Quantification of hydrogen sulfide and methanethiol and the study of their scavenging by biocides of the isothiazolone family

Article abstract of DOI:10.1002/cplu.201300234

A sensitive analytical method for the quantification of malodourous hydrogen sulfide (H2S) and methanethiol (CH3SH) was developed and validated. The method consisted of solidphase microextraction with concomitant in-fibre derivatisation of the mercaptan using N-ethylmaleimide. The adducts were analysed by GC-MS using a triple quadrupole in the selected reaction monitoring mode. The limits of detection of solutions of H2S and CH3SH in water were 100 and 10 ngL-1, respectively, below their odour-perception thresholds. The analytical method was used to show the scavenging effect of 1,2-benzisothiazol- 3(2H)-one (Proxel) and 2-methyl-4-isothiazolinone. The mechanism by which H2S and CH3SH are trapped with Proxel was studied by ultraperformance liquid chromatography (UPLC)-MS.

Full text of DOI:10.1002/cplu.201300234

CHEMPLUSCHEM
FULL PAPERS
DOI: 10.1002/cplu.201300234
Quantification of Hydrogen Sulfide and Methanethiol and
the Study of Their Scavenging by Biocides of the
Isothiazolone Family
Eric Frerot,* Alain Bagnoud, and Esmeralda Cicchetti[a]
Dedicated to Alain Bagnoud on the occasion of his retirement after almost 40 years at Firmenich
A sensitive analytical method for the quantification of mal-
odourous hydrogen sulfide (H2S) and methanethiol (CH3SH)
was developed and validated. The method consisted of solid-
phase microextraction with concomitant in-fibre derivatisation
of the mercaptan using N-ethylmaleimide. The adducts were
analysed by GC–MS using a triple quadrupole in the selected
reaction monitoring mode. The limits of detection of solutions
of H2S and CH3SH in water were 100 and 10 ngLÀ1, respective-
ly, below their odour-perception thresholds. The analytical
method was used to show the scavenging effect of 1,2-benzi-
sothiazol-3(2H)-one (Proxel) and 2-methyl-4-isothiazolinone.
The mechanism by which H2S and CH3SH are trapped with
Proxel was studied by ultraperformance liquid chromatography
(UPLC)-MS.
Introduction
Hydrogen sulfide (H2S) and methanethiol (CH3SH) are very
powerful malodourous chemicals. They can have natural or an-
thropological origins. Thus, H2S causes the rotten-egg odour of
compost or sewer. In January 2013, an accidental discharge of
CH3SH from a factory in northern France caused a stink as far
away as South East England.[1] In lower concentrations, the two
mercaptans may cause aroma defects in beer[2] or white
wine.[3]
sensitive detectors such as a pulsed-flame photometric detec-
tor,[8] an atomic emission detector, or a sulfur chemilumines-
cence detector.[9] In addition, the high volatility of these com-
pounds makes their sampling and gas chromatographic sepa-
ration difficult. Static headspace sampling, a technique in
which several millilitres of air is injected, was used to quantify
H2S, CH3SH, dimethylsulfide and dimethyldisulfide in air.[8] Cryo-
trapping was also used to analyse the atmosphere with better
sensitivity.[9] Solid-phase microextraction (SPME) is a convenient
technique to analyse volatile chemicals. In the analysis of gas-
eous industrial effluents, a satisfactory sensitivity for H2S and
CH3SH and other sulfur compounds was achieved by using
a polydimethylsiloxane (PDMS)-Carboxen fibre.[10] However, the
easy-to-use external calibration could not be used for the most
volatile compounds, H2S and CH3SH, which required the stan-
dard addition method.[10]
In the flavour and fragrance industry, H2S and CH3SH find
beneficial applications in the flavouring of foods such as meat
or coffee products.[4] These sulfur compounds may contribute
to toilet,[5] kitchen[6] or breath[7] malodours, and thus for home-
care products, they should be prevented from being formed or
should be counteracted with a perfume.
During a study on laundry softeners, we observed that in-
tentionally added H2S did not smell after a few hours in certain
products, whereas the odour was persistent in other softener
bases. The development of a GC–MS method for H2S and
CH3SH determination, and the identification of the possible
scavenger in certain softeners, is the subject of this study.
Derivatisation of thiols prior to their injection onto a high-
performance liquid chromatograph or
a GC is another
option.[11] It improves both the selectivity and the ease of de-
tection. Mateo-Vivaracho et al.[12] analysed higher thiols, but
not H2S or CH3SH, in wine by using pentafluorobenzyl deriva-
tives together with GC–MS in the negative chemical ionisation
mode with sensitivity as low as 0.5 ngLÀ1. In this case, the deri-
vatisation was performed in the liquid phase, and a subsequent
extraction had to be performed to get this impressive sensitivi-
ty. The quantification of H2S, CH3SH, ethanethiol and propane-
thiol in waste water was achieved with derivatisation using N-
ethylmaleimide (NEM) in the liquid phase, followed by SPME of
the liquid phase and GC–MS.[13] The limits of detection (LODs)
were below 1 ngLÀ1, with relative standard deviation values
lower than 11%. The mechanism of the derivatisation reaction
of thiols with NEM was studied. In particular, the product of
the first Michael addition of H2S to NEM, namely, NEMÀSH, re-
Results and Discussion
The quantification of H2S and CH3SH by gas chromatography
(GC) presents two main difficulties. Because of the extremely
powerful odour of these compounds, very high sensitivity
measurements have to be achieved. This requires specific and
[a] E. Frerot, A. Bagnoud, E. Cicchetti
Analytical Innovation Group
Corporate R&D Division
Firmenich S.A. 1
Route des Jeunes, 1211 Geneva 8 (Switzerland)
E-mail: eric.frerot@firmenich.com
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemPlusChem 2014, 79, 77 – 82 77

CHEMPLUSCHEM
FULL PAPERS
www.chempluschem.org
acted with a second NEM molecule to give (NEM)2ÀS, which
showed two diastereoisomeric peaks in a 1:1 ratio from GC. In
the water phase, the reaction was never complete and a better
reproducibility was obtained by using the minor peak of NEMÀ
SH instead of that of (NEM)2ÀS.
MS was selected as the precursor ion for one quantifying and
two qualifying transitions, each optimised with respect to colli-
sion energy.
Before the method could be studied in more detail, the sta-
bility of the cumbersome H2S and CH3SH in aqueous solution
was evaluated. Freshly prepared solutions at pH 3.0 were split
into five headspace vials that were analysed at different times.
The solutions can be considered stable for a maximum of 2 h
with a loss not greater than 5% (Figure 1). Even if the analyti-
cal method can be fully automated, the lack of stability of stan-
dard solutions is a serious limitation, precluding a long series
of overnight calibrations and measurements.
Wang et al.[14] determined thiols in liquids, but also in solid
samples, by using N-phenylmaleimide as an in-fibre derivatisa-
tion reagent. The fibre was first exposed to the headspace of
the reagent, and then to the headspace of the analyte. As no
extraction step had to be performed, the process could be au-
tomated. CH3SH, ethanethiol, propanethiol and butanethiol
could be quantified with LODs of about 1 mgLÀ1. H2S was not
included in the study. The method was further improved by
the use of a needle-trap device, and many thiols were quanti-
fied in garlic.[15]
We decided to investigate the use of NEM 1 with in-fibre de-
rivatisation. Because its molecular weight is lower than that of
N-phenylmaleimide, the double adduct of H2S NEM2ÀS 2 was
anticipated to be easier to analyse by GC. The first preliminary
trials in which with the SPME fibre was exposed to the head-
space of NEM crystals were not successful. When the fibre was
first exposed to a mixture of triethylamine (TEA) and NEM,
then to the headspace of a H2S solution, the reaction was very
effective and proceeded to the bis-adduct 2. TEA may act as
a base catalyst in the Michael addition, and may also improve
the extraction of the weak acid H2S by the SPME fibre. With
the optimal ratio of the mixture TEA/NEM (7:3), we did not ob-
serve the monoadduct NEMÀSH (Scheme 1). The in-fibre reac-
Figure 1. Stability of H2S and CH3SH over time in aqueous solution (pH 3.0).
Working with this limitation, the method was validated ac-
cording to the International Conference on Harmonisation of
Technical Requirements for Registration of Pharmaceuticals for
Human Use (ICH) guidelines.[16] Carryover was estimated by an-
alysing a series of empty headspace vials after the injection of
a 1 mgLÀ1 solution of H2S or CH3SH, that is, the maximum con-
centration in the calibration range. Peaks corresponding to
around 200 ngLÀ1 for NEM2ÀS but less than 10 ngLÀ1 for
NEMÀSCH3 were obtained in the first blank run after the analy-
sis of a 1 ppm solution of H2S. This is unfortunate, but it corre-
sponds to only 0.02% of carryover. The user should take this
into account and include several blank runs after the highest
concentration range of the calibration.
Generally, the limit of detection of an analyte is defined as
the lowest concentration that gives a peak with a certain
signal-to-noise ratio. For this study, we adapted the Interna-
tional Organization of the Flavor Industry rules for GC–MS
quantification.[17] Determination of LODs was based on the
area ratios between the transitions 284/158.1 and 284/99.2 for
NEM2ÀS 2 and between 173/99.2 and 173/84.1 for NEM-S-CH3
3. The ratios were first determined at different concentrations
by using calibration solutions, and LOD was estimated by
using validation solutions of different concentrations. To avoid
problems of carryover, the solutions were analysed in order of
increasing concentration and after obtaining a clean blank run
Scheme 1. In-fibre derivatisation of H2S and CH3SH with N-ethylmaleimide
(NEM).
tion proceeded equally well with CH3SH to give NEMÀSCH3 3,
as expected. The electron-impact (EI) mass spectra of NEM2ÀS
2 and NEM-S-CH3 3 were in accordance with those in the litera-
ture.[13] The PDMS fibre was used because of its high loading
capacity.
To increase the selectivity and sensitivity further, we devel-
oped a GC–MS-MS method using a triple quadrupole mass
spectrometer in the multiple reaction monitoring (MRM) mode.
For both NEM2ÀS 2 and NEMÀSCH3 3, the molecular ion in EI
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemPlusChem 2014, 79, 77 – 82 78

CHEMPLUSCHEM
FULL PAPERS
www.chempluschem.org
if needed. The LOD was 100 ngLÀ1 for H2S and 10 ngLÀ1 for
CH3SH. This is better than the method with N-phenylmalei-
mide.[14] We could detect those compounds at concentrations
close to or below their olfactive perception threshold, which
we roughly estimated to be 100 and 50 ngLÀ1 for H2S and
CH3SH, respectively. This is the result of the high concentration
factor on the SPME fibre and the high sensitivity of the MS de-
tector in the MRM mode.
The validated method presented above was used to demon-
strate this assumption. In separate experiments, buffered solu-
tions (pH 3.0, typical pH of fabric softeners) of H2S and CH3SH
at 1 mgLÀ1 were split into many headspace vials. A concentra-
tion of 100 mgLÀ1 of 4 was added, or not added, to the vials.
Note than this concentration of 100 mgLÀ1 of 4 is typical in
fabric softeners and that it represents 20 molar equivalents
with respect to 1 mgLÀ1 of H2S. The concentration of the mer-
captans in each vial was measured as a function of time. The
point at t=0 could not be obtained because of the duration
of the SPME derivatisation process, so that the first point of
the kinetics was measured exactly 12 min after the solutions
were prepared. Two replicates were performed and the data
are shown in Figure 2. It is clear that the decay of the H2S con-
centration was accelerated by the presence of 4. A similar
effect was observed for CH3SH, with the scavenging being
slower (Figure 3).
The method was found to be linear over a range of 1 to
1000 mgLÀ1, with the calculated amounts not differing by more
than 5% from the actual amounts. The intermediate precision
was checked over three different days for both compounds
over the whole calibration range (Table 1). For both com-
Table 1. Intermediate precision of the method for H2S and CH3SH in the
linear range of the method (1–1000 mgLÀ1).
We showed that 4 is a very efficient scavenger of H2S and
CH3SH since less than 4% of these malodour compounds re-
Concentration
Precision [%]
[a]
[mgLÀ1
]
H2S
CH3SH
1
10
19
4
31
5
100
1000
1
4
2
4
[a] In aqueous solution.
pounds, the intermediate precision of the method was excel-
lent, being in the 10–1000 mgLÀ1 range (CVR <5%). This shows
that the preparation of standard solutions, the SPME derivatisa-
tion and the GC–MS measurement were all reproducible.
The limits of quantification (LOQs) were estimated from the
accuracy profiles with a tolerated bias of Æ10%. For H2S and
CH3SH, the LOQ was 10 mgLÀ1. Considering the accuracy and
the linearity profiles, the range of the method was 10–
1000 mgLÀ1 for both analytes (Table 2).
Figure 2. Stability of H2S over time in buffer solution (pH 3.0) with/without
4. The relative MS area is normalised to 100% for the buffer solution con-
taining no benzisothiazolone. Two replicates of each experiment are shown.
Table 2. Summary of validation parameters of the method.
H2S[a]
100 ngLÀ1
10 mgLÀ1
1–1000 mgLÀ1
10–1000 mgLÀ1
0.02%
CH3SH[a]
10 ngLÀ1
10 mgLÀ1
1–1000 mgLÀ1
10–1000 mgLÀ1
<0.01%
LOD
LOQ
linearity
range
carryover
[a] Concentrations in aqueous solution.
During a study on fabric softeners,[18] we observed that cer-
tain products never developed rotten-egg malodours. Exami-
nation of their composition made us suspect 1,2-benzisothia-
zol-3(2H)-one (4), a well-known biocide commercialised as
Proxel GXL, to be the cause of the malodour decay.
Figure 3. Stability of CH3SH over time in buffer solution (pH 3.0) with/with-
out 4.
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemPlusChem 2014, 79, 77 – 82 79

CHEMPLUSCHEM
FULL PAPERS
www.chempluschem.org
Figure 4. Stability of H2S over time in buffer solution (pH 3.0) with/without
5.
mained after 6 h of contact. The analogous biocide 2-methyl-4-
isothiazolinone (5) was much less efficient since more than
30% of H2S remained after 6 h (Figure 4). Similar results were
obtained for CH3SH.
In the literature, biocide 4 was shown to target intracellular
thiols of bacteria such as glutathione or the cysteine residues
of proteins.[19] In addition, aliphatic thiols have been shown to
react with N-substituted 1,2-benzisothiazol-3-ones to open the
heterocyclic ring and yield a disulfide.[20,21] The reaction of H2S
with isothiazol-3-ones was never described.
To elucidate the mechanism of the trapping, we reacted mil-
ligram quantities of H2S and a twofold excess of 4 in a buffer
at pH 3. A white precipitate formed almost immediately and
was identified as 2,2’-disulfanediyldibenzamide (6) by NMR
spectroscopy and MS (Scheme 2). The filtered solution was ex-
tracted with ethyl acetate and analysed by GC–MS. Benzodi-
thiolone 7 was identified by means of its mass spectrum and
by comparison with the commercially available product. When
the same reaction solution was analysed by ultra-performance
liquid chromatography (UPLC)-UV-MS, product 7 was not pres-
ent (Figure 5), which indicated that it formed in the hot GC in-
jector.
Scheme 2. Reaction of 1,2-benzisothiazol-3(2H)-one (4) on H2S and CH3SH.
359 [M+Na+]. A product-ion spectrum of ion 337 showed two
fragments at 152 and 109, respectively. This product was thus
tentatively identified as 2,2’-trisulfanediyldibenzamide (8).
Peak 2 presented a UV absorption at 249 and 284 nm and
a molecular ion at m/z 310, but could not be identified.
The presence of 6, 7 (in GC) and 8 led us to propose the
mechanism of Scheme 2 for the trapping of H2S by 4. The re-
action of 4 with CH3SH was analysed by GC–MS, and product 9
was tentatively identified on the basis of its EI mass spectrum.
Product 10 could well have formed by nucleophilic attack of
4 by H2S. The intramolecular ring-closing reaction to give 7 oc-
Besides the peak of the remaining reagent 4, two other
peaks were present (Figure 5). Peak 1 displayed a UV spectrum
showing two maxima at 249 and 300 nm, a molecular ion in
positive electrospray ionisation (ESI) at m/z 337 [M+H+] and
Figure 5. UPLC-UV trace of the reaction of 4 with H2S in buffer (pH 3).
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemPlusChem 2014, 79, 77 – 82 80

CHEMPLUSCHEM
FULL PAPERS
www.chempluschem.org
and then a gradient of 68CminÀ1 to 2408C, followed by a 5.7 min
isotherm step. The injection was performed in splitless mode. MS
conditions: MRM mode in EI (70 eV). Transitions: 173/127.1 (quanti-
fier, 3 V collision energy, CE); 173/99.2 (10 V CE) and 173/84.1 (10 V
CE) for NEM-S-CH3 3 with a gain set at 1. Transitions: 284/127
(quantifier, 9 V CE), 284/158.1 (1 V CE) and 284/99.2 (15 V CE) for
NEM2ÀS 2 with a gain set at 100. Peaks were integrated using
MassHunter.
curred only in the GC injector, but it may constitute indirect
evidence of 10. Precipitate 6 and product 8 could have formed
from intermediate 11. It is interesting to notice that the reac-
tion of 4 with CH3SH did not form any precipitate 6.
Conclusion
We developed a very selective and sensitive SPME-GC–MS
method for the quantification of H2S and CH3SH in aqueous so-
lution using external calibration. The method can also be used
for more difficult matrices such as fabric softeners or foods.
However in this case, the standard addition method rather
than external calibration would have to be applied for true
quantitative measurements. This method was used to demon-
strate that biocides of the isothiazolinone family are able to
scavenge these thiols, thereby preventing malodour formation
in the industrial products in which they are used.
Validation of the method
The method was validated according to ICH requirements.[16] LOD
determination was based on the concordance of nonpolar linear
retention indices[22] (Æ10 units) and good transition ratios (Æ20%):
173/99.2 and 173/84.1 for NEM-S-CH3 and 284/158.1 and 284/99.2
for NEM2ÀS. The method linearity was checked by analysis of cali-
bration and validation solutions, as well as verification that the
concentrations calculated with the calibration curve were directly
proportional to the concentration of H2S or CH3SH in the sample
(R2 >0.994). Intermediate precision (CVR) was estimated over 3 days
with the preparation of fresh calibration and validation solutions.
The LOQs were determined by using the accuracy profile as the
minimum amount of compound that could be quantified with
good accuracy (relative bias Æ10%) and good precision (CVR <
10%). The range of the method was defined as the linear part of
the method starting from the LOQ in which both the accuracy and
the precision of the method were sufficient. The carryover of the
method was estimated by successive blank injections after the
analysis of a 1 mgLÀ1 solution of H2S or CH3SH.
Experimental Section
Reagents, standards, buffer, calibration and validation solu-
tions
CH3SH was available in-house (Firmenich S.A., Geneva, Switzerland)
as a solution in ethanol. It was titrated as 0.33% with Ellman’s re-
agent (5,5’-dithiobis(2-nitrobenzoic acid), Sigma–Aldrich, Buchs,
Switzerland). The mother solution for H2S was made from sodium
hydrogen sulfide (100.8 mg, anhydrous, Alfa Aesar, Karlsruhe, Ger-
many) in 0.05m sodium hydroxide (100 mL). This solution was ti-
trated using Ellman’s reagent and was at 917 mgLÀ1 of NaSH thus
equivalent to 557 mgLÀ1 of H2S. Then the mother solutions of H2S
and CH3SH were diluted using micropipettes (Socorex, Ecublens,
Switzerland) in a buffer solution at pH 3 prepared from Titrisol
phosphate concentrate (Merck, Darmstadt, Germany). The calibra-
tion solutions and validation solution should not be the same for
the validation of the method. NEM 1, triethylamine and 2-methyl-
4-isothiazolinone hydrochloride (5) were purchased from Sigma–Al-
drich. 1,2-Benzisothiazol-3(2H)-one (4) was purchased from the
Tokyo Chemical Industry Co. Ltd. (Tokyo, Japan).
UPLC-UV-MS method
An Acquity UPLC system (Waters Corp., Milford, MA) was used with
a HSS-T3 column (100ꢁ2.1 mm i.d., 1.8 mm) operated at 408C. The
injection volume was 1 mL. The solvents were as follows: A: water,
0.1% HCOOH; B: acetonitrile, 0.1% HCOOH. Flow: 0.3 mLminÀ1
The detection was performed with the diode array detector (DAD)
at 250–490 nm. The solvent program was set as follows: 10% B, 0–
0.5 min; 10–90% B, 0.5–9 min; 100%, 9–11 min; equilibration at
10% B for 1 min.
.
The MS analyses were performed in ESI positive mode with a TSQ
Quantum Ultra AM triple-quadrupole MS (Thermo Scientific, Basel,
Switzerland) equipped with an H-ESI II source and operated with
XCalibur as follows: spray voltage 3000 V, vaporiser temperature
3008C, sheath gas pressure 50 (arbitrary unit), auxiliary gas 15,
heated capillary temperature 3508C, skimmer offset 5 V. MS and
MS/MS were triggered automatically using the data-dependent
mode. CID argon gas pressure: 1 mtorr. Scan event 1: detection in
Q3 in full-scan mode with a mass range between 130 and 1000.
Scan events 2 and 3: data-dependent MS/MS of the first most-in-
tense ion of scan 1 at 25 and 35 V of collision energy.
SPME and in-fibre derivatisation
A PDMS fibre (100 mm) (Supelco, Buchs, Switzerland) was used. The
derivatisation was performed using a PAL autosampler (CTC, Zwin-
gen, Switzerland) with a cycle composer/macro editor. The reagent
NEM/TEA consisted of a weekly prepared 30:70 w/w solution
(400 mL) of NEM and triethylamine placed in a headspace vial
(volume of 20 mL). The analyte consisted of buffer solution (1 mL)
placed in a headspace vial (20 mL). The fibre was exposed to the
headspace of the NEM/TEA reagent for 30 s, and then to that of
the analyte for 10 min prior to desorption in the GC injector at
2808C.
NMR spectroscopy
1H and 13C NMR spectra were recorded with a 500 MHz Bruker
Avance (Bremen, Germany) spectrometer at 500.13 and
125.76 MHz, respectively. Chemical shifts (d) and coupling con-
stants (J) are expressed in parts per million and hertz, respectively.
The solvent was [D6]DMSO with tetramethylsilane as the internal
standard. The 13C signal assignments were obtained from standard
gradient-selected correlated spectroscopy, heteronuclear single-
quantum coherence and heteronuclear multiple-bond correlation
GC–MS method
Analyses were performed using a GC 7890A coupled to a triple-
quadrupole mass spectrometer 7000B (Agilent Technologies, Palo
Alto, CA), with a HP1 inert nonpolar column (30 m, 0.25 mm inside
diameter (i.d.), 0.25 mm film thickness). The flow rate of helium,
used as the carrier gas at constant flow, was set at 1.2 mLminÀ1
The oven temperature was set as follows: 1008C, 1 min isotherm,
.
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemPlusChem 2014, 79, 77 – 82 81

CHEMPLUSCHEM
FULL PAPERS
www.chempluschem.org
experiments. NMR spectra were processed with Bruker TopSpin 2.0
(s, singlet; d, doublet; t, triplet; m, multiplet).
Keywords: desulfurization
spectrometry · mercaptan · trace analysis
· gas chromatography · mass
Kinetics of H2S and CH3SH decay in buffer solution
The degradation was evaluated on 1 ppm solutions of H2S or
CH3SH in pH 3 buffer solution over nine injection events (nine vials
prepared at the same time) corresponding to a total of 348 min.
The first injection (12 min) was considered as the reference and
the results were expressed as a relative MS area compared with
this reference (set at 100%). The same experiments were repeated
in the presence of 0.01% of 4 or 5.
[1] http://www.telegraph.co.uk/earth/earthnews/9818749/French-gas-
cloud-causes-stink-across-the-South-East.html; accessed March, 2013.
[2] P. Hughes, Cerevisia 1999, 24, 21–25.
[3] Y. Fang, M. C. Qian, J. Chromatogr. A 2005, 1080, 177–185.
[4] These compounds are generally recognised as safe (GRAS): FEMA N8
3779 (H2S) and N8 2716 (CH3SH). They can be used as food flavourings.
[5] A. Tangerman, J. Chromatogr. B 2009, 877, 3366–3377.
[6] W.-H. Chen, J.-H. Lin, Y.-C. Lin, Aerosol Air Qual. Res. 2012, 12, 1386–
1397.
Experiments toward the determination of the scavenging
mechanism
[7] M. Mieth, J. K. Schubert, T. Groger, B. Sabel, S. Kischkel, P. Fuchs, D.
Hein, R. Zimmermann, W. Miekisch, Anal. Chem. 2010, 82, 2541–2551.
[8] K.-H. Kim, Atmos. Environ. 2005, 39, 2235–2242.
[9] M. A. H. Khan, M. E. Whelan, R. C. Rhew, Talanta 2012, 88, 581–586.
[10] F. Lestremau, V. Desauziers, J. C. Roux, J. L. Fanlo, J. Chromatogr. A 2003,
999, 71–80.
[11] K. Shimada, K. Mitamura, J. Chromatogr. B 1994, 659, 227–241.
[12] L. Mateo-Vivaracho, J. Cacho, V. Ferreira, J. Chromatogr. A 2007, 1146,
242–250.
[13] C. Salgado-Petinal, R. Algaza, C. Garcꢂa-Jares, M. Llompart, J. M. Bayona,
Anal. Chem. 2005, 77, 6012–6018.
[14] Q. Wang, J. M. Chong, J. Pawliszyn, Flavour Fragrance J. 2006, 21, 385–
394.
[15] J. M. Warren, D. R. Parkinson, J. Pawliszyn, J. Agric. Food Chem. 2013, 61,
492–500.
[16] International conference of harmonisation (ICH) of technical require-
ments for registration of pharmaceuticals for human use, topic Q2 (R1):
Validation of analytical procedures: Text and methodology Q 2(R1),
Geneva. 2005.
[17] IOFI (International Organization of the Flavor Industry) Working Group
on Methods of Analysis, Flavour Fragrance J. 2012, 27, 224–226.
[18] D. Reichlin, E. Frerot, P. L. Schneider (Firmenich S.A.), PCT Int. Appl., WO
2012113746, 2012.
[19] S. J. Fuller, S. P. Denyer, W. B. Hugo, D. Pemberton, P. M. Woodcock, A. J.
Buckley, Lett. Appl. Microbiol. 1985, 1, 13–15.
1,2-Benzisothiazolone 4 (51.6 mg, 0.34 mmol) was diluted in pH 3
buffer solution (100 mL) using the ultrasound bath for at least 1 h.
Then solid NaSH (9.2 mg, 0.16 mmol) was added. This corresponds
to a 2.1-fold excess amount of 4. Almost immediately after the ad-
dition of NaSH, the solution that was totally clear became cloudy
(white). After standing overnight at RT, precipitate 6 was collected.
The pH of the solution was still 3. The extraction was performed
with ethyl acetate (3ꢁ5 mL), dried over anhydrous sodium sulfate,
concentrated under an argon flow and injected onto the GC–MS.
For LC/MS analysis, the solution was diluted 10 times in ethanol
and filtered over a 0.45 mm regenerated cellulose syringe filter
before injection. The same reaction was carried out with CH3SH,
but with equimolar concentrations with respect to 4. No precipi-
tate was observed. 2,2’-Disulfanedyildibenzamide (6):1H NMR: d=
7.28 (td, J=7.6, 1.0 Hz; 2H-C(5,5’)), 7.44 (td, J=7.6, 1.3 Hz; 2H-
C(4,4’), 7.63 (dd, J=7.6, 1.0 Hz; 2H-C(3,3’), 7.72 ppm (dd, J=7.6,
1.3 Hz; 2H-C(6,6’)); 13C NMR: d=125.4 (d; C(3,3’)), 125.6 (d; C(5,5’)),
128.2 (d; C(6,6’)), 131.1 (d; C(4,4’)), 133.1 (s; C(1,1’)), 137.1 (s;
C(2,2’)), 168.8 ppm (s; 2CONH2); ESI: m/z: 305 [M+H+].
[20] M. Bao, M. Shimizu, S. Shimada, J. Inoue, T. Konakahara, Tetrahedron
2004, 60, 11359–11366.
[21] S. Sivaramakrishnan, K. Keerthi, K. S. Gates, J. Am. Chem. Soc. 2005, 127,
10830–10831.
[22] B. d’Acampora Zellner, C. Bicchi, P. Dugo, P. Rubiolo, G. Dugo, L. Mon-
dello, Flavour Fragrance J. 2008, 23, 297–314.
Acknowledgements
Philippe L. Schneider and Daniel Rechlin (Perfumery Division, Fir-
menich) performed the first experiments showing the absence of
malodour in the fabric softeners that contained Proxel. Andreas
Herrmann (R&D, Firmenich) is acknowledged for fruitful discus-
sions.
Received: June 19, 2013
Published online on October 11, 2013
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemPlusChem 2014, 79, 77 – 82 82