate cartridge (1 g, 6 mL) to remove the traces of residual
water. The methanol fraction was concentrated to 100 ␮L by
gentle stream of nitrogen, before GC–MS analysis.
2.4. GC–MS analysis
The GC–MS analyses were carried out on Agilent 6890 GC
system equipped with a 5973 inert mass selective detector and
7863 autosampler (Agilent Technologies, USA). A CP SIL 8
CB (5% phenyl, 95% dimethylpolysiloxane) (Varian, Middle-
burg, The Netherlands) column of 30 m length, 0.25 mm internal
393
diameter and 0.25 ␮m film thickness was used. The oven was
programmed from an initial temperature 50 ◦ C (hold for 2 min)
to the final temperature 280 ◦ C at the rate of 10 ◦ C/min. The
final temperature hold up time was 5 min. Helium at the rate of
1 mL/min was used as the carrier gas in constant flow mode.
The inlet and interface temperatures were kept at 280 ◦ C. The
EI source was operated at 230 ◦ C and the quadrupole temper-
ature was 150 ◦ C. The MS was scanned from 30 to 600 units
for recording full scan spectra. For calculating recoveries for
IP-SPE method and its detection limits, the MS was operated
in selected ion monitoring (SIM) mode. For SIM experiment,
multiple ions from each compound (m/z 79, 94 and 124 for 1; 79
and 110 for 2; 79, 110 and 124 for 3, 139 and 111 for chemicals
4 and 5) were selected for the valid characterization of com-
pounds. One microlitre of the sample was injected in split mode
at a split ratio of 5:1 by the autosampler.
3. Results and discussion
Activated carbons are generally highly hydrophobic and have
high surface area. Hence, the amount of the substance adsorbed
is more when compared to the normal RP adsorbents. The sur-
face area of the activated carbon depends on the temperature at
which it is prepared. The chemicals are generally adsorbed phys-
ically and can be easily desorbed with proper solvents. Hence,
activated carbons are widely used as stationary phases in GC,
HPLC and SPE as discussed in the introduction section. Some
authors used activated charcoal having a very high surface area
(1600 m2 /g) for the extraction of organic acids and obtained sat-
isfactory recoveries [14]. In the present method we used the
activated carbon having BET surface area 900 m2 /g for the IP-
SPE of alkylphosphonic acids from aqueous samples and the
results were compared to those obtained on the C18 cartridges
and two polymeric sorbents DSC-6S and Oasis HLB (surface
area 450 m2 /g, 600 m2 /g and 800 m2 /g, respectively).
The test samples spiked with the test chemicals 1, 2, 3, 4
and 5, were diluted to a concentration of 5 ␮g/mL. Five millil-
itres of the dilute solution was mixed with 500 ␮L of 1.0 M
PTMAH in methanol solution and allowed to stand for 10 min.
The mixture was subjected to SPE on activated charcoal car-
tridge as described in the Section 2. The contents retained on the
cartridge were eluted with 5 mL methanol and the methanol frac-
tion was concentrated to 100 ␮L by gentle stream of nitrogen.
The aliquot was analysed by GC–MS after adding the internal
standard TPP in EI full scan and SIM modes. The total ion chro-
matogram (TIC) obtained in the full scan mode showed that
the alkylphosphonic acids 1–5 were converted to their corre-
sponding methyl esters, dimethyl methylphosphonate, dimethyl
ethylphosphonate, dimethyl propylphosphonate, methyl propyl
propylphosphonate, methyl butyl propylphosphonate, respec-
tively. The TIC obtained from the GC–MS analysis of the aliquot
from the IP-SPE of aqueous sample in SIM mode representing
the methyl esters of the test chemicals 1–5 is presented in Fig. 1.
The methylation occurred in the hot injection port by the mech-
anism of pyrolytic methylation, which has been described in
the literature for acidic compounds. TMPAH forms an ion-pair
complex with the acidic analytes and when injected into the hot
394 U.V.R. Vijaya Saradhi et al. / J. Chromatogr. A 1157 (2007) 391–398
Fig. 1. SIM TIC of the methyl esters of the test chemicals 1, 2, 3, 4 and 5 obtained from the GC–MS analysis of the aliquot obtained after the IP-SPE analysis with
PTMAH on activated charcoal.
GC injection port maintained at a temperature of 250–300 ◦ C,
the analytes will be methylated. The mechanism of reaction was
described as bimolecular nucleophilic substitution reaction by
means of the electrospray ionization studies [33]. A similar kind
of reaction might have occurred in the GC injection port leading
to the methylation of the analytes.
The method was also tested by using other tetra-
alkylammonium hydroxides namely, tetramethylammonium
hydroxide, tetraethylammonium hydroxide, tetrapropylam-
monium hydroxide and tetra(n-butyl)ammonium hydroxide
solutions. The chemicals were not alkylated with TMAH and
TEAH solutions, whereas they were converted to their corre-
sponding propyl and n-butyl esters when TPAH and TBAH were
used as ion-pairing agents.
For quantification studies, the concentrated methanol aliquot
obtained from IP-SPE experiment was mixed with the same
quantity of the internal standard and made up to 1 mL with
methanol. The recoveries of the test chemicals (1–5) and in turn
extent of methylation by PTMAH were assessed by both internal
and external standard methods. Linear calibration curves were
plotted individually for all the standard chemicals in the concen-
tration range of 0.05–50 ␮g/mL (0.05, 0.5, 1, 25 and 50 ␮g/mL).
Table 1
Extraction recoveries and detection limits obtained in the IP-SPE and GC–MS analysis with PTMAH as IP reagent on activated charcoal
S. no. Name of the compound Structrue
(1) Methylphosphonic acid
(2) Ethylphosphonic acid
(3) Propylphosphonic acid
(4) Propyl propylphosphonate
(5) n-Butyl propylphosphonate
a Recovery ± standard deviation represents five replicate tries.
The recoveries of the test chemicals were calculated by using
peak areas obtained in the GC–MS analysis and the linear regres-
sion equations of the standard chemicals. The recoveries for 1–5
were above 95%. In our previous communication we reported
20–30% recovery for the chemicals 1 and 2 on C18 RP-SPE
cartridges [4]. The analysis was also performed on the reversed
phase C18 cartridges in order to test the efficiency of the method.
The recovery of these compounds was not improved even with
the present method on C18 cartridges. Hence, it can be concluded
that change of IP reagent did not improve the recovery of the test
chemicals. Instead, the recoveries were above 95% on activated
charcoal cartridges. This may be due to more hydrophobicity
of the activated carbon and efficient adsorption on the carbon
surface when compared to that of octadecylsilica surface. The
analysis was also performed on two polymer based SPE phases
namely DSC-6S and HLB, which have almost comparable sur-
face area to that of activated charcoal, for understanding the
retention mechanism of the sorbents. The recovery of the test
chemicals is less than 0.5% on these polymeric sorbents. This
suggests the better adsorption on the activated charcoal when
compared to the other adsorbents. The adsorption on carbon
surfaces can be either physical or chemical with weak van der
Recoverya Minimum detection limit
Scan (␮g/mL) SIM (ng/mL)
96.5 ± 3.1% F = 0.89 0.20 15
97.4 ± 2.5% F = 0.95 0.10 10
98.2 ± 2.3% F = 1.10 0.10 10
95.1 ± 3.5% F = 0.90 0.10 10
96.5 ± 2.8% F = 0.94 0.10 10
 U.V.R. Vijaya Saradhi et al. / J. Chromatogr. A 1157 (2007) 391–398 395

Fig. 2. Plot of concentration of chemical no. 3 against GC–MS-SIM peak area.
Waals forces between the analyte and the active groups on the
carbon surface. The compounds bound with these forces can
be desorbed either thermally with the increase of temperature
or chemically with the use of organic solvents of appropriate
polarity.
The F-test values for the recoveries were calculated at two dif-
ferent sets of concentration levels viz., 0.05 ␮g/mL and 5 ␮g/mL,
50 ␮g/mL and 100 ␮g/mL, and the values are in the range of
0.91–1.05 in order to evaluate the standard deviations of the
recoveries. The recoveries of the chemicals along with their
percentage relative standard deviations and F-test values are
summarized in Table 1. The inter-day and intra-day precision
of the recoveries were assessed by using the data from five
replicate analyses of the test chemicals individually at three con-
centration levels viz., 1 ␮g/mL, 10 ␮g/mL and 50 ␮g/mL. The

Fig. 3. GC–MS-SIM chromatogram of 15 ng/mL of dimethylester of chemical No. 1 after IP-SPE at S/N > 10:1.

Fig. 4. Total ion chromatograms obtained from the GC–MS analysis of IP-SPE aliquot of water sample-1 of the 19th proficiency test in (A) full scan and (B) SIM
modes representing the methyl esters of the all spiked chemicals.
396 U.V.R. Vijaya Saradhi et al. / J. Chromatogr. A 1157 (2007) 391–398
precision was determined at each concentration level by calcu-
lating the percentage relative standard deviations. The precision
levels were below 2.9% over the defined concentration levels.
The effect of various solvents on the extraction of the spiked
test chemicals was studied with different solvents namely,
dichloromethane, n-hexane, ethylacetate, acetone, acetonitrile
and methanol. The recovery of the compounds is negligible
with the hydrophobic solvents dichloromethane, n-hexane and
ethylacetate, 80% recovery was obtained with acetone and the
chemicals were recovered above 95% with acetonitrile and
methanol. The extraction recovery was also tested with various
methanol–water mixtures. The chemicals started eluting with
60% methanol onwards and could be completely eluted with
80% methanol. But, the aliquots obtained from the methanol
water mixture must be evaporated to dryness. In order to avoid
water from the aliquot, in the present method, pure methanol is
used for the extraction of spiked chemicals.
In view to understand the effect of concentration of PTMAH
for the extraction of test chemicals in the present method,
Fig. 5. GC–MS TIC obtained from the analysis of the aliquot obtained from IP-SPE of W2 sample representing the chemical e at 8.04 min, (B) GC–MS SIM TIC
for the ion m/z 110, and (C) mass spectrum of chemical e.
the experiments were performed with varied concentration of
PTMAH (ranging from 0.001 to 0.5 M) for the test solution
having 10 ␮g/mL concentration of spiked chemicals. The recov-
eries of the test chemicals are consistent and reproducible up to
the concentration of 0.005 M and started decreasing thereon.
The recoveries of the test chemicals are poor below 0.005 M of
PTMAH. The method is linear in the test concentration range
0.05–250 ␮g/mL of test chemicals at 0.05 M concentration of
PTMAH.
The effect of pH on the extraction of spiked chemicals was
studied. Three kinds of water samples were prepared in acetate
buffer with pH values 4.0, 7.0 and 10.0. The test chemicals 1–5
were spiked at a concentration level of 10 ␮g/mL. Then 500 ␮L
of the PTMAH solution was added to 5 mL each of test samples.
The pH of the solution became 13–14 due to the basic nature
of PTMAH. After the SPE of the test samples, the recoveries of
the spiked chemicals were calculated and were observed to be
above 95%. In the next step, the pH of the water samples were
drastically changed to 0–1 with the drop wise addition of 1.0 M
 U.V.R. Vijaya Saradhi et al. / J. Chromatogr. A 1157 (2007) 391–398
HCl and 13–14 with the addition of 0.5 M Na2 CO3 solution (the
addition of NaOH liberates ammonia from the solution). The test
chemicals 1–5 were spiked at 10 ␮g/mL concentration levels.
The IP-SPE was performed after the addition 500 ␮L of PTMAH
to 5 mL of the test samples. The test chemicals could not be
recovered in the highly acidic medium. This may be due to the
rapid conversion of PTMAH to its corresponding hydrochloride
salt and hence the ion pairing ability becomes poor due to this
rapid salt equilibrium. The recovery of the test chemicals in the
highly basic medium was above 90%.
The present method was compared to the methods recom-
mended by the OPCW. The test samples were prepared by
spiking 10 ␮g/mL of the test chemicals 1–5 in 50 mL of the
water sample as described in the Section 2.2. Fifty microlitres of
poly(ethylene glycol)-200 was added as the interference chemi-
cal. The analysis was performed by both direct evaporation and
cation exchange method as described in the recommended oper-
ating procedures and final aliquots were silylated with BSTFA
[2,3]. The recovery of the chemicals was tested by compari-
son to their corresponding authentic trimethylsilyl esters. The
recoveries of the chemicals analysed by the direct evapora-
tion method varied between 50–60% and in the case of cation
exchange method, the recoveries were above 93%. The results
of the present method can be compared to the cation exchange
method but the present method is superior due to its simplicity,
reduction in the number of steps and the highest recoveries.
The linearity of the method is tested using the test sam-
ples spiked with propylphosphonic acid. The mass spectrometric
integrated peak area response over the range of 0.05–250 ␮g/mL
at concentrations 0.05, 0.5, 1.0, 50, 100 and 250 ␮g/mL exhib-
ited a linear response with regression equation y = 13327
x + 51457 and R2 = 0.9979. The plot is shown in Fig. 2. The
minimum detection levels (at a signal to noise ratio of 10:1) of
the method for the actual water samples serially diluted up to
a concentration level of 0.1 ng/mL were tested using GC–MS
Fig. 6. Mass spectra of chemicals a, b, c and d obtained from the GC–MS analysis of IP-SPE aliquots of W1 sample.
397
operating in the full scan and SIM modes. The minimum detec-
tion limits for the test chemicals are also represented in Table 1.
The SIM-TIC (for m/z 79, 94 and 124) of the aliquot obtained
from the IP-SPE of the test sample spiked with 15 ng/mL con-
centration of chemical 1 is presented in Fig. 3.
This method has been successfully employed for the anal-
ysis of aqueous samples during the 19th official proficiency
test conducted by the OPCW. The samples contain the inor-
ganic salts magnesium sulfate, sodium hydrogencarbonate and
calcium chloride and poly(ethyleneglycol) monomethyl ester as
interference along with the spiked chemicals. Poly(ethylene gly-
col)s are commonly added as interferences as their presence
always masks the presence of spiked chemicals and the inor-
ganic salts mask the acidic compounds by transforming them to
their corresponding salts. The spiked chemicals were detected by
present method from the samples. Two millilitres each of aque-
ous samples (W1 and W2) were mixed with 200 ␮L of 1.0 M
PTMAH in methanol solution and subjected to SPE on the acti-
vated charcoal cartridges. The retained chemicals on cartridge
were eluted with 5 mL methanol and the aliquots were concen-
trated to 100 ␮L and analysed by GC–MS in EI and CI modes.
The TICs obtained from the GC–MS analysis of the W1 and
W2 showed the methyl esters of all the spiked phosphonic acids
namely, methylphosphonic acid, pinacolyl methylphosphonic
acid, 1-methylpentyl methylphosphonic acid, 4-methylpentyl
methylphosphonic acid in the W1 sample and isopropylphos-
phonic acid in the W2 sample. These chemicals after the
IP-SPE with PTMAH were converted to their methyl esters,
dimethyl methylphosphonate (a), methyl pinacolyl methylphos-
phonate (b), methyl 1-methylpentyl methylphosphonate (c),
methyl 4-methylpentyl methylphosphonate (d) and dimethyl
isopropylphosphonate (e), respectively. The TICs and extracted
ion chromatograms (EICs) showing the presence of all spiked
chemicals as methyl esters in the W1 sample is presented in
Fig. 4 and W2 sample is presented in Fig. 5. Pinacolyl methyl
398 U.V.R. Vijaya Saradhi et al. / J. Chromatogr. A 1157 (2007) 391–398
methylphosphonate and methyl 1-methylpentyl methylphos-
phonate showed split peaks due to the presence of diastereomers.
The quality of mass spectra of the derivatized chemicals is also
good and the mass spectra are represented in Fig. 6.
4. Conclusions
Here, we have successfully developed an efficient IP-SPE
method to obtain maximum recoveries (>95%) of methyl, ethyl
and propylphosphonic acids by using activated charcoal car-
tridges. The test chemicals present in the aqueous sample were
ion-paired with PTMAH and the solution was passed through the
activated charcoal cartridges. The analytes retained in the car-
tridge were eluted with methanol and the concentrated methanol
extract was analysed by GC–MS. The analytes were methy-
lated by pyrolytic methylation in the GC–MS injection port. The
recoveries of the test chemicals were above 98%. The effect of
various hydrophobic and hydrophilic solvents on the recovery of
the test chemicals was studied and it was found that the chemicals
could be extracted only with the hydrophilic solvents acetone,
acetonitrile and methanol. The chemicals could not be recovered
by the present method from the highly acidic samples but the
method was efficient for the solutions with pH range 4.0–14.0.
The present method was also tested with other tetraalkylammo-
nium hydroxides, namely, TMAH, TEAH, TPAH and TBAH.
The chemicals were not alkylated with TMAH and TEAH
and were converted to their corresponding propyl and n-butyl
esters with TPAH and TBAH. The method was also applied to
two highly cross-linked polymeric sorbents DSC-6S and Oasis
HLB. The recovery of the chemicals on these sorbents was
observed to be poor. The chemicals could be detected in the
parts per billion levels on the activated charcoal sorbent and
hence highly useful for the proficiency tests conducted by the
OPCW.
Acknowledgements
The authors thank Dr. J.S. Yadav, Director, IICT, and Dr. M.
Vairamani for continuous encouragement and support.
References
[1] E.W.J. Hooijschuur, C.E. Kientz, U.A.Th. Brinkman, J. Chromatogr. A 982
(2002) 177.
[2] M. Mesilaakso, Chemical Weapons Convention Chemical Analysis: Sam-
ple Collection, Preparation and Analytical Methods, Wiley, Chichester,
2005, pp. 163.
[3] M. Rautio, Recommended Operating Procedures for Sampling and Anal-
ysis in the Verification of Chemical Disarmament, Finnish Ministry of
Foreign Affairs, Helsinki, 1993.
[4] U.V.R. Vijaya Saradhi, S. Prabhakar, T. Jagadeshwar Reddy, M. Vairamani,
J. Chromatogr. A 1129 (2006) 9.
[5] R.W. Frei, Swiss Chem. 6 (1984) 55.
[6] C.E. Werkhoven-Goewie, U.A.Th. Brinkman, R.W. Frei, Anal. Chem. 53
(1981) 2072.
[7] C. Borra, A. Di Corcia, M. Marchetti, R. Samperi, Anal. Chem. 58 (1986)
2048.
[8] A. Di Corcia, M. Marchetti, R. Samperi, J. Chromatogr. 405 (1987) 357.
[9] F. Andreolini, C. Borra, F. Caccamo, A. Di Corcia, R. Samperi, Anal. Chem.
59 (1987) 1720.
[10] Hennion, J. Chromatogr. A 885 (2000) 73.
[11] M. Hutta, E. Simunicova, D. Kaniansky, J. Tkacova, J. Brtko, J. Chro-
matogr. 470 (1989) 223.
[12] O. Chiantore, I. Novak, D. Berek, Anal. Chem. 60 (1988) 638.
[13] K. Coulibaly, I.J. Jeon, J. Agric. Food. Chem. 40 (1992) 612.
[14] J.T. LePage, V.R. Hebert, E.M. Tomaszewska, J.E. Rothlein, L. McCauley,
J. AOAC. Int. 88 (2005) 1788.
[15] B.I.A. Kaipper, L.A.S. Madureira, H.X. Corseuil, J. Braz. Chem. Soc. 12
(2001) 514.
[16] C.P. Cai, M. Liang, R.R. Wen, Chromatographia 40 (1995) 417.
[17] H.S. Dorea, J.L. Tadeo, C.S. Sanchez, Chromatographia 43 (1996) 380.
[18] S.Y. Szeto, H.R. Maccarthy, P.C. Oloffs, J. Environ. Sci. Health, Part B B14
(1979) 635.
[19] K.M. Sundaram, R. Nott, J. Chromatogr. 627 (1992) 300.
[20] Document EPA/600/R-05/054: Determination of nitrosamines in drinking
water by solid phase extraction and capillary column gas chromatography
with large volume injection and chemical ionization tandem mass spec-
trometry (MS/MS), U.S. Environmental Protection Agency, Cincinnati,
Ohio 45268, September 2004.
[21] R.M. Black, B. Muir, J. Chromatogr. A 1000 (2003) 253.
[22] J.M. Halket, V.G. Zaikin, Eur. J. Mass Spectrom. 11 (2005) 127.
[23] J.M. Halket, V.G. Zaikin, Eur. J. Mass Spectrom. 10 (2004) 1.
[24] J.M. Halket, V.G. Zaikin, Eur. J. Mass Spectrom. 10 (2004) 555.
[25] J. Riches, I. Morton, R.W. Read, R.M. Black, J. Chromatogr. B 816 (2005)
251.
[26] M. Palit, A.K. Gupta, R. Jain, S.K. Raza, J. Chromatogr. A 1043 (2004)
275.
[27] A. Miki, M. Katagi, H. Tsuchihashi, M. Yamashita, J. Anal. Toxicol. 23
(1999) 86.
[28] D. Pardasani, M. Palit, A.K. Gupta, P.K. Kanaujia, D.K. Dubey, J. Chro-
matogr. A 1059 (2004) 157.
[29] M. Rompa, E. Kremer, B. Zygmunt, Anal. Bioanal. Chem. 377 (2003) 590.
[30] R.J. Wells, J. Chromatogr. A 843 (1999) 1.
[31] J. Segura, R. Ventura, C. Jrado, J. Chromatogr. B 713 (1998) 61.
[32] J.Aa. Tornes, B.A. Johnsen, J. Chromatogr. 467 (1989) 129.
[33] S. Gronert, J. Azebu, Org. Lett. 3 (1999) 503.