Identification of acidic degradation products of chemical warfare agents by methylation with trimethylsilyldiazomethane and gas chromatography–mass spectrometry – Journal of Analytical Science and Technology

Chemicals and reagents

Residue analysis grade methanol was purchased from BDH (Poole, UK) and Fluka (Buchs, Switzerland). Dichloromethane (CH₂Cl₂) used in derivatization was residue analysis grade:SupraSolv® from Fluka and from Merck (Darmstadt, Germany). Residue analysis grade acetonitrile for silylation was from BDH. Deionized water was produced using a Millipore Direct-Q® 3UV water purification system (Millipore, Billerica, MA, USA). Acidic methanol (AcMeOH) solutions were prepared from methanol and gaseous hydrogen chloride (HCl) (99.5%, AGA, Espoo, Finland) in the laboratory.

The derivatization reagents TMSDAM (as 2 M solution in n-hexane) and MTBSTFA (silylation grade in 1 ml ampoules) were from Aldrich (Milwaukee, WI, USA). To produce gaseous DAM as described (Hazai and Alexander 1982; Wrolstad et al. 2004) methylation reagent, Diazald® (99%, Fluka) was dissolved in HPLC-grade methyl tert-butyl ether (MTBE, Rathburn, Walkerburn, Scotland). A saturated potassium hydroxide solution in methanol was prepared from pellets (85%) supplied by J.T. Baker (Deventer, the Netherlands). The apparatus is described in Additional file 1: Section 1 Figure S1.

All analytes were of 95% or higher purity determined by NMR. MPA (Fluka), ethyl methylphosphonic acid (EMPA, Aldrich), pinacolyl methylphosphonic acid (PMPA, Aldrich) and BA (Merck) were from commercial sources, whereas propyl propylphosphonic acid (PPPA), isopropyl methylphosphonic acid (IPMPA), methyl ethylphosphonic acid (MEPA), 2-(N,N-dimethylamino)ethanesulfonic acid (DMAESA), DEAESA and DIPAES were synthezed in-house (Spiez Laboratory). The internal standard hexachlorobenzene (HCB, Aldrich) was dissolved in CH₂Cl₂ (residue analysis grade, Fluka).

GC–MS analysis

The GC–MS analyses were performed with an Agilent 6890N GC (Agilent Technologies, Palo Alto, CA, USA) coupled to an Agilent 5973 mass selective detector (MSD) or an Agilent 7890GC coupled to an Agilent 5975C inert XL MSD. Injector temperature was 250 °C and injection volume 1 μl in splitless mode with 1 min splitless time. Helium was used as the carrier gas (AGA) at a constant flow rate of 1.0 ml/min. A DBWax^etr column (polyethylene glycol phase, 30 m, 0.25 mm i.d., 0.25 µm film thickness, J&W Scientific, Folsom, CA, USA) was used for analysis of methylated samples with the following temperature program: 40 °C (1 min)–10 °C/min–260 °C/10 min. Silylated samples were analyzed with a DB-5 ms UI column (phenyl arylene polymer phase, 30 m, 0.25 mm i.d., 0.25 µm film thickness, Agilent Technologies, Folsom, CA, USA) with temperature program: 40 °C (1 min)–10 °C/min–280 °C/15 min). The MS source was operated in electron impact ionization (EI) mode (70 eV), EI source temperature was 230 °C and the scan range at m/z 40–500.

Matrices

Laboratory tap, spring, artificial seawater and a water sample from an OPCW PT containing an inorganic ion background were used as aqueous matrices. Spring water was collected from Ruotsinkylä spring (Tuusula, Finland). The composition of the spring water was determined by the Geological Survey of Finland and is presented in Additional file 1: Section 4 Table S1. Artificial seawater was prepared by diluting 3.8 g of Instant Ocean® Sea Salt (Instant Ocean, Blacksburg, VA, USA) in 100 ml of deionized water. The water sample matrix containing inorganic background ions (‘PT water’) was prepared by diluting 200 mg of both calcium chloride (94%, J.T. Baker) and sodium sulfate (SupraPur, Merck) in 100 ml of deionized water to simulate a PT water sample.

Fine sand, clay, humus and sand soil were used as soil matrices. Before use, the soil samples were dried in air and sieved through a 2 mm screen to remove debris. Fine sand, humus and clay soil were collected from the experimental farm of the Agricultural Research Centre of Finland in Jokioinen and sand was obtained from Soil Analysis Service Ltd. (Helsinki, Finland). The suppliers provided the details presented in Additional file 1: Section 4 Table S2.

Solutions, extraction and cation exchange

Two sets of separate stock solutions of spiking chemicals were used: the first set containing 200 μg/ml of APAs and 200 μg/ml of DEAESA each separately in deionized water and the other set containing 1 mg/ml of SAs and 1 mg/ml BA each separately in deionized water. Stock solution of HCB (1 mg/ml) was prepared in CH₂Cl₂.

For preparation of the linear calibration plots for each set of experiments, 6.5, 7.5, 8.5 and 10 µg/ml of the spiking chemicals were added into matrix blank samples. However, when SAs and BA were silylated and methylated, calibration standards contained 5, 10, 15 or 20 μg/ml of each analyte and the solutions were prepared in deionized water for water samples and in tap water for aqueous extracts of soil samples. The standard addition method for calculating the responses of the spiking chemicals was used.

Usually, 10 µg/ml of each analyte was spiked into an aqueous sample or soil extract and HCB was added in 10 µl to each sample before analysis. However, when methylation of the SAs and BA with TMSDAM was compared with silylation, a spiking level of 20 μg/ml was used to ensure reproducible results. The sample volume was 0.5 ml.

The aqueous extract of each soil was prepared by shaking a portion of 5 g of soil in a glass vial with 5 ml of deionized water for 10 min. The sample was centrifuged at 2000 G (Hettich Universal 16 centrifuge, Hettich, Tuttlingen, Germany) for 5 min and the extract was filtered through Whatman 4 filter paper (Ø 90 cm, Whatman, Maidstone, UK). The extraction was repeated with another 5 ml of water, and the centrifuged and filtered extracts were combined.

10 ml of the water sample or the prepared aqueous extract of soil was cation-exchanged with a 500 mg SCX cartridge (LRC, 3 ml, 40 µm, Varian, Harbor City, CA, USA), which was conditioned with 3 ml of methanol and 6 ml of deionized water.

Evaporation and derivatization

All experiments were done in triplicate. 0.5 ml of the water samples and aqueous extracts was placed in a 1.5 ml screw cap vial and evaporated to dryness with a TurboVap evaporator (TurboVap® LV Concentration Workstation, Caliper Life Sciences, Hopkinton, MA, USA) at 45 °C and 7 psi (for approximately 55 min).

For methylation with TMSDAM, the residue was dissolved in 0.1 ml of 0.2 M dry AcMeOH and vortexed (IKA MS1 minishaker, IKA, Wilmington, NC, USA). 0.85 ml of CH₂Cl₂ was added followed by 0.1 ml of TMSDAM drop by drop without mixing. The vial was capped immediately and heated at 60 °C for 30 min shaking the vial occasionally after every 5 min. To increase the recoveries, the molarity of AcMeOH was increased to 1 M. When the yellow color of the TMSDAM in the solution disappeared, another portion of 0.5 ml of the sample was processed and the derivatization procedure repeated using 0.7 ml of CH₂Cl₂ and 0.2 ml of TMSDAM.

For methylation with DAM, the residue was dissolved in 0.1 ml of 0.2 M dry AcMeOH and vortexed. Approximately 2 g of Diazald® was dissolved in a screw-capped storage bottle (100 ml) in ca. 30 ml of MTBE. The bottle was capped with a three-hole, three-valve screw cap, and 2 ml of saturated KOH/methanol solution (4 g/100 ml) was added to start the formation of DAM. Nitrogen was used as the DAM carrier for the sample methylation. The reaction was continued until the sample color changed to yellow, which marked the end of the reaction and indicated an excess of DAM (see Additional file 1: Section 1). Especially with soil extracts, the color change is often difficult to detect, and therefore, it was always checked that the pH of the sample was not acidic after methylation to be sure that all HCl had reacted with DAM and that methylation had continued for long enough.

For silylation with MTBSTFA, the evaporation residue was dissolved in 0.25 ml of MTBSTFA and vortexed. 0.25 ml of acetonitrile was added, and the vial was capped and heated at 60 °C for 30 min.