Chapter 3  |  Chapter 5


4-1. Background

In animals exposed to the toxic organophosphorus nerve agents, substantial amounts of the parent compounds are hydrolyzed to their corresponding phosphonic acids, the rest are covalently bound to enzymes and tissue proteins.1-3 Analytical procedures for quantifying the hydrolyzed phosphonic acids in environmental samples have been reported by many analysts.4-8 For more complex matrices such as biological samples, there has not yet been a method reported to detect these polar acids for verification of exposure. The method described in this chapter is a GC/MS method for the detection of the metabolites of three toxic organophosphorus compounds in urine (sarin, soman, and GF), extracted from a published report.9 Urinary excretion of the metabolite is the primary elimination route for these three compounds. The major differences among them are primarily the extent and rate of excretion. Nearly total recoveries of the given doses for sarin and GF in metabolite form were obtained from the urine in rats dosed subcutaneously.10 Soman was excreted at a slower rate with a recovery of only 62 percent. The acid metabolites can be detected in urine for 4 to 7 days after exposure in rats.

4-2. Materials and methods

a. Materials.

(1) Isopropyl methylphosphonic acid (IMPA) and pinacolyl methylphosphonic acid (PMPA) were synthesized by personnel at U.S. Army Chemical Biological Defense Agency, Aberdeen Proving Ground, MD and their methyl deuterated analogs (d3-IMPA and d3-PMPA) by Chemsyn Science Laboratories (Lenexa, KS). Cyclohexyl methylphosphonic acid (CMPA) was obtained by hydrolyzing GF in base as described in the literature.1 Their respective structures are shown in figure 4-1.

Figure 4-1. Chemical structures of IMPA, PMPA, and CMPA

(2) D3-PMPA was used as an internal standard for both PMPA and CMPA, and deuterated analog for IMPA. Other chemicals and their respective sources are the following:

(a) Concentrated HCl, powdered anhydrous potassium carbonate, sodium chloride (NaCl) (Fisher Scientific, Pittsburgh, PA).

(b) 100 mg Bond-Elut
TM C18 cartridges (Analytichem International, Harbor City, CA).

(c) 1.9 ml microfuge tube (Elkay Products Incorporated, Shrewsbury, MA).

(d) Methanol and methylene chloride (Burdick and Jackson Labs, Muskegon, MI).

(e) 18-crown-6 ether (Eastman Kodak Company, Rochester, NY).

(f) Pentafluorobenzyl bromide (PFBBr) (Pierce Chemical Company, Rockford, IL).

(g) Carbon tetrachloride (Mallinckrodt, Paris, KY).

(h) Nitrogen and helium gases (UHP grade), (Matheson Gases and Equipment, Montgomeryville, PA).

(i) Isobutane reagent gas (99.5 percent), (Specialty Products, Houston, TX).

b. Equipment.

(1) Centra-MTM centrifuge (IEC Company, Needham Heights, MA).

(2) Heating block (Reacti-ThermTM, heating/stirring module; Pierce Chemical Company, Rockford, IL).

(3) Oil bath (N-EVAPTM Analytical Evaporator; Organomation Association Incorporated, South Berlin, MA).

(4) Mixer (Thermolyne Maxi MixTM; Thermolyne Corporation, Dubuque, IA).

(5) Hewlett-Packard 5970B MSD and 5890 GC (Hewlett-Packard, Wilmington, DE).

(6) Finnigan Incos 50B MS (Finnigan MAT, San Jose, CA).

(7) DB-17 and DB-5 bonded-phase capillary column (30 meters x 0.25 mm I.D., film thickness 0.25 mm; J and W Scientific, Folsom, CA).


The chemicals involved in these procedures are toxic. Follow all safety precautions listed below.

c. Sample collection and safety. The three acid metabolites are stable compounds at neutral pH. The samples should be kept frozen to prevent bacterial growth and thawed at analysis time. If noticeable amounts of sediment or particulate matter are present, the sample should be centrifuged and only the supernatant should be used. PFBBr is a lacrimator and eye irritant. All derivatization procedures should be carried out in a ventilated hood.

d. Preparation of standard solution. Weigh about 50 mg of IMPA, PMPA, CMPA, d3-IMPA and d3-PMPA in individual plastic scintillation vials. Add distilled and deionized water to reach a concentration of 10 mg/ml. Dilute each stock solution to 10 µg/ml (10 µl stock solution to 10 ml water). These solutions are stable for at least 3 months if kept refrigerated.

e. Linear standard curve. Spike 1 ml control urine with the prepared standard solutions (10 mg/ml) to make six concentrations of the acids ranging from 20 to 200 ng/ml (e.g., 5
µ l of standard solution added to 1 ml urine makes 50 ng/ml).

f. Sample extraction.

(1) Prepare the C18 solid phase cartridge with 2 x 1 ml of methanol and 2 x 1 ml of purified water.

(2) Add 10 ml each of the deuterated internal standard solution (d3-IMPA and d3-PMPA, 10 
µg/ml) to 1 ml of unknown or spiked urine sample. Acidify the urine sample to approximately pH one by adding concentrated HCl and then pass it through the C18 Bond-ElutTM extraction cartridge.

(3) After rinsing with 1 ml 20 percent NaCl in 0.1 N HCl, elute the extraction cartridge with 1 ml methanol into a 1.9 ml microfuge tube containing 20 mg powdered anhydrous potassium carbonate.

(4) Evaporate the eluant to dryness at 85 oC in an oil bath under nitrogen.

(5) Add 1 ml methylene chloride containing 3 mg 18-crown-6 ether. Add 10
µl PFBBr. Heat the capped tube in a reaction block at 50 oC for 1 hour and vortex the tube every 15 minutes.

(6) Centrifuge the tube for 1 minute. Decant the organic phase to a clean tube and evaporate to dryness at room temperature with nitrogen.

(7) Reconstitute with 100
µl each of carbon tetrachloride and purified water.

(8) Inject 1
µl of the organic layer onto GC/MS for electron impact (EI) analysis.

(9) If necessary, GC/MS analysis using chemical ionization (CI) for positive and/or negative ions can be carried out to further confirm suspected samples.

g. GC/MS.

(1) Helium, the carrier gas, is set at a linear velocity of 35 cm/second (butane injection at 120 oC). The oven temperature is held initially at 60 oC for 1 minute, programmed from 60oC to 200 oC at 20 oC/minute, held at 200 oC for 4 minutes, and then programmed from 200oC to 260 oC at 30 oC/minute and held for 5 minutes. Splitless injections of 1 ml are made with the split delay set at 0.5 minute. The injection port temperature is set at 180 oC; split flow 40 ml/minute; transfer line temperature 280 oC and septum purge 2 ml/minute. The EI MS operating conditions are as follows: ion source pressure 1.5 x 10-5 torr; source temperature 200 oC; electron energy 70 electron volt (eV); and electron emission current 220 microamperes (mA). The MS is operated in the selected ion mode. Three fragment ions characteristic for the derivatized compound of interest are monitored (303, 277, 256 for PMPA; 303, 276, 256 for IMPA; 277, 256, and 80 for CMPA; see relative abundance of the ions in table 4-2), along with the m/z ion of 259 for the deuterated internal standard. The dwell time for each ion is 20 msec resulting in a total scan rate of 6.7 cycles/second.

(2) PMPA is composed of four stereoisomers due to the asymmetric centers at the phosphorus atom and the pinacolyl carbon. Under the chromatographic conditions employed, the stereoisomers are separated into two diastereomer pairs. Quantitation is performed by combining the integrated peak area for each peak. Calibration curves for EI are constructed by plotting the ratio of the peak area of the fragment ion at m/z 256 versus the area of the corresponding m/z 259 fragment ion of the deuterated internal standard. CMPA and IMPA produced only a single peak.

(3) Chemical ionization GC/MS analyses were performed on a Finnegan Incos 50B MS interfaced to a Hewlett-Packard 5890 GC. The GC is fitted with a 30 m x 0.25 mm I.D. DB-5 bonded- phase capillary column, film thickness 0.25
µ m. Helium is set at a linear velocity of 37 cm/second (butane injection at 120 oC). The oven temperature is held initially at 40 oC for 2 minutes, programmed from 40 to 200 oC at 20 oC/minute, held at 200 oC for 4 minutes, and then programmed from 200 to 250 oC at 20 oC/minute. Splitless injections of 1 µl are made with the split delay set at 1.5 minutes. Injection port temperature is set at 180 oC; split flow 40 ml/minute; transfer line temperature 240 oC and septum purge 2 ml/minute. The isobutane chemical ionization MS operating conditions are as follows: ion source pressure 1.0 torr; source temperature 115 oC; electron energy 110 eV; and electron emission current 750 mA. The MS is operated in the selected ion mode. In positive CI mode, the MH+ ion of the derivatized compound is monitored along with its primary fragment ion. In negative CI mode, only the methyl phosphonate anion is observed and monitored (m/z 179 for PMPA, 137 for IMPA, and 177 for CMPA). The scan rate in both positive and negative modes is 0.36 scans/second.

4-3. Results and discussion

a. Solid phase extraction. The recovery from solid-phase extraction was determined for PMPA and IMPA using 14C-labeled compounds. From urine samples the values were 94 percent and 85 percent with coefficients of variation (CV) 1.7 percent and 4.5 percent for PMPA and IMPA respectively (N=6). The hydrophobic pi value of the cyclohexyl group is close to the pinacolyl group. Hence, the recovery for CMPA is expected to be similar to PMPA.

b. Linearity and reproducibility. In spiked urine samples the peak area ratios were linear over concentration ranges from 10 to 200 ng/ml for PMPA, IMPA, and CMPA with a correlation coefficient better than 0.99. The coefficient of variation of repetitive assay was less than 5 percent for urine samples (N=6). The lower quantitation limit was 1 ng/ml in urine for PMPA (CV = 5.7 percent), 5 ng/ml for CMPA (CV = 10.2 percent), and 10 ng/ml for IMPA (CV = 9.6 percent).

c. Chromatograms.

(1) Table 4-1 shows the retention times of the perfluorobenzyl ester of the various phosphonic acids and the deuterated analogs. The retention time increases proportionately as the lipophilicity of the alkyl side chain of the methylphosphonic acid increases. The PMPAs were resolved into two pairs of diastereomer separated by 0.23 minutes. Control human urine (N=10) did not show any interfering peaks at the region where PMPA and CMPA appeared. An interfering peak did appear at the retention time of IMPA when m/z 256 ion was monitored but not 303 ion. For quantitation purposes a correction from the control sample has to be made if possible. If m/z 256 ion was detected in the EI spectra for unknown samples, CI spectra should be performed to confirm the finding.

(2) The EI mass spectra of all three derivatized acids shared several characteristics. A molecular ion was absent for all three derivatized acids. The base peak for all three compounds was at m/z 181, a non-specific ion of the PFBBr derivatizing reagent. The relative abundance of the major ions observed in the EI mass spectra are summarized in table 4-2. The ready loss of the entire R group under EI conditions produced two fragmentation pathways designated as class specific. Both pathways produced abundant ions for all three compounds. In the first fragmentation scheme the R group was lost along with proton transfer to the remaining oxygen. In addition, protonation of the -P=O oxygen also occurred. This process was observed for both the derivatized PMPA and CMPA and has been reported previously for related organophosphorus compounds.11,12 Unlike the other two compounds, protonation of just one of the oxygens produced a more prominent ion at m/z 276 for the derivatized IMPA. Further fragmentation with loss of the -PFB (pentafluorobenzyl) group followed by loss of an OH group produced ions at m/z 97 and 80, respectively. The second class specific pathway resulted from the net loss of the R group along with loss of a single fluorine atom which produced a m/z 256 ion.

(3) Compound specific fragmentation pathways where the R group was only partially fragmented or remained attached in whole to the phosphonyl backbone varied greatly among the three compounds in their relative importance. Derivatized PMPA produced several compound specific ions, but the derivatized CMPA produced virtually none. Loss of a methyl group from -R alkyl chain was observed for the pinacolyl and isopropyl compounds producing ions at m/z 345 and 303 for PMPA and IMPA, respectively. The ion abundance was very small for the pinacolyl group but very prominent for the isopropyl group. An additional compound specific pathway existed for the derivatized PMPA due to the tendency of the pinacolyl moiety to lose an isobutene group. Loss of the isobutene produced an ion at m/z 303 with further loss of the derivatizing group producing an m/z 123 ion. The methyl deuterated label attached directly to the phosphorus atom remained intact in all class and compound specific ions, indicating that the methyl group was not affected by the ionization process.

(4) Under isobutane positive CI conditions, all three derivatized acids produced a strong MH+ ion with a relative abundance of at least 93 percent and was the base peak for both IMPA and CMPA. Other compound specific ions observed resulted from the loss of the PFB derivatizing group with protonation of both oxygens. This process was observed for CMPA and IMPA but the corresponding m/z 181 ion for PMPA was nearly absent (table 4-3). Isobutane adduct ions were also observed for all three compounds. A prominent class specific m/z 277 fragment ion was also observed for all three compounds. While the protonation of the oxygen under EI conditions also produced an m/z 277 ion, the ready availability of protons under CI conditions appeared to make this a much more predominant pathway and produced the base peak for PMPA.

(5) Monitoring of the compound specific ions in EI and CI spectra along with the retention time difference from the deuterated internal standard allows positive identification of each of these metabolites in alleged victims. The presence of closely separated diastereomer pairs produced a distinctive ion chromatogram for PMPA.

Table 4-1. Retention times of the derivatized acids
Compound Retention times (minutes) Difference*
PIMPA 9.992, 10.213 0.034, 0.026
d3-PMPA 9.958, 10.187  
IMPA 8.571 0.024
d3-IMPA 8.547  
CMPA 13.292 3.300

* The difference in minutes between the retention times of the analyte and its internal

Table 4-2. Electron impact mass spectra
Relative abundance (%)
m/z Proposed structure PMPA IMPA CMPA

A. Compound specific ions
345 M-CH3 (PMPA) 0 1 0
303 M-CH3 (IMPA)/ --- 14 ---
  M-C(CH3)3 (PMPA) 20 --- 0
123 M-C(CH3)3-PFB+H 42 0 0

B. Class specific ions
227 M-R+2H 25 3 15
276 M-R+H 4 10 3
256 M-RF 54 66 32
97 CH3P(OH)3 23 9 9
80 CH3P(OH)2 13 44 22

C. Non-specific ions
181 PFB 100 100 100
161 PFB-HF 8 10 4

Table 4-3. Positive CI mass spectra

Relative abundance (%)

m/z Proposed structure PMPA IMPA CMPA

A. Compound specific ions
399 [M+C3H3]+ (PMPA) 7 0 0
397 [M+C3H3]+ (CMPA) 0 0 6
361 [M+H]+ (PMPA) 93 0 3
359 [M+H]+ (CMPA) 1 0 100
357 [M+C3H3]+ (IMPA) 0 1 1
319 [M+H]+ (IMPA) 6 100 9
179 [M-PFB+2H]+ (CMPA) 0 1 72
139 [M-PFB+2H]+ (IMPA) 0 23 4

B. Class specific ions
277 [M-R+2H]+ 100 12 65
256 [M-RF]+ 4 3 9

C. Non-specific ions
181 [PFB]+ 2 7 9


1 Harris, L.W., L.M. Braswell, J.P. Fleisher, and W.J. Cliff. "Metabolites of Pinacolyl Methylphosphonofluoridate (Soman) after Enzymatic Hydrolysis in Vitro," Biochem. Pharmacol. 13, (1964), p. 1129.

2 Reynolds, M.L., P.J. Little, B.F. Thomas, R.B. Bagley, and B.R. Martin. "Relationship Between the Biodisposition of [3H] Soman and its Pharmacological Effects in Mice," Toxicol. Appl. Pharmacol. 80, (1985), p. 409.

3 Lenz, D.E., J. Boisseau, D.M. Maxwell, and E. Heir. "Pharmacokinetics of Soman and its Metabolites in Rats," Proceedings of the 6th Medical Chemical Defense Bioscience Review , (1987), p. 201. AD B121516.

4 Verweij, A., C.E.A.M. Degenhardt, and H.L. Boter. "The Occurrence and Determination of PCH3-containing Compounds in Surface Water," Chemosphere 8, (1979), p. 115.

5 Schiff, L.J., S.G. Pleva, and E.W. Sarver. In Ion Chromatographic Analysis of Environmental Pollutants, Vol. 2 , ed. by J. D. Mulik and E. Sawicki, Ann Arbor Science Publishing, Ann Arbor (1972), p. 329.

6 Bossle, P.C., J.J. Martin, E.W. Sarver, and H.Z. Sommer. "High Performance Liquid Chromatography Analysis of Alkyl Methylphosphonic Acids by Derivatization," J. Chromato. 267, (1983), p. 209.

7 Wils, E.R.J. and A.G. Hulst. "Determination of Organophosphorus Acids by Thermospray Liquid Chromatography-Mass Spectrometry," J. Chromato. 454, (1988), p. 261.

8 Tornes, J.A. and B.A. Johnsen. "Gas Chromatographic Determination of Methylphosphonic Acids by Methylation with Trimethylphenylammonium Hydroxide," J. Chromato. 467, (1989), p. 129.

9 Shih, M.L., J.R. Smith, J.D. McMonagle, T.W. Dolzine, and V.C. Gresham. "Detection of Metabolites of Toxic Alkylmethylphosphonates in Biological Samples," Biol. Mass. Spec. 20, (1991), p. 717.

10 Shih, M.L., J.D. McMonagle, T.W. Dolzine, and V.C. Gresham. "Metabolite Pharmacokinetics of Soman, Sarin, and GF in Rats and Biological Monitoring of Exposure to Toxic Organophosphorus Agents," J. Appl. Tox. 14, (1994), p. 195.

11 Occolowitz, J.L. and G.L. White. "The Mass Spectrometry of Esters of Phosphorous and Phosphonic Acids," Anal. Chem. 35, (1963), p. 1179.

12 Gillis, R.G. and J.L. Occolowitz. In Analytical Chemistry of Phosphorous Compounds , ed. by M. Halmann, Wiley, New York, (1972), p. 313.

Chapter 3  |  Chapter 5

Questions or comments? Send mail to: MRICD Public Affairs Officer or MRICD Web Administrator.
Security & Privacy Notice
Featured Story
The U.S. Army Medical Research Institute of Chemical Defense (USAMRICD) is the DoD lead laboratory for medical chemical defense research.