Improving Quantification of tabun, sarin, soman, cyclosarin, and sulfur mustard by focusing agents: A field portable gas chromatography-mass spectrometry study

Commercial gas chromatograph-mass spectrometers, one of which being Inficon’s HAPSITE® ER, have demonstrated chemical detection and identification of nerve agents (G-series) and blistering agents (mustard gas) in the field.

Trang 1

mustard by focusing agents: A ﬁeld portable gas

John T Kellya, ∗, Anthony Qualleya, Geoffrey T Hughesa, Mitchell H Rubensteinb, ∗,

Thomas A Malloyc, Tedeusz Piatkowskic

a UES, Inc., Air Force Research Laboratory, 711th Human Performance Wing/RHMO, 2510 Fifth Street, Area B, Building 840, Wright-Patterson AFB, OH 45433,

USA

b United States Air Force 711th Wing – Air Force Research Laboratory, 711th Human Performance Wing/RHMO, 2510 Fifth Street, Area B, Building 840,

Wright-Patterson AFB, OH 45433, USA

c Hazardous Materials Research Center (HMRC), Battelle Columbus Laboratories, Battelle Memorial Institute, Columbus, OH, USA

Article history:

Received 27 July 2020

Revised 30 November 2020

Accepted 2 December 2020

Available online 13 December 2020

Commercialgas chromatograph-massspectrometers, one of whichbeingInficon’s HAPSITE® ER,have demonstratedchemicaldetectionandidentificationofnerveagents(G-series)andblisteringagents (mus-tardgas)inthefield;howevermostanalysesreliesonself-containedorexternalcalibrationthat inher-entlydriftsovertime.Wedescribeananalyticalapproachthatusestarget-basedthermaldesorption stan-dards,calledfocusing agents,toaccuratelycalculateconcentrationsofchemical warfareagents thatare analyzedbygaschromatograph-massspectrometry.Here,weproviderelativeresponsefactorsof focus-ingagents(2-chloroethylethylsulfide,diisopropylfluorophosphate,diethylmethylphosphonate,diethyl malonate,methylsalicylate,anddichlorvos)thatareusedtoquantifyconcentrationsoftabun,sarin, so-man,cyclosarinand sulfurmustardloaded onthermal desorptiontubes (Tenax® TA).Agingeffectsof focusingagentsareevaluatedbymonitoringdeviationsinquantificationasthermaldesorptiontubesage

instorageatroomtemperatureandrelativehumidity.Theadditionoffocusingagentsimprovesthe quan-tificationoftabun,sarin,soman,cyclosarinandsulfurmustardthatisanalyzedwithinthesamedayas wellasa14-dayperiod.Amongthesixfocusingagentsstudiedhere,diisopropylfluorophosphatehasthe bestperformancefornerveagents(G-series)andblisteringagents(mustardgas)comparedtoother fo-cusingagentsinthisworkandisrecommendedforfielduseforquantification.Theuseoffocusingagent

intheﬁeld leadstomoreaccurateand reliablequantiﬁcation ofTabun (GA),Sarin (GB),Soman (GD), Cyclosarin(GF)andSulfurMustard(HD)thanthetraditionalinternalstandard.Futureimprovementson thedetectionofchemical,biological,radiological,nuclear,andexplosivematerials(CBRNE)canbesafely demonstratedwithstandardscalibratedforharmfulagents

1 Introduction

The need for detecting chemical warfare agents (CWAs) is

ubiquitous and is considered of the highest importance for pro-

tecting soldiers, sailors, airmen, and Marines The military com-

munity supports current methodology for detecting nerve agents

with signiﬁcant gaps in reliability and transferability in ﬁeld-

portable chemical identiﬁcation methods The organophosphorus

∗ Corresponding authors

E-mail address: jkelly@ues.com (J.T Kelly)

nerve agents, tabun (GA), sarin (GB), soman (GD), cyclosarin (GF) and the blistering agent sulfur mustard (HD) are recognized as some of the most lethal CWAs with respect to persistency and tox- icity [1–3] A common point-sensing approach to detecting nerve agents in the field is by mass spectrometry; which has the rep- utation of being the “gold standard” of chemical identification with high sensitivity and high selectivity Gas chromatography- mass spectrometry (GC-MS) is commonly used in the identification of CWAs, as well as chemical precursors and decomposition products however inter- and intra-instrument reproducibility puts into question absolute quantification The chemical structures of the CWAs investigated in this work are shown in Fig.1

https://doi.org/10.1016/j.chroma.2020.461784

( http://creativecommons.org/licenses/by-nc-nd/4.0/ )

Trang 2

J.T Kelly, A Qualley, G.T Hughes et al Journal of Chromatography A 1636 (2021) 461784

Fig 1 Structures of the chemical warfare agents: (a) GA, (b) GB, (c) GD, (d) GF and (e) HD Carbon is black, oxygen is red, nitrogen is blue, ﬂuorine is olive, chlorine is lime,

sulfur is mustard, and phosphorus is orange (For interpretation of the references to color in this ﬁgure legend, the reader is referred to the web version of this article.)

For a number of years, response factors have been of inter-

est for developing novel mass spectrometer motifs and fundamen-

tal aspects: mass analyzers type [4–8] (magnetic sector, time-of

ﬂight, quadrupole, ion trap, or ion cyclotron resonance), ionization

method [9–11] (electron impact, chemical, and electrospray) and

sample introduction [12–15] (liquid injection, solid phase micro-

extraction and thermal desorption) Seto and coworkers have re-

ported the response factors for G-series nerve agents for two ﬁeld-

portable GC-MS spectrometers, the HAPSITE® (Hazardous Air Pol-

lutants on Site) and the HAPSITE® ER (Hazardous Air Pollutants

on Site Extended Range) [ 16, 17] The results of earlier experiments

concluded with signiﬁcant residual carryover effects from adsorp-

tion to the air sampling probe and transfer line to the concen-

trator The HAPSITE® ER has an onboard internal standard, bro-

mopentaﬂuorobenzene (BPFB), that is commonly used for quantiﬁ-

cation however is known to ﬂuctuate between days (26.3 %RSD)

and even more within a single day (32.9%RSD) [17]

While GC-MS has been shown to be capable of deconvoluting

complicated chemical mixtures in solution, there has been signif-

icant work in coupling sorbent traps to automated thermal des-

orption to improve detection limits Combining thermal desorp-

tion with gas chromatography has been demonstrated for the ap-

plication of CWAs in the laboratory setting [18–22] however lit-

erature for ﬁeld-portable techniques is more sparse Kanamori-

Kataoka and Seto reported a thorough study on comparing CWAs

performance on Tenax® TA, Tenax® GR, and Carboxen® 1016 ther-

mal desorption tubes The thermal desorption GC-MS analysis re-

vealed Tenax® TA as the most favorable of the three sorbents in

dry conditions as well as humid conditions up to 50% relative hu-

midity [23] The use of thermal desorption improves the detection

limits substantially however a number of considerations must be

evaluated ( e.g. thermal desorption tube stability, capacity, carry-

over, and sampling rates.) [24]

The primary aim of this work is to provide relative response

factors [25] (RRFs) as target-based standards, referred to in this

study as focusing agents, for G-series agents and sulfur mustard

quantiﬁcation after analyzed by two different HAPSITE® ER sys-

tems Our previous work shows system-level improvement in the

data quality when switching from external and internal standards

to in-situ calibration with isotopic analogues [26] Here, we demon-

strate the quality control in quantiﬁcation of by using the fol-

lowing focusing agents: 2-chloroethyl ethyl sulﬁde (2-CEES), diiso-

propyl ﬂuorophosphate (DIFP), diethyl methylphosphonate (DEMP),

diethyl malonate (DEM), methyl salicylate (MES), and dichlorvos

(DCV) as an alternative focusing agent to demonstrate a transfer-

able RRF for all ﬁeld-portable GC-MS systems Fig.2shows struc-

tures of the focusing agents used in this work Thermal desorption

tubes are spiked with focusing agent and exposed to select con-

centrations of nerve gases (G-series) and blistering agents (mustard gas) to establish RRFs and are monitored over 14 days eval- uate storage at room temperature and relative humidity ranging from 24% to 34% This transferring of RRFs across instruments that would otherwise rely on external calibration alone is referred to as

calibration transportability.

2 Experimental

2.1 Reagents and material

A total set of 6 focusing agents were purchased from at Sigma– Aldrich: 2-CEES, DIFP, DEMP, DEM, MES, and DCV Isopropyl alcohol and acetonitrile (American Chemical Society grade or equiva- lent) were used as solvents for liquid dilution GB, GD, GA, GF, and

HD are not commercially available; however, independent standard solutions were prepared by two different analysts CWA purity was determined by preparing high concentration stock solutions in acetonitrile and analyzing the solution by gas chromatography-ﬂame ionization detection (GC-FID) and followed standard operating procedure Hazardous Materials Research Center [HMRC] IV-056

2.2 Instrumentation

This investigation examined the performance of the HAPSITE®

ER (Inﬁcon) with respect to ﬁve chemical warfare agents: sarin, tabun, soman, cyclosarin, and distilled sulfur mustard by thermal desorption from Supelco Tenax® TA (35/60) thermal desorption tubes The temperature of the thermal desorber sampling system (TDSS) was set to 310 °C during which nitrogen carrier gas trans- ferred the desorbed sample to a tri-bed concentrator that is held

at 45 °C for 12:00, where time in mm:ss A new tri-bed concentrator was installed in each instrument at the start of testing and was not replaced throughout the duration of the present work The tri- bed concentrator is then heated to 280 °C in 11 seconds and then introduced to a DB-1ms GC column (15 m, 0.25 mm ID, 1.0 μm

d f) The column temperature, membrane and valve oven were set

at 60 °C, 120 °C, and 120 °C respectively The temperature proﬁle for all measurements held the GC column at 60 °C for 01:15, followed by a thermal ramp at a rate of 8 °C min −1 for 03:45 The temperature was then ramped at a rate of 25 °C min −1 for 04:24 after reaching 90 °C The GC column temperature was limited to

200 °C for the remaining time of the 15:30 experiment The carrier gas (nitrogen) ﬂow rates through the GC column were depen- dent on a constant inlet pressure of 88 kPa and the column temperature The spatial and temporally separated eluates then pass through a hydrophobic membrane that interfaced with the electron ionization (EI, positive) quadrupole mass spectrometer The

Trang 3

Fig 2 Structures of focusing agents: (a) 2-CEES, (b) DIFP, (c) DEMP, (d) DEM (e) MES, and (f) DCV Carbon is black, oxygen is red, hydrogen is grey, ﬂuorine is olive, chlorine

is lime, sulfur is mustard and phosphorus is orange (For interpretation of the references to color in this ﬁgure legend, the reader is referred to the web version of this article.)

employed source settings were previously established for consis-

tency with Automated Mass Spectral Deconvolution and Identiﬁ-

cation System [ 27, 28] or AMDIS (Version 2.72, 2014, National In-

stitute of Standards and Technology), using NIST mass spectral li-

braries Mass spectra were acquired for the mass-to-charge range

from 45 to 300 (dwell time is 300 μs dwell time, 0.765 scans per

second) The internal standard for this work was bromopentaﬂuo-

robenzene (BPFB, 5.5 ppm) and has previously shown to have poor

reproducibility (R 2ranging from 0.88 to 0.982)

2.3 Thermal desorption tube preparation

For baseline testing and calibration, Supelco Tenax® TA (35/60)

thermal desorption tubes were spiked with 1, 2, 5, 10, and 50 ng of

each focusing agent thermal desorption tubes were characterized

for stability: aging, carry-over and residual focusing agent on the

thermal desorption tube were evaluated by two HAPSITE® ER (in-

struments 112 and 121) analysis as previously described [ 26, 29]

In brief, carryover was determined by a subsequent blank tube

thermally desorbed after a spiked tube analysis Spiked thermal

desorption tubes were analyzed in repeat desorptions to measure

residual focusing agent on the thermal desorption tube Aging ef-

fects of focusing agents spiked on thermal desorption tubes were

measured by comparing mid-point signal response of 5 ng of fo-

cusing agent (capped and stored at ambient temperatures and rel-

ative humidity levels) on days 0, 3, 7, and 14

External calibrations for CWAs (GA, GB, GD, GF, and HD) were

created in triplicate by a six-point curve (1, 2, 5, 10, 20, and 50 ng)

and internal standard values reported in the literature (26.3 %RSD)

were consistent with the values reported here (24.2 %RSD) [17]

The Stability and thermal desorption tube aging effects were estab-

lished for CWAs following the same approach as for the focusing

agents Relative response factors (RRFs) are compared through in-

ternal standard and focusing agents in the quantiﬁcation of CWAs

Area determinations were made for the internal standard and fo-

cusing agents using the HAPSITE® ER IQ software package (v 2.32,

Inﬁcon) and AMDIS protocol

Sample preparation and thermal desorption tube conditioning was previously described [26] In brief, thermal desorption tubes were procured from Supelco and conditioned using a Markes Inter- national TC-20 with dry-purging for multiple sorbent tubes Flow rates did not exceed 50 mL min −1 and conditioning temperatures were set to 280 °C for 120 mins at Wright-Patterson Air Force Base (WPAFB) Thermal desorption tubes were loaded with a Markes International Calibration Solution Loading Rig (CSLR TM) with ﬂow rates not exceeding 50 mL min −1of ultra-high purity nitrogen gas Single and dual regulator pneumatics controllers (Markes Interna- tional Gas01 and Gas03) were used for setting backing pressures

2.4 RRFs

RRFs are calculated for GA, GB, GD, GF, and HD using the areas

of the CWA (A CWA) and focusing agent (A FA) and the ratio of the mass of focusing agent (m FA) and mass of CWA (m CWA) loaded on the thermal desorption tube as shown in Eq 1 The derivative of the relative response function accounts for the area response ratio (A CWA/A FA) and m CWAtherefore the RRF is a product of the derivative and m FA

RRF=ACWAmFA

AFAmCWA

(1)

RRFs are calculated by ﬁtting a six-point analysis at 1, 2, 5, 10,

20 and 50 ng of CWAs by the area response ratio of the CWA and focusing agent Using a forced-origin linear analysis (Origin- Pro 2020b), values for relative response functions including slopes, standard errors, R 2and adjusted-R 2are reported Tabulated values can be found in the Supplemental Materials

2.5 Safety considerations

The CWAs investigated here are highly toxic and were han- dled with personal protective equipment (PPE) Full-face respira- tory masks, self-contained breathing apparatuses, and liquid-proof and vapor-impermeable suits are highly recommended for han- dling CWAs [1] Safety protocol established at Battelle followed

Trang 4

Fig 3 Mass spectra of (a) GA, (b) GB, (c) GD, (d) GF, and (e) HD and asterisk denoting ions used in quantiﬁcation and conﬁrmation Total ion chromatogram of (f) CWAs via

thermal desorption GC-MS (HAPSITE® ER) including focusing agents Combines extracted ion chromatogram of (g) CWAs and (h) focusing agents Area response of 50 ng of

GB, GD, GA, GF, and HD loaded on to Tenax® TA (35/60) thermal desorption tubes

standard operating procedure Hazardous Materials Research Cen-

ter [HMRC] IV-056

3 Results and discussion

3.1 Determination of CWAs

We examined the capability to use focusing agents on ther-

mal desorption tubes as a method of improving quantiﬁcation of

GA, GB, GD, GF and HD in the ﬁeld Six different focusing agents

were nominated and characterized by performance, chemical sta-

bility and reproducibility The NIST library mass spectra of CWAs

is shown in Fig 3a-e and a representation of the total ion chro-

matogram (TIC) containing CWAs and focusing agents is shown in

Fig 3f Extracted ion chromatograms (EIC) of all CWAs is shown

in Fig.3g and the EIC for all focusing agents is shown in Fig 3h

The retention time (mm:ss) and quantiﬁcation ions (m/z) for CWAs

are presented in Table1and are compared to previously reported

values [17] The four-step AMDIS approach was used in the iden-

tiﬁcation of CWAs (noise characterization, perception of CWAs, ex-

traction of CWA spectrum, and library comparison with a target

library for mass spectrum and retention time) [28] Fig.3(f-h) are

reduced representations of experimental data by the HAPSITE® ER

Table 1

CWA retention times (RT) and mass-to- charge ratio for ions used in quantiﬁcation and conﬁrmation (m/z) compared to pre- cious values [17]

This work [Ref 17]

GB 02:48 99 02:47 99

GD 06:13 126 05:57 99

GA 06:58 133 06:47 70

HD 07:27 111 07:05 109

GF 07:44 99 07:52 99

by adapting retention times and response areas to a summation of

a gaussian functions with the peak areas reﬂecting the observed values The experimental TICs and EICs can be found in Fig S1 in the Supplemental Materials

Carryover and residual CWAs remaining on a thermal desorbed thermal desorption tube was measured by comparing area response from a 50 ng spike of CWAs along with that of the internal standard Carryover for GA, GD, GF and HD were 0.26%, 0.04%, 0.03% and 0.02% respectively The thermal desorption tube residual for GA, GD, GF and HD were 0.26%, 0.04%, 0.01% and 0.02%

Trang 5

Fig 4 Mass spectra of (a) 2-CEES, (b) DIFP, (c) DEMP, (d) DEM, (e) MES and (f) DCV and asterisks denoting ions used in quantiﬁcation and conﬁrmation

respectively The nominally larger variations in internal standard

for BPFB range from 58% to 78% for carryover and 48% to 87% for

residual BPFB This lack of stability and reproducibility mandates

an alternative quantitative methodology, and we employ focusing

agents to circumvent this discrepancy Carryover and residual fo-

cusing agent on thermal desorption tubes were less than 0.1% for

both instruments and solvents (isopropyl alcohol and acetonitrile)

There is no significant influence of carryover on the quantifica-

tion of the analysis Neither solvent showed favorable chromato-

graphic response or instrument response leading to the inference

that either could be used in standard preparation The results for

all carry-over and residual experiments for four HAPSITE® ER sys-

tems are in the Supporting Information (Table S1)

Inﬂuence of carryover on the quantiﬁcation of the following

sample Inﬂuence of carryover on the quantiﬁcation of the follow-

ing sample Inﬂuence of carryover on the quantiﬁcation of the fol-

lowing sample Inﬂuence of carryover on the quantiﬁcation of the

following sample Inﬂuence of carryover on the quantiﬁcation of

the following sample

3.2 Characterization of focusing agents

Previous evaluation of thermal desorption analysis on a

portable GC–MS systems compares performance to that of a stan-

dard bench-top instrument with one of the most problematic in-

consistencies is the response area of the internal standards, 1,3,5-

tris(triﬂuoromethyl)benzene and bromopentaﬂuorobenzene [29]

We have demonstrated the improvement of data quality for ﬁeld-

portable GC-MS through the use of focusing agents and em-

ploy them for determining reliable RRFs and demonstrating inter-

instrument transportability [26] The retention time (mm:ss) and

quantiﬁcation ions (m/z) for the selected focusing agents are 2-

CEES – 04:38 (75), DIFP – 05:08 (101), DEMP – 05:52 (79), DEM –

06:31 (115), MES – 07:48 (120) and DCV – 08:08 (109) as seen in

the extract ion chromatogram in Fig.3f and mass spectra in Fig.4

External calibrations for focusing agents, as seen for DIFP in Fig.3,

are consistent and reliably linear within the data set RRFs are his-

torically a signiﬁcant methodology for determining concentration

of species in an analytical setting ranging for gas chromatography

In 1976, Pacer states “The use of relative response factors in an ex-periment illustrates the fact that components of a mixture, present in equal amounts, need not respond equally in order for that response

to be useful for quantitative purposes As long as the response is re-producible, a quantitative method is feasible” and later states that RRFs can provide as little as 1% error by a GC approach [30] Par- due et al [ 31] and Karasek et al [32]demonstrate the peak area methodology for determining RRFs by GC Here, focusing agent area responses are used for obtaining the RRFs for CWAs

Here, we report the RRFs for GA, GB, GD, GF, and HD for focusing agents (2-CEES, DIFP, DEMP, DEM, MES, and DCV) This quantitative relationship provides a reliable and reproducible approach for quantifying G-series CWAs by thermal desorption GC-MS All CWAs and focusing agents were analyzed in a single thermal desorption tube and were done in triplicate across four HAPSITE® ER systems The data that is used for discussion is selected with nearly average results in attempt to reﬂect the overall system performance The 5 ng of each 2-CEES, DIFP, DEMP, DEM, MES, and DCV reﬂected area responses of 5612366, 9170962, 8204334, 7931053,

391090 03, and 2730 0138 These values are used for the RRFs for all CWAs that are used to calculate each relative response function

3.2.1 Tabun (GA)

RRFs for GA are by a six-point quantitative analysis at 1, 2,

5, 10, 20 and 50 ng of GA loaded on thermal desorption tubes and analyzed by thermal desorption GC-MS The linear relative response functions have slopes of 0.241, 0.105, 0.129, 0.092, 0.0192, and 0.0303 for 2-CEES, DIFP, DEMP, DEM, MES and DCV respectively The derivative of the relative response function provides the calculated RRFs without accounting for the focusing agent mass loaded on the thermal desorption tubes The RRFs are 1.203, 0.523, 0.647, 0.458, 0.0962, and 0.1515 for 2-CEES, DIFP, DEMP, DEM, MES and DCV respectively with 5 ng of focusing agent All relative response functions have a ﬁxed y-intercept of 0 for the linear analysis The reliability of focusing agents determined by looking at

Trang 6

Fig 5 RRFs for GA for (a, red) 2-CEES, (b, orange) DIFP, (c, yellow) DEMP, (d, green) DEM, (e, blue) MES and (f, purple) DCV for day 0 The six-point quantitative analysis at 1,

2, 5, 10, 20 and 50 ng are denoted by dots, the relative response function are black traces and the colored shadows represent the standard error in the ﬁt (For interpretation

of the references to color in this ﬁgure legend, the reader is referred to the web version of this article.)

the aging effects of the focusing agents have on the quantiﬁca-

tion of GA after 3, 7 and 14 days after being capped and stored

at ambient temperatures and relative humidity levels ranging from

24% to 34% The percent relative standard deviation (%RSD) for 2-

CEES, DIFP, DEMP, DEM, MES and DCV are 45%, 14%, 12%, 22%, 30%,

and 20% respectively over 14 days In the case of GA, the focus-

ing agents DIFP, DEMP, DEM, MES, and DCV are an improvement

on reliability beyond the BPFB internal standard (32.9%) [17] Fig.5

shows the relative response functions of GA and different focusing

agents (a-f) along with the standard error in the slope The co-

eﬃcient of determination (R 2) and adjusted-R 2 values for relative

response functions of GA are 0.994 or better See Table S1 in the

Supplemental Materials for the slope, standard error in the slope,

R 2and adjusted-R 2values

3.2.2 Sarin (GB)

The linear relative response functions for GB have slopes of

0.179, 0.078, 0.096, 0.068, 0.0143, and 0.0225 and RRFs 0.894,

0.389, 0.481, 0.340, 0.0716, and 0.1126 for 2-CEES, DIFP, DEMP,

DEM, MES and DCV respectively The RRFs are calculated for 5 ng

of focusing agent The reliability of all focusing agents, with the

exception of 2-CEES, was signiﬁcantly better for GB than GA with

%RSDs of 51%, 3%, 8%, 10%, 21%, and 9% for 2-CEES, DIFP, DEMP,

DEM, MES and DCV respectively over 14 days In the case of GB,

the focusing agents DIFP, DEMP, DEM, MES, and DCV are an im-

provement on reliability beyond the BPFB internal standard (32.9%)

[17] Fig.6 shows the relative response functions of GB and dif-

ferent focusing agents (a-f) along with the standard error in the

slope The coeﬃcient of determination (R 2) and adjusted-R 2 values

for relative response functions of GB are 0.997 or better See Table

S2 in the Supplemental Materials for the slope, standard error in

the slope, R 2and adjusted-R 2values

3.2.3 GD

RRFs for GD are calculated for 2-CEES, DIFP, DEMP, DEM, MES

and DCV by the relative response functions for day 0 The slopes

of these functions are 0.40, 0.172, 0.213, 0.151, 0.0317, and 0.0499

The RRFs are 1.983, 0.861, 1.067, 0.754, 0.1587, and 0.2497 The

%RSD over a 14-day aging study reveals that DIFP (18%), DEMP

(25%), DEM (9%), MES (2%) and DCV (9%) are more reliable than the internal standard, BPFB (32.9%) The only focusing agent that performed worse than BPFB is 2-CEES (37%) GD has a retention time of 06:13 and there are two neighboring focusing agents in the TIC for DEMP (05:52) and DEM (06:31) Detection techniques other than mass spectrometry can encounter problems in quantiﬁcation

in the presence of poor peak shapes or exceedingly high concen- trations of GD Fig.7shows the relative response functions of GD and different focusing agents (a-f) along with the standard error

in the slope The coeﬃcient of determination (R 2) and adjusted-

R 2 values for relative response functions of GD are 0.997 or better See Table S3 in the Supplemental Materials for the slope, standard error in the slope, R 2and adjusted-R 2values

3.2.4 GF

RRFs for GF are calculated through the relative response functions for day 0 for 2-CEES, DIFP, DEMP, DEM, MES and DCV The slopes of these functions are 2.1, 0.93, 1.15, 0.82, 0.171 and 0.27 The RRFs are 10.7, 4.66, 5.75, 4.08, 0.856, and 1.35 The %RSD over a 14-day aging study reveals that DIFP (12%), DEMP (18%), DEM (2%), MES (9%) and DCV (3%) are more reliable than the internal standard, BPFB (32.9%) The only focusing agent that performed worse than BPFB is 2-CEES (41%) We report the retention time for GF is 07:44 and the retention time for the focusing agent MES is 07:48 While this method does not resolve individual peaks in the TIC, quantiﬁcation of GF is performed by using unique features in mass spectra where GF and MES can be separated Detection techniques such as FID or PID can encounter problems in quantiﬁcation for GF

if using MES as a focusing agent Fig.8shows the relative response functions of GF and different focusing agents (a-f) along with the standard error in the slope The coeﬃcient of determination (R 2) and adjusted-R 2 values for relative response functions of GF are 0.983 or better See Table S4 in the Supplemental Materials for the slope, standard error in the slope, R 2and adjusted-R 2 values

3.2.5 HD

RRFs for HD are calculated through the relative response functions for day 0 The slopes of these functions are 0.74, 0.32, 0.40, 0.28, 0.059, and 0.093 and the RRFs are 3.71, 1.31, 1.99, 1.41, 0.296,

Trang 7

Fig 6 RRFs for GB for (a, red) 2-CEES, (b, orange) DIFP, (c, yellow) DEMP, (d, green) DEM, (e, blue) MES and (f, purple) DCV for day 0 The six-point quantitative analysis at 1,

Fig 7 RRFs for GD for (a, red) 2-CEES, (b, orange) DIFP, (c, yellow) DEMP, (d, green) DEM, (e, blue) MES and (f, purple) DCV for day 0 The six-point quantitative analysis at 1,

and 0.466 for 2-CEES, DIFP, DEMP, DEM, MES and DCV respectively

The %RSD over a 14-day aging study reveals that 2-CEES (29%),

DIFP (32%), DEM (22%), MES (12%) and DCV (23%) are more reliable

than the internal standard, BPFB (32.9%) The only focusing agent

that performed worse than BPFB was DEMP (37%) The linear ﬁt

analysis shows the largest error in the slope for the relative re-

sponse function for HD and the focusing agents in this work, how-

ever this is still an improvement from the average BPFB internal

standard quantiﬁcation method Fig.9shows the relative response

functions of HD and different focusing agents (a-f) along with the

standard error in the slope The coeﬃcient of determination (R 2)

and adjusted-R 2 values for relative response functions of HD are

0.983 or better See Table S4 in the Supplemental Materials for the slope, standard error in the slope, R 2and adjusted-R 2 values

3.3 RRFs and transferability

The acquisition of data on a single instrument provides reliability and reproducibility, however it is often difficult to compare inter-instrument data with respect to a number of propagating errors [33] Here, we define calibration transferability as an analytical method that lacks the strict protocols applied to laboratory-based analyses through the use of focusing agents and analysis by field- portable thermal desorption GC-MS Table2shows the RRFs for the

Trang 8

Fig 8 RRFs for GF for (a, red) 2-CEES, (b, orange) DIFP, (c, yellow) DEMP, (d, green) DEM, (e, blue) MES and (f, purple) DCV for day 0 The six-point quantitative analysis at 1,

Fig 9 RRFs for HD for (a, red) 2-CEES, (b, orange) DIFP, (c, yellow) DEMP, (d, green) DEM, (e, blue) MES and (f, purple) DCV for day 0 The six-point quantitative analysis at 1,

Table 2

RRFs for GA, GB, GD, GF, and HD when compared to 5 ng of focusing agent RRFs were calculated using

Eq 1 and the slopes of the relative response function for two different HAPSITE® ER systems (Instru- ment A and Instrument B) The %RSD is calculated [33] from the difference in area responses by each instrument for all CWAs and focusing agents Similar RRFs indicate a stable focusing agent across systems

2-CEES 1.20 0.89 1.98 10.73 3.71 0.90 0.66 1.12 6.24 2.53 21%

DIFP 0.52 0.39 0.86 4.66 1.61 0.62 0.46 0.77 4.30 1.75 6%

DEMP 0.65 0.48 1.07 5.75 1.99 0.66 0.49 0.83 4.64 1.87 6%

DEM 0.46 0.34 0.75 4.08 1.41 0.71 0.52 0.88 4.92 2.00 16%

MES 0.10 0.07 0.16 0.86 0.30 0.14 0.11 0.18 0.99 0.40 13%

DCV 0.15 0.11 0.25 1.35 0.47 0.19 0.14 0.24 1.33 0.54 7%

Trang 9

provide an accurate methodology for intra-instrument calibration

but also provides calibration transportability for inter-instrument

comparison The inconsistencies in the performance that were pre-

viously identiﬁed ( i.e variations in internal standard peak areas

and daily response) [29] have been improved by the implemen-

tation of focusing agents Seto previously compares the HAPSITE®

ER to other point sensor technologies which compares most closely

with electrochemical sensors (limits of alarms in the sub mg/m 3

range) [34] The use of focusing agent in the ﬁeld leads to more

accurate and reliable quantiﬁcation of GA, GB, GD, GF and HD than

the traditional internal standard (BPFB) within a single day as well

as across a multiple day timespan BPFB has shown to have as high

as 32.9 %RSD within a single day and 26.3 %RSD between days

[17] Aging effects are tabulated in the Supplemental Material (Ta-

ble S6) in term of %RSD with respect to time Future work will es-

tablish measurement detection limits (MDLs) for the HAPSITE® ER

and relative response factors for VX and Russian VX using focus-

ing agents While focusing agents have been demonstrated against

chemical warfare agents in our laboratory settings, they have yet to

be reported for ﬁeld use Future work expands upon this work by

extending experiments beyond laboratories and into the intended,

more harsh environments

Declaration of Competing Interest

The authors declare that they have no known competing ﬁnan-

cial interests or personal relationships that could have appeared to

inﬂuence the work reported in this paper

CRediT authorship contribution statement

John T Kelly: Writing original draft, Writing review & edit-

ing, Data curation, Visualization Anthony Qualley: Writing orig-

inal draft, Writing review & editing, Data curation, Visualization

Geoffrey T Hughes: Writing original draft, Writing review &

editing, Data curation, Visualization Mitchell H Rubenstein: Con-

ceptualization, Methodology, Writing original draft, Writing re-

view & editing, Project administration Thomas A Malloy: Con-

ceptualization, Formal analysis Tedeusz Piatkowski: Conceptual-

ization, Formal analysis

Acknowledgements

We would like to thank the United StatesAir Force for fund-

ing as well as Drs Darrin Ott and Claude C Grigsby for their sup-

port and encouragement We recognize Mr Will Bell and Dr R

Craig Murdoch who provided program and ﬁnancial management

The views expressed in this article are those of the author and do

not reﬂect the oﬃcial policy or position of the United States Air

Force, Department of Defense, or the U.S Government This work

has been approved for public distribution (Distribution-A, Public

88ABW-2020-2344)

ical and biological warfare agents, Int J Environ Health Res 14 (1) (2004) 31–41

[4] M Sinha , A Tomassian , Development of a miniaturized, light-weight magnetic sector for a ﬁeld-portable mass spectrograph, Rev Sci Instrum 62 (11) (1991) 2618–2620

[5] S.A Ecelberger , T.J Cornish , B.F Collins , D.L Lewis , W.A Bryden , Suitcase TOF:

a man-portable time-of-ﬂight mass spectrometer, Johns Hopkins APL Tech Dig

25 (1) (2004) 14–19 [6] T Spaeder , R Walton , Hapsite GC/MS analysis of low level chemical warfare agents, Toxic Industrial Compounds and Toxic Industrial Materials in the Field

by First Responders, Abstracts of Papers of the American Chemical Society,

2003 U128-U128 [7] D.T Snyder , C.J Pulliam , Z Ouyang , R.G Cooks , Miniature and ﬁeldable mass spectrometers: recent advances, Anal Chem 88 (1) (2016) 2–29

[8] Z Ouyang , R.G Cooks , Miniature mass spectrometers, Annu Rev Anal Chem

2 (2009) 187–214 [9] J Liigand , A Kruve , P Liigand , A Laaniste , M Girod , R Antoine , I Leito , Trans- ferability of the electrospray ionization eﬃciency scale between different instruments, J Am Soc Mass Spectrom 26 (11) (2015) 1923–1930

[10] W Lotz , An empirical formula for the electron-impact ionization cross-section,

Z Physik 206 (2) (1967) 205–211 [11] W.L Fitch , A.D Sauter , Calculation of relative electron impact total ionization cross sections for organic molecules, Anal Chem 55 (6) (1983) 832–835 [12] K Demeestere , J Dewulf , K De Roo , P De Wispelaere , H Van Langenhove , Quality control in quantiﬁcation of volatile organic compounds analysed by thermal desorption–gas chromatography–mass spectrometry, J Chromatogr A

1186 (1-2) (2008) 348–357 [13] S Popiel , M Sankowska , Determination of chemical warfare agents and related compounds in environmental samples by solid-phase microextraction with gas chromatography, J Chromatogr A 1218 (47) (2011) 8457–8479

[14] D Patterson Jr , L Alexander , L Gelbaum , R O’Connor , V Maggio , L Needham , Synthesis and relative response factors for the 22 tetrachlorodibenzo-p-diox- ins (TCDDs) by electron-impact ionization mass spectrometry, Chemosphere 15 (9-12) (1986) 1601–1604

[15] K Lundgren , C Rappe , M Tysklind , Low-resolution mass spectrometric relative response factors (RRFs) and relative retention times (RRTs) on two common gas chromatographic stationary phases for 87 polychlorinated dibenzofurans, Chemosphere 55 (7) (2004) 983–995

[16] H Sekiguchi , K Matsushita , S Yamashiro , Y Sano , Y Seto , T Okuda , A Sato , On-site determination of nerve and mustard gases using a field-portable gas chromatograph-mass spectrometer, Forensic Toxicol 24 (1) (2006) 17–22 [17] H Nagashima , T Kondo , T Nagoya , T Ikeda , N Kurimata , S Unoke , Y Seto , Identification of chemical warfare agents from vapor samples using a field- -portable capillary gas chromatography-membrane-interfaced electron ionization quadrupole mass spectrometry instrument with tri-bed concentrator, J Chromatogr A (2015) (1406) 279–290

[18] R.M Black , R.J Clarke , D.B Cooper , R.W Read , D Utley , Application of headspace analysis, solvent extraction, thermal desorption and gas chromatography—mass spectrometry to the analysis of chemical warfare samples containing sulphur mustard and related compounds, J Chromatogr A 637 (1) (1993) 71–80

[19] W.A Carrick , D.B Cooper , B Muir , Retrospective identiﬁcation of chemical warfare agents by high-temperature automatic thermal desorption–gas chromatography–mass spectrometry", J Chromatogr A 925 (1-2) (2001) 241–249 [20] J Oostdijk , C Degenhardt , H Trap , J Langenberg , Selective and sensitive trace analysis of sulfur mustard with thermal desorption and two-dimensional gas chromatography–mass spectrometry", J Chromatogr A 1150 (1-2) (2007) 62–69

[21] O Terzic , I Swahn , G Cretu , M Palit , G Mallard ,Gas chromatography–full scan mass spectrometry determination of traces of chemical warfare agents and their impurities in air samples by inlet based thermal desorption of sorbent tubes, J Chromatogr A 1225 (2012) 182–192

[22] Y Juillet , C Dubois , F Bintein , J Dissard , A Bossee , Development and valida- tion of a sensitive thermal desorption–gas chromatography–mass spectrometry (TD-GC-MS) method for the determination of phosgene in air samples", Anal Bioanal Chem 406 (21) (2014) 5137–5145

[23] M Kanamori-Kataoka , Y Seto , Measurement of breakthrough volumes of volatile chemical warfare agents on a poly(2, 6-diphenylphenylene ox- ide)-based adsorbent and application to thermal desorption–gas chromatography–mass spectrometric analysis, J Chromatogr A 1410 (2015) 19–27

Trang 10

[24] S.W Harshman , V.L Dershem , M Fan , B.S Watts , G.M Slusher , L.E Flory ,

C.C Grigsby , D.K Ott , The stability of tenax TA thermal sesorption tubes in

simulated ﬁeld conditions on the HAPSITE® ER, Int J Environ Anal Chem 95

(11) (2015) 1014–1029

[25] L Ettre , Nomenclature for Chromatography (IUPAC Recommendations 1993),

Pure Appl Chem 65 (4) (1993) 819–872

[26] A Qualley , G.T Hughes , M.H Rubenstein , Data quality improvement for ﬁeld-

-portable gas chromatography-mass spectrometry through the use of iso-

topic analogues for in-situ calibration", Environ Chem 17 (1) (2020) 28–

38

[27] S.E Stein , An integrated method for spectrum extraction and compound iden-

tiﬁcation from gas chromatography-mass spectrometry data, J Am Soc Mass

Spectrom 10 (8) (1999) 770–781

[28] W.G Mallard , AMDIS in the chemical weapons convention, Anal Bioanal

Chem 406 (21) (2014) 5075–5086

[29] S.W Harshman , M.H Rubenstein , A.V Qualley , M Fan , B.A Geier , R.L Pitsch , G.M Slusher , G.T Hughes , V.L Dershem , C.C Grigsby , Evaluation of thermal desorption analysis on a portable GC–MS system", Int J Environ Anal Chem

97 (3) (2017) 247–263 [30] R.A Pacer , Quantitative gas chromatography using peak heights and relative response factors an undergraduate student experiment, J Chem Educ 53 (9) (1976) 592

[31] H.L Pardue , M.F Burke , J.R Barnes , Quantitative gas chromatography, J Chem Educ 44 (11) (1967) 695

[32] F Karasek , E De Decker , J Tiernay , Qualitative and quantitative gas chromatography for the undergraduate, J Chem Educ 51 (12) (1974) 816

[33] G.L Long , J.D Winefordner , Limit of detection A closer look at the IUPAC def- inition, Anal Chem 55 (7) (1983) 712A–724A

[34] Y Seto , On-site detection of chemical warfare agents, in: Handbook of Toxicol- ogy of Chemical Warfare Agents, Elsevier, 2015, pp 897–914

Tiêu đề	Improving Quantification of Tabun, Sarin, Soman, Cyclosarin, and Sulfur Mustard by Focusing Agents: A Field Portable Gas Chromatography-Mass Spectrometry Study
Tác giả	John T. Kelly, Anthony Qualley, Geoffrey T. Hughes, Mitchell H. Rubenstein, Thomas A. Malloy, Tedeusz Piatkowski
Trường học	United States Air Force Research Laboratory
Chuyên ngành	Chemical Detection and Analysis
Thể loại	Journal Article
Năm xuất bản	2021
Thành phố	Wright-Patterson AFB

Định dạng
Số trang	10
Dung lượng	2,42 MB