Identification and toxicological evaluation of cyclic sulfonium ion degradation products of sulphur mustard

Article abstract of DOI:10.1016/j.marenvres.2020.105047

In the aftermath of WWII large amount seized German chemical munitions were dumped in the Baltic Sea by Allied forces. In this work, we have compared the chemical content of the solidified blocks of dumped WWII mustard gas collected from the Baltic Sea with solid precipitate from stored mustard gas, known as heel. We have identified the same cyclic sulfonium ions in both samples. In assessing the environmental and toxicological impact of dumped sulphur mustard munitions on the world's oceans the potential risk posed by cyclic sulphur mustard salts have so far not been incorporated. The toxicity of 1-(2-chloroethyl)-1,4-dithiane and its hydrolysis product 1-(2-hydroxyethyl)- 1,4-dithiane was evaluated using three different cell lines. Their effect on released pro-inflammatory cytokines was also measured. The toxicity tests showed low toxicity and low pro-inflammatory response and we therefore conclude that the environmental threat posed by these compounds is low.

Full text of DOI:10.1016/j.marenvres.2020.105047

Journal Pre-proof
Identification and toxicological evaluation of cyclic sulfonium ion degradation products
of sulphur mustard
P. Hemström, K.H. Holmgren, B.E. Hammarström, A. Larsson, A. Östin
PII:
S0141-1136(20)30199-9
DOI:
https://doi.org/10.1016/j.marenvres.2020.105047
MERE 105047
Reference:
To appear in: Marine Environmental Research
Received Date: 28 February 2020
Revised Date: 30 May 2020
Accepted Date: 9 June 2020
Please cite this article as: Hemström, P., Holmgren, K.H., Hammarström, B.E., Larsson, A., Östin,
A., Identification and toxicological evaluation of cyclic sulfonium ion degradation products of sulphur
mustard, Marine Environmental Research (2020), doi: https://doi.org/10.1016/j.marenvres.2020.105047.
This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition
of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of
record. This version will undergo additional copyediting, typesetting and review before it is published
in its final form, but we are providing this version to give early visibility of the article. Please note that,
during the production process, errors may be discovered which could affect the content, and all legal
disclaimers that apply to the journal pertain.
© 2020 Published by Elsevier Ltd.

CRediT author statement
Hemström, P: Investigation-Analytical Chemistry, Writing - Original Draft-Chemistry
Holmgren, K. H: Investigation-Analytical Chemistry, Visualization
Hammarström, B. E: Investigation-toxicology, Writing - Original Draft-Toxicology
Larsson, A: Investigation-Synthesis
Östin, A. Conceptualization, Supervision, Project administration, Writing - Review & Editing

Identification and toxicological evaluation of cyclic sulfonium ion degradation products of sulphur
mustard
Hemström, P., Holmgren, K. H, Hammarström, B. E., Larsson, A. and Östin, A.
FOI, CBRN-Defence & Security, The Swedish Defence Research Agency, Cementvägen 20, SE-901 82
Umeå, Sweden
Corresponding Author Anders Östin anders.ostin@foi.se
Abstract
In the aftermath of WWII large amount seized German chemical munitions were dumped in the
Baltic Sea by Allied forces. In this work, we have compared the chemical content of the solidified
blocks of dumped WWII mustard gas collected from the Baltic Sea with solid precipitate from stored
mustard gas, known as heel. We have identified the same cyclic sulfonium ions in both samples. In
assessing the environmental and toxicological impact of dumped sulphur mustard munitions on the
world’s oceans the potential risk posed by cyclic sulphur mustard salts have so far not been
incorporated.
The toxicity of 1-(2-chloroethyl)-1,4-dithiane and its hydrolysis product 1-(2-hydroxyethyl)- 1,4-
dithiane was evaluated using three different cell lines. Their effect on released pro-inflammatory
cytokines was also measured. The toxicity tests showed low toxicity and low pro-inflammatory
response and we therefore conclude that the environmental threat posed by these compounds is
low.

1. Introduction
In the aftermath of WWII, large amounts of seized German chemical weapons, mainly mustard gas,
were disposed by dumping at sea (Beldowski et al., 2016; Duursma, 1999; Glasby, 1997; Greenberg,
Sexton, & Vearrier, 2016) . Today, the dumped mustard gas is in the form of blocks, thought to have
resulted from polymerisation processes that occur when slowly corroding containers rupture and the
hydrophobic mustard gas is exposed to seawater. These blocks preserve some mustard gas internally
for long times, the corrosion/degradation process could take several hundred years (Jurczak &
Fabisiak, 2017). Thus, the presence of large amounts of mustard gas munitions in the sea pose a clear
occupational hazard for workers at sea, e.g. fishermen who may accidentally catch remains of
dumped munitions during trawling. When analysing Baltic Sea sediments, neither mustard gas nor its
expected primary hydrolysis product thiodiglycol are generally found, if this reflects an absence or
the difficulty of analysing sediments is unclear (Söderström, 2014). If mustard gas is detected, it will
be at ppt levels close to leaking objects (Söderström, 2014). Instead, the cyclic products dithiane [28]
[substance numbers according, table 1 and 2] and oxathiane [7] are commonly found and used as
markers for potential leakage of mustard gas degradation products to the surrounding sediments
(Black, Clarke, Cooper, Read, & Utley, 1993; Magnusson, Nordlander, & Ostin, 2016; Roen, Unneberg,
Tornes, & Lundanes, 2010). The formation of dithiane [28] and oxathiane [7] is thought to proceed
through a charged cyclic intermediate: 1-(2-chloroethyl)-1,4-dithiane [3] and 4-(2-chloroethyl)-1,4-
oxathiane [D], respectively (Figures 1 and 2). In the USA, tests connected to the destruction of old
one-ton containers of mustard gas, untouched for 50 years, showed that up to 50% of the content
consisted of a solid precipitate, called ‘mustard heel’ (Rohrbaugh & Yang, 1997). We hypothesise that
formation of polymerized blocks on the seafloor and mustard heel involve the same chemistry, in
which the charged cyclic species 1-(2-chloroethyl)-1,4-dithiane [3] and 4-(2-chloroethyl)-1,4-
oxathiane [D] are key intermediates. In the work reported here, we analysed the polar fraction of a
polymerised mustard block dredged up by a trawler and relate the results to information on mustard
heel.’ The toxicity of the intermediates in the formation of dithiane [28], 1-(2-chloroethyl)-1,4-
dithiane [3] and 1-(2-hydroxyetyl)-1,4-dithiane [2], were also evaluated.
2. Material & Methods
2.1. Chemicals
Solidified samples of sulphur mustard from dumped chemical munitions were caught accidentally by
the fishing vessel Tanja av Grebbestad while fishing near Bornholm in the Baltic Sea. The lump
analysed in this study was solid with a porous structure, and had been stored in a closed vial at room
temperature since 1991.
Mustard heel was obtained from sulphur mustard stored in steel barrels since the 1950s by the
Swedish Armed Forces. The barrels were emptied and the green tar-like residue was sampled with a
hollow glass rod, the barrels were then rinsed with deionized water (10 mL), and a sample of the
resulting solution was collected. This water extract formed a three phase system with a sulphur
mustard phase, a water phase (fraction 2) and an emulsion phase. The mustard heel had similar
consistency to moist sugar, it was soluble in water, and accounted for less than one percent of the
initial mustard gas.
Mustard gas salts are difficult to find commercially. Therefore the reference substances for their
chlorinated [D,3] [substance letter according figure 1] and hydrolysed forms [1,2] were synthesised
following previously reported procedures with slight modifications (Davies & Oxford, 1931;

Stahmann, Fruton, & Bergmann, 1946), Figures 1 and 2. Final products were white solids [D,3] or
colourless crystals [1,2].
All other chemicals were purchased from Sigma-Aldrich (www.sigmaaldrich.com) with at least pro-
analysi purity.
Figure 1. Synthesis of oxathian sulphur mustard salt, 4-(2-chloroethyl)-1,4-oxathian [D], and its hydrolysis product 4-(2-
hydroxyethyl)-1,4-oxathian [1].
Figure 2. Synthesis of dithiane sulphur mustard salt, 1-(2-chloroethyl)-1,4-dithiane [3], and its hydrolysis product 1-(2-
hydroxyetyl)-1,4-dithiane [2].
2.2. Toxicological analysis
Toxicological analysis were performed on 1-(2-chloroethyl)-1,4-dithiane [3], and its hydrolysis
product 1-(2-hydroxyetyl)-1,4-dithiane [2].
In the statistical evaluation in the toxicological experiments below are the results expressed as mean
values ± standard deviation (S.D). The data was analysed by one-way-ANOVA followed by Dunnett’s
post-hoc test. Data was considered significant at p<0.05 (*p<0.05, **p<0.01, ***p<0.001). GraphPad
Prism v.5.02 (GraphPad Software, San Diego, CA, USA) for Microsoft®, Windows, was used for the
statistical analyses.
2.2.1.Analysis of cell toxicity
Cells of the human type II alveolar epithelial cell line A549 (ATCC CCL-185; American Type Culture
Collection) and the mouse fibroblast cell line L929 (ATCC CCL-1) were cultured in RPMI 1640 medium
supplemented with 10% foetal calf serum (FCS; Hyclone, Perbio Science, Aalst, Belgium) and 50
µg/mL gentamicin. The human bronchial epithelial cell line BEAS-2B transformed by an adenovirus
12-SV40 hybrid (ATCC CRL-9609) was grown in serum-free bronchial epithelial cell basal medium with
supplements (BEGM; Cambrex, Verviers, Belgium). BEAS-2B cells were cultivated in tissue culture
flasks or on plates pre-coated with fibronectin, vitrogen and bovine serum albumin. All cells were
maintained at 37° C in a humidified atmosphere containing 5% CO2. For experiments, the cells were
seeded in 24- or 96-well culture plates and allowed to attach overnight before exposure to sulphur
mustard salt or sulphur mustard. The cells were exposed for 1 h, washed and incubated in fresh
medium for another 18 h. Their viability (percentage of apparently living cells) was assessed
fluorescently, using an AlamarBlue assay kit (Thermo Fisher), 24 h after exposure. The significance of
differences between treated cells and controls (exposed to medium only) was also assessed.

2.2.2. Analysis of released pro-inflammatory cytokines
The release of the pro-inflammatory cytokines interleukin (IL)-8 and IL-6 into the culture media was
measured using an immunoassay (ELISA). Lung epithelial cells were seeded at 5x104 in 24-well plates.
After sulphur mustard salt or sulphur mustard exposure for 1 h the medium was removed. Fresh
media was added in a total volume of 0.5 mL for an additional 23 h. The supernatants were then
separated from the cells by centrifugation. IL-8 and IL-6 were measured in the cell-free fluid using the
DuoSet ELISA Development kit (R&D Systems, Abingdon, UK) according to manufacturer’s protocol.
Stimulation with TNF-α was used as a positive control and exposure to complete medium served as
negative control. Each experiment was performed with 4 replicates.
2.3. Sample preparation
2.3.1. Extraction of mustard gas lumps
A sample of 1.26 mg of a solid sulphur mustard lump was extracted with 2 mL dichloromethane
(DCM) for 3 hours at room temperature. The extract was concentrated 10-fold then analysed by gas
chromatography/ mass spectrometry with electron impact ionization (GC/MS EI: see section 2.4 for
instrumental details). Trimethyl silanol (TMS) derivatives of the analytes were also prepared, prior to
GC/MS analysis, by adding 10 µL of N,O-bis(trimethylsily)trifluoroacetamide (BSTFA) and 1%
trimethylchlorosilane (TMCS) to 90 µL DCM extract and heating for 60 minutes at 60 °C.
The DCM-extracted sample was then subjected to a second extraction by incubation in 2 mL 50/50
ACN-water for 3 hours at room temperature. The extract was subsequently divided and half was
evaporated to 500 µL and analysed by liquid chromatography-high resolution mass spectrometry (LC-
HRMS), as described below.
A second lump (4.66 mg) was extracted first with 2 mL acetonitrile and then with 2 mL water for two
hours each in an attempt to extract more compounds for LC-MS analysis. 1 mL of each extract was
then combined and analysed by LC-HRMS (data not shown). The remaining part of the acetonitrile
fraction was analysed by GC/MS after derivatization as described above.
2.3.2. Mustard heel
Hundred microliter of the aqueous phase of the mixture obtained by rinsing mustard heel, as
described above (fraction 2), was diluted with 100 µL ACN then 500 µL DCM was added. DCM and
ACN are miscible while the water forms a separate phase, which was removed, then the organic
phase was analysed (in both native form and after silylation) by GC/MS, as described below.
The aqueous phase (fraction 2) was also diluted 1000-fold with 50/50 acetonitrile/water and
subjected to LC-HRMS analysis, as described below.
2.4. Chemical analysis
Mustard heel from sealed long-term stored mustard gas containers and mustard lumps from dumped
mustard gas were analysed for mustard gas-related products by GC/MS and LC/MS, as described
below.
2.4.1. GC/MS
An Agilent Technologies 7890A gas chromatograph equipped with a 30 m, 0.25 mm id DB-5MS
column (0.25 µm film) coupled to an Agilent 5975C mass selective detector was used for all GC/MS EI
analyses. The carrier gas was helium at a constant 1.0 mL/min, and the samples were introduced via
splitless injection of 1 µL samples at 250 ºC. The GC temperature program for analysis of non-polar
fractions consisted of 40 ºC for 1 minute, followed by a linear increase of 10 ºC/min to 280 ºC, which

was held for 5 minutes. The program for analysing polar silylated fractions was identical except that
the initial temperature (held for the first minute) was 60 ºC. In both cases, the temperatures of the
transfer line, ion source and quadrupole were 280, 230 and 150 ºC, respectively. The mass
spectrometer was used in scan mode (29-500 m/z).
The analytes were identified using the deconvolution software AMDIS software (Automated Mass
Spectral Deconvolution and Identification System, NIST, version 2.73, 2017) and the OPCW OCAD MS-
library v 21 (OPCW central analytical database, 2019). An AMDIS net match factor of ≥ 80 and RI
match of ± 30 units from a reference value were the identification criteria. An in-house AMDIS library
including compounds related to sulphur mustard was also used. Peaks with fragmentation patterns
similar to sulphur mustard-related compounds were sought in NIST databases.
2.4.2. LC-MS
The water-soluble fraction extracted from sulphur mustard containers was analysed by LC-HRMS.
The extracts had a strong green colour, probably due to metal ions from the stainless steel
containers, and they were too paramagnetic for NMR analysis.
Polar fractions were analysed using a Dionex Ultima 3000 chromatographic system fitted with a
Waters AcQuity UPLC HSS C18 (1.8 µm) 2.1x100 mm column coupled to a Bruker Impact QTOF mass
spectrometer operated in positive electrospray ionization (ESI) mode. The mobile phase consisted of
10 mM ammonium acetate + 0.1% formic acid (A) and 99% methanol containing 10 mM ammonium
acetate + 0.1 % formic acid (B). The A:B ratio was held at 95:5 from 0 to 1 min, then changed linearly
to 85:15 at 1.5 min, and 5:95 at 4.5 min. After a hold to 5 min, the column was re-equilibrated with
the initial 95:5 A:B mixture from 5.1 to 6 min. The flowrate was 0.4 mL/min throughout each run. The
MS settings were: capillary voltage 4500 V, end plate offset -500 V, dry gas flow 9 L/min, heater
temperature 200 °C, scanned mass range 100-800 m/z. Sodium formate clusters injected at the start
of analysis were used for internal calibration.
Mass spectra of the two hydroxyl ethyl compounds [1 and 2] were also compared to those of
synthesized standards. Identified compounds are listed in Table 1, in which compounds with multiple
plausible structures for the same formula are noted as tentative.
3. Results
3.1. Toxicological analysis
The toxicity of the mustard salts [2 and 3] in Figure 1, was tested following Karacsonyi (Karacsonyi,
Shanmugam, & Kagan, 2009).
3.1.1. Analysis of cell toxicity
In order to test cell toxicity, three cell systems were used: pulmonary epithelial cell line A549,
pulmonary epithelial cell line BEAS-2B, and L929 mice fibroblast cells. Exposure to concentrations ≥
250 μM for 24 hours significantly reduced the viability of A549 and BEAS-2B pulmonary epithelial
cells, and the highest test concentration (1000 μM) reduced their viability by ca. 20%. L929 mouse
fibroblast cells were slightly more susceptible to exposure to the mustard salts, which reduced their
survival by ca. 30% at 1000 µM. No difference in toxicity towards the three cell systems between the
mustard salt [2] and [3] samples was detected. Exposure to mustard gas [30] at concentrations
ranging from 0.19 to 100 µM caused a significant dose-dependent reduction in viability of both A549
and BEAS-2B cells. At 100 µM (the highest test concentration) it reduced these cells’ viability by 53
and 50%, respectively (Figure 3).

Figure 3. Effects of 1-(2-chloroethyl)-1,4-dithiane [3] and its hydrolysis product 1-(2-hydroxyethyl)- 1,4-dithiane [2] on
viability of (A) an alveolar cell line (A549), (B) a bronchial epithelial cell line (BEAS-2B) and (C) a mouse fibroblast cell
line (L929). Cell viability was assed using the AlamarBlue assay and the results are presented as percentages of the
viability of unexposed control cells. The significance of differences between treated cells and controls (exposed to
medium only) was assessed by one-way ANOVA and Dunnett’s post-hoc-test. Differences that are significant at the p
<0.05 level are marked in presented figures by one (*).
3.1.2.Analysis of released pro-inflammatory cytokines
The expression of pro-inflammatory cytokines in pulmonary epithelial cells A549 and BEAS-2B as a
response to exposure to 1-(2-chloroethyl)-1,4-dithiane [3], and its hydrolysis product 1-(2-
hydroxyetyl)-1,4-dithiane [2] was also measured and correlated to unexposed cells.
Both sulphur mustard salts induced a low upregulation of IL-8 secretion in the alveolar A549 cell line
and in the bronchial BEAS-2B cell line compared to control. 1-(2-hydroxyetyl)-1,4-dithiane [2] had a
slightly higher effect of IL-8 secretion in BEAS-2B cells compared 1-(2-chloroethyl)-1,4-dithiane [3], in
samples at 500 µM (figure 4).
Exposure to sulphur mustard [30] however induced a significant dose-dependent production of IL-8
in both A549 and BEAS-2B cells. The highest concentration of IL-8 was achieved in supernatants at
25µM for A549 (4.3±0.07 fold increase compared to control) and for BEAS-2B (2.0±0.1 fold increase
compared to control).
Neither of the sulphur mustard salt [2 and 3] samples tested nor sulphur mustard [30] samples
induced IL-6 production above background levels, data not shown.

Figure 4. Release of pro-inflammatory cytokine IL-8 from (A- B) an alveolar cell line (A549) as a result of exposure to 1-
(2-chloroethyl)-1,4-dithiane [3], and its hydrolysis product 1-(2-hydroxyetyl)-1,4-dithiane [2] (A) and mustard gas [30]
(B), respectively. Release of pro-inflammatory cytokine IL-8 from (C-D) a bronchial epithelial cell line (BEAS-2B) as a
result of exposure to 1-(2-chloroethyl)-1,4-dithiane [3], and its hydrolysis product 1-(2-hydroxyetyl)-1,4-dithiane [2] (C)
and mustard gas [30] (B), respectively. Error bars indicate standard deviations. Statistical analysis was performed using
one-way-ANOVA followed by Dunnett`s post- test testing significances vs. untreated cells (* p < 0.05, ** p < 0.01).
3.2. Identification of mustard gas degradation products with GC/MS
The GC/MS analysis of lump extracts detected sulphur mustard and a number of sulphur mustard-
related compounds such as 1,2-bis(2-chloroethylthio)ethane [33], bis(2-chloroethyl) disulphide [32]
and bis(2-chloroethylthioethyl)ether [15] (Figure 5). Three cyclic compounds were detected: 1,4-
oxathiane [7], 1,4-dithiane [28] and 1,2,5-trithiepane [29]. The overall findings were similar to those
previously reported (Mazurek, Witkiewicz, Popiel, & Sliwakowski, 2001). Following derivatization,
thiodiglycol [34] and other hydrolysis products were detected in the acetonitrile extract of the
mustard lump (Figure 6). Names, analytical data, formulae and structures of compounds identified in
mustard lump and heel are presented in Table 2. Most of the compounds identified by GC/MS were
found both in lump and heel.

Figure 5. GC/MS Reconstructed ion chromatogram (m/z 104, 109, 210, 123, 152, 154, 182 and 190) of compounds
related to sulphur mustard found in the dichloromethane extract of lump. (Identified compounds numbered according to
Table 2).
Figure 6. GC/MS extracted ion chromatogram (m/z 116, 283 and 298) of TMS-derivatized compounds related to sulphur
mustard found in the acetonitrile extract of lump. (Identified compounds numbered according to Table 2)
3.3. Identification of mustard gas degradation products with LC-HRMS
Analysis of sulphur mustard degradation products by LC-HRMS detected numerous compounds.
Most importantly, some of the most abundant constituents of both extracts of dumped munition
blocks and stored sulphur mustard were hydrolysed cyclic sulphur mustard salts. The two hydrolysis
products [1] and [2] in Figures 7 and 8 were identified by comparing their retention times and exact
masses with those of synthesised standards as shown in Table 1. The other compounds were
tentatively identified by exact masses, isotope patterns and manual spectra interpretation Table 1.

Figure 7. LC-HRMS chromatogram of water-extractable compounds (numbered according to Table 1) in a solid lump
from dumped German WWII mustard gas munitions.
Figure 8. LC-HRMS chromatogram of compounds (numbered according to Table 1) extracted from sulphur mustard
(heel) in storage containers.

Table 1. Sulphur mustard related compounds in lumps and heel detected by LC-HRMS analysis.
Comp.no.
Retention
time
m/z
Formula[M+H]+
CAS
Strucure
Found in sample
1
2
0.63
0.72
149.063
165.040
C6H13O2S
C6H13OS2
255861-28-0
255861-27-9
Lumps and Heel
Lumps and Heel
(R)
(R)
3
0.79
183.006
C6H12S2Cl
199982-97-3
Heel
4
0.97
200.061
C16H32O3S4
N. A.
Heel
5
6
0.97
1.03
177.094
253.093
C8H17O2S
Lumps and Heel
Heel
C10H21O3S2
N. A.
7
1.09
1.14
1.54
1.63
105.037
211.056
193.072
245.043
C4H9OS
15980-15-1
N. A.
Heel
Heel
Heel
Heel
8
C8H16O2SCl
C8H17OS2
9
10
C8H18O2S2Cl
150640-37-2
N.A
11
1.87
269.070
C10H21O2S3
Lumps and Heel
12
13
2.36
2.76
227.033
243.010
C8H16OS2Cl
C8H16S3Cl
N. A.
Heel
Heel
224949-02-4
14
15
2.86
3.38
331.107
263.009
C12H27O4S3
Heel
C8H17OS2Cl2
63918-89-8
N. A.
Lumps and Heel
16
17
3.52
3.54
349.073
373.100
C12H26O3S3Cl
C14H29O3S4
Heel
Heel
18
19
1.48
0.53
137.009
181.035
C4H9OS2
19087-70-8
N. A.
Heel
C6H13O2S2
Lumps and Heel
20
0.54
167.074
C6H15O3S
44910-50-1
Lumps
21
22
23
0.63
0.66
1.02
163.079
179.056
197.012
C7H15O2S
C7H15OS2
C6H13OS3
N. A.
N. A.
N. A.
Lumps
Lumps
Lumps
24
25
1.29
1.42
211.028
273.065
C7H15OS3
C9H21O3S3
N. A.
Lumps
Lumps
1.86: 2.22
and 2.33
26
267.109
C11H23O3S2
N. A.
N. A.
Lumps
2.94 and
3.14
27
283.086
C11H23O2S3
Lumps
R= structure confirmed by reference analysis. Remaining structures are tentative and based on elemental composition.

Table 2. Sulphur mustard related compounds in lumps and heel detected by GC/MS analysis.
Comp.no.
RI
Formula
CAS
Strucure
Found in sample
7
891
C4H8OS
15980-15-1
63918-89-8
505-29-3
Lumps and Heel
Lumps and Heel
Lumps and Heel
15
28
29
30
31
32
33
34
35
36
37
1991
1077
1382
1180
1219
1402
1701
1413
1872
C8H16OS2Cl2
C4H8S2
C4H8S3
6576-93-8
505-60-2
Lumps and Heel
Lumps and Heel
C4H8SCl2
C5H10SCl2
C4H8S2Cl2
C6H12S2Cl2
C10H26SO2Si2
C12H30S2O2Si2
71784-01-5
1002-41-1
3563-36-8
20486-03-7
936623-39-1
Lumps and Heel
Lumps and Heel
Lumps and Heel
Lumps
Lumps and Heel
2148
1672
C14H34S2O3Si2
C10H26SO4Si2
959082-71-4
97916-04-6
Lumps and Heel
Lumps
38
1811
C12H10AsCl
712-48-1
Lumps
4. Discussion
In order to understand the fate of sulphur mustard [30] in WWII dumped munition in the Baltic Sea
we analysed a sample of solidified sulphur mustard lumps and identified the major degradation
products. In our GC/MS study, sulphur mustard [30], four sulphur mustard analogues [15, 31-33] and
arsine oil [38] were identified. These are components of the original weapons grade sulphur mustard
where addition of arsine oil [38] often was used to lower the freezing point (Söderström et al., 2018).
The hydrolysis product of sulphur mustard [30], thiodiglycol [34], as well as its oxidized form is
observed [37] after silylation of the hydroxyl-groups. Furthermore, three cyclic degradation products
are identified where dithiane [28] and oxathiane [7] is thought to form through the cyclic
intermediates 1-(2-chloroethyl)-1,4-dithiane [3] and 4-(2-chloroethyl)-1,4-oxathian [D], as presented
in Figure 9 for dithiane [3]. However, these intermediates are not analysable by GC/MS but require
LC-MS analysis. In the GC/MS analysis the use of deconvolution program with combined
spectra/retention index libraries will provide a good basis for identification. The LC-MS analysis is
dependent on reference chemicals and HRMS for elucidation of elemental composition. In this work
we use authentic reference standards and mustard heel, a product in long-term stored sulphur
mustard that are known to form cyclic sulphur degradation products, to support the identification of
the key intermediates in polymerised lumps of sulphur mustard from the Baltic Sea. Based on this,
the hydrolysis products 1-(2-hydroxyetyl)-1,4-dithian [2] and 4-(2-hydroxyethyl)-1,4-oxathian [1]
were identified in the sulphur mustard lump (Figure 7) while in the mustard heel also 1-(2-

chloroethyl)-1,4-dithiane [3] was identified (Figure 8). Further identifications in the LC-HRMS analysis
lacks comparison with authentic reference and are dependent on manual interpretation and should
therefore be considered tentative. In these products the major difference between lumps and heels
is that almost all chloro-groups are hydrolysed to the corresponding hydroxyl group in the lumps.
Our analysis of sulphur mustard [30] degradation products in both mustard heel and lumps of
solidified dumped sulphur mustard yields comparable results to previous investigations (Mazurek et
al., 2001; Rohrbaugh et al., 1997). That our findings are so similar to the more dedicated work done
by Mazurek et.al. on GC/MS analysis of sulphur mustard lumps indicates that the degradation
processes and end products are relatively similar at least for munitions dumped in the same
geographical area and likely from the same source. Bizzigotti et al. (Bizzigotti, Castelly, Hafez, Smith,
& Whitmire, 2009) stated that cyclic mustard salts are likely formed in dumped sulphur mustard
munitions and that they are water soluble and could thus potentially be released into the
environment but that their fate once released was unknown. Neither sulphur mustard [30], nor the
end product of sulphur mustard polymerisation are water soluble. However, parts of the sulphur
mustard lump/heel fraction are water-soluble and could be major sources of products extracted by
seawater from dumped sulphur mustard. In contact with water, the 1-(2-chloroethyl)-1,4-dithiane [3]
and 4-(2-chloroethyl)-1,4-oxathiane [D] products will eventually be hydrolysed, resulting in the
dithiane [28] and oxathiane [7] observed in sediment surrounding dumping sites(Beldowski et al.,
2016; Magnusson et al., 2016). These compound could also be produced in hydrophobic conditions in
the mustard heel. These two processes are shown for dithiane in figure 9.
Since 1-(2-chloroethyl)-1,4-dithiane [3] and its hydrolysed product could be major products leaking
into the environment, we investigated their toxicity. Epithelial cells from the human respiratory tract
are routinely used in in vitro studies of the toxicity of sulphur mustard and related skin-damaging
substances(Karacsonyi et al., 2009). Here, the toxicity of the cyclic mustard salts, 1-(2-chloroethyl)-
1,4-dithiane [3] and 1-(2-hydroxyetyl)-1,4-dithiane [2] were tested and the results compared with
those of sulphur mustard [30]. The in vitro experiments for cell viability showed that the mustard
salts are less toxic than mustard gas, a 50% cell viability is observed at 100 µM for sulphur mustard
[30] compared to 90% for the compounds [2] and [3] at the same concentration. The released pro-
inflammatory cytokines for 1-(2-chloroethyl)-1,4-dithiane [3] and 1-(2-hydroxyetyl)-1,4-dithiane [2]
are only 1/10 or less as compared to the corresponding concentration of sulphur mustard [30] .
Additionally, the toxicity of sulphur mustard [30] follows a dose-dependent manner while the
response for [2] and [3] is low and dose-independent, similar to the effect on cell viability expressed
by most organic compounds.
Toxicity observed in living systems and published toxicity data are highly relevant for marine systems.
The latter include results of tests of the toxicity of the mustard gas salts towards the bacterium
Aliivibrio fischeri (EC50, 75-400 mg/L)(Storgaard, Christensen, & Sanderso, 2018). Although these are
very different taxa, the results confirm the lower toxicity of the salts, which is supported by the
change in structure (involving elimination of the ability to form the reactive three membered
sulfonium ion rings). Our toxicology studies indicate that the cyclic sulphur mustard salts are not
overly toxic and their leaching from rusted sulphur mustard munitions is not likely to pose great
environmental threats.

Figure 9. The formation of dithane [28] from sulphur mustard [30] may occur through two degradation pathways. In long-
term storage of sulphur mustard the formation of 1-(2-chloroethyl)-1,4-dithiane [3] will result in a release of dithiane [28]
as shown in the yellow box. The mustard gas salt may be extracted by sea water (blue box), and in contact with water 1-
(2-chloroethyl)-1,4-dithiane [3] will be hydrolysed to the corresponding alcohol. These salts can be further hydrolysed to
dithiane [28], which is frequently used as a marker for mustard gas leakage at dumping sites. (HD = sulphur mustard
[30]). The numbers in the figure are according to table 1 and 2. A is detected by LC/MS analysis of heel, B is detection
by GC/MS in lumps or heel, C is toxicity tested on alveaolar cells, bronchial cells and mouse fibroblast cell line and D is
detected in analysis of sediment at duping sites (Magnusson et al., 2016).
5. Conclusion
Our GC/MS analysis results corroborate findings of an extensive analysis of the extractable
hydrophobic volatile fraction of recovered polymerised mustard block by Mazurek et al. (Mazurek et
al., 2001). This clearly showed that active mustard gas and related blister agents are still present in
the polymerised block, and thus pose hazards for direct contact for marine organisms and
occupational health problems for people encountering the object.
Our work supports the hypothesis that the degradation of sulphur mustard in both dumped and
long-term stored mustard gas should be similar. To the best of our knowledge, no previous scientific
investigations have been conducted to support this.
Our work support the hypothesis that cyclic sulphur mustard salts could constitute a large part of the
release of sulphur mustard degradation products at the dumpsites to the environment where further
degradation forms the dithiane and oxathiane that frequently are found in the surrounding
sediment.
We have conducted a toxicological evaluation of these cyclic sulphur mustard salts and found them
to possess some toxicity in cell viability assays and inflammatory response but nothing that is of
special concern compared with the reactivity of mustard gas.
Our conclusion is that the release of these cyclic sulphur mustard salts are not of immediate
environmental concern and that our finding do not warrant a re-evaluation of the environmental
aspects of dumped sulphur mustard munitions.
Acknowledgement: the Swedish Ministry of Defence financed this work

6. References
Beldowski, J., Klusek, Z., Szubska, M., Turja, R., Bulczak, A., Rak, D., . . . Schmidt, B. (2016). Chemical
Munitions Search & Assessment-An evaluation of the dumped munitions problem in the
Baltic Sea. Deep-Sea Research Part Ii-Topical Studies in Oceanography, 128, 85-95.
https://doi.org/10.1016/j.dsr2.2015.01.017.
Bizzigotti, G. O., Castelly, H., Hafez, A. M., Smith, W. H. B., & Whitmire, M. T. (2009). Parameters for
Evaluation of the Fate, Transport, and Environmental Impacts of Chemical Agents in Marine
Environments. Chemical Reviews, 109(1), 236-256. https://doi.org/10.1021/cr0780098.
Black, R., Clarke, R., Cooper, D., Read, R., & Utley, D. (1993). Application Of Headspace Analysis,
Solvent-Extraction, Thermal-Desorption And Gas-Chromatography Mass-Spectrometry To
The Analysis Of Chemical Warfare Samples Containing Sulfur Mustard And Related-
Compounds. Journal of Chromatography, 637(1), 71-80. https://doi.org/10.1016/0021-
9673(93)83100-7.
Davies, J. S. H., & Oxford, A. E. (1931). XXX- Formation of sulphonium chlorides and of unsaturated
substances by the action of water and of aqueous alcoholic potash on ββ′-dichlorodiethyl
sulphide. J. Chem. Soc., 224-236.
Duursma, E. K. (1999). Dumped Chemical Weapons in the Sea - Options. Amsterdam: Dr. A.H
Heineken Foundation for the Environment.
Glasby, G. (1997). Disposal of chemical weapons in the Baltic Sea. Science of the Total Environment,
206(2-3), 267-273.
Greenberg, M. I., Sexton, K. J., & Vearrier, D. (2016). Sea-dumped chemical weapons: environmental
risk, occupational hazard. Clinical Toxicology, 54(2), 79-91.
https://doi.org/10.3109/15563650.2015.1121272.
Jurczak, W., & Fabisiak, J. (2017). Corrosion of ammunition dumped in the Baltic sea. Journal of
KONBiN, 41, 227-246.
Karacsonyi, C., Shanmugam, N., & Kagan, E. (2009). A clinically relevant in vitro model for evaluating
the effects of aerosolized vesicants. Toxicology Letters, 185(1), 38-44.
https://doi.org/10.1016/j.toxlet.2008.11.015.
Magnusson, R., Nordlander, T., & Ostin, A. (2016). Development of a dynamic headspace gas
chromatography-mass spectrometry method for on-site analysis of sulfur mustard
degradation products in sediments. Journal of Chromatography A, 1429, 40-52.
https://doi.org/10.1016/j.chroma.2015.12.009.
Mazurek, M., Witkiewicz, Z., Popiel, S., & Sliwakowski, M. (2001). Capillary gas chromatography-
atomic emission spectroscopy-mass spectrometry analysis of sulphur mustard and
transformation products in a block recovered from the Baltic Sea. Journal of Chromatography
a, 919(1), 133-145. https://doi.org/10.1016/S0021-9673(01)00672-0.
Roen, B., Unneberg, E., Tornes, J., & Lundanes, E. (2010). Headspace-trap gas chromatography-mass
spectrometry for determination of sulphur mustard and related compounds in soil. Journal of
Chromatography a, 1217(14), 2171-2178. https://doi.org/DOI
10.1016/j.chroma.2010.01.088.
Rohrbaugh, D., & Yang, Y. (1997). Liquid chromatography electrospray mass spectrometry of
mustard-related sulfonium ions. Journal of Mass Spectrometry, 32(11), 1247-1252.
Stahmann, M. A., Fruton, J. S., & Bergmann, M. (1946). Chemical reactions of mustard gas and
related compounds. VI. The chemistry of sulfonium salts related to mustard gas., 11, 704-
718.

Storgaard, M. S., Christensen, I., & Sanderso. (2018). Environmental toxicity of CWAs and their
metabolites. In J. Beldowski, R. Been & E. K. Turmus (Eds.), Towards the monitoring of
dumped munitions threat (MODUM). Dordrecht: Springer.
Söderström, M. (2014). WP3. Summary of chemical analysis of sediment samples. CHEMSEA
Chemical Munitions search an assessment, www.chemsea.eu.
Söderström, M., Östin, A., Qvarnström, J., Magnusson, R., Rattfelt-Nyholm, J., Vaher, M., . . .
Bełdowski, J. (2018). Chemical Analysis of Dumped Chemical Warfare Agents During the
MODUM Project. In J. Beldowski, R. Been & E. K. Turmus (Eds.), Towards the monitoring of
dumped munitions threat (MODUM). Dordrecht: Springer.

Highlights
•
•
•
Analysis of dumped mustard gas revealed cyclic sulfonium ion degradation products
The identified sulfonium ions may be the major source of leakage to the environment
The identified cyclic sulfonium ion products possess low toxicity

Declaration of interests
☒ The authors declare that they have no known competing financial interests or personal relationships
that could have appeared to influence the work reported in this paper.
☐The authors declare the following financial interests/personal relationships which may be considered
as potential competing interests:
No conflict of interest