Structure, stability, and generation of CH3CNS

Article abstract of DOI:10.1071/CH10303

The unstable acetonitrile N-sulfide molecule CH₃CNS has been photolytically generated in inert solid argon matrix from 3,4-dimethyl-1,2,5- thiadiazole by 254-nm UV irradiation, and studied by ultraviolet spectroscopy and mid-infrared spectroscopy. The molecule is stable in the matrix to 254-nm UV irradiation, but decomposes to CH₃CN and a sulfur atom when broad-band UV irradiation is used. Chemiluminescence due to S₂ formation from triplet sulfur atoms was detected on warming the matrix to ～20-25K. The ground-state structure and potential uni- and bimolecular reactions of CH₃CNS are investigated using B3LYP, CCSD(T), and MR-AQCC quantum-chemical methods. CH₃CNS is demonstrated to be stable under isolated conditions at room temperature, i.e. in the dilute gas phase or in an inert solid matrix, but unstable owing to bimolecular reactions, i.e. in the condensed phase. CSIRO 2010.

Full text of DOI:10.1071/CH10303

RESEARCH FRONT
Full Paper
CSIRO PUBLISHING
Aust. J. Chem. 2010, 63, 1686–1693
www.publish.csiro.au/journals/ajc
Structure, Stability, and Generation of CH3CNS
A
B
A
A C
,
Melinda Krebsz, Bala´zs Hajgato´, Ga´bor Bazso´, Gyorgy Tarczay,
¨
A C
,
and Tibor Pasinszki
A
Institute of Chemistry, Eotvos Lora´nd University Budapest, PO Box 32,
¨
¨
H-1518 Budapest 112, Hungary.
B
Department of Chemistry, Hasselt University, Research Group of Theoretical Chemistry
and Molecular Modelling, Agoralaan, Gebouw D, B-3590 Diepenbeek, Belgium.
C
Corresponding authors. Email: pasinszki@chem.elte.hu; tarczay@chem.elte.hu
The unstable acetonitrile N-sulfide molecule CH3CNS has been photolytically generated in inert solid argon matrix from
3,4-dimethyl-1,2,5-thiadiazole by 254-nm UV irradiation, and studied by ultraviolet spectroscopy and mid-infrared
spectroscopy. The molecule is stable in the matrix to 254-nm UV irradiation, but decomposes to CH3CN and a sulfur atom
when broad-band UV irradiation is used. Chemiluminescence due to S2 formation from triplet sulfur atoms was detected
on warming the matrix to ,20–25 K. The ground-state structure and potential uni- and bimolecular reactions of CH3CNS
are investigated using B3LYP, CCSD(T), and MR-AQCC quantum-chemical methods. CH3CNS is demonstrated to be
stable under isolated conditions at room temperature, i.e. in the dilute gas phase or in an inert solid matrix, but unstable
owing to bimolecular reactions, i.e. in the condensed phase.
Manuscript received: 13 August 2010.
Manuscript accepted: 23 September 2010.
Introduction
bond, and suggests that the instability is possibly not a uni-
molecular process. The identification of CH3CNS is based on a
single IR band at 2230 cmꢁ1 and on mass spectrometric mon-
itoring of the FVP products of 5-methyl-2,2-bis(trichloro-
methyl)-1,3,4-oxathiazole.[7] No other spectroscopic data exist
for CH3CNS, and the experimental structure of this molecule is
unknown.
Nitrile sulfides (X–CꢀN-S) are well-known transient species
in synthetic organic chemistry.[1–4] They are unstable in the
pure state and in solutions; therefore, their direct spectroscopic
identification is challenging. Nitrile sulfides have never been
detected by spectroscopic methods in the pure condensed phase
or in solution, and only four derivatives, viz. HCNS,[5] FCNS,[6]
CH3CNS,[7] and PhCNS,[7,8] have been identified in low-tem-
perature inert matrices by IR spectroscopy so far. The identifi-
cation of PhCNS has also been supported by matrix-isolation
UV spectroscopy.[8] It is important to note that small nitrile
sulfides (X ¼ H, Cl, Br, I, NH2, CH3S, CH3, Et, Pr, Ph) have
been generated and identified as neutral species in the gas phase
by neutralization–reionization mass spectroscopy,[7,9,10] which
established the existence of these species with lifetimes of at
least microseconds in the dilute gas phase. Nitrile sulfides are
believed to be thermally unstable, readily losing sulfur. The
mechanism of this process has been the subject of some debate
and is not yet understood, although it has been realized that the
decomposition of nitrile sulfides is not a simple unimolecular
reaction;[1,2] it has been suggested that the decomposition is
accelerated by short sulfur chains and initial S2 formation may
occur via dimerization or sulfur atom transfer mechanisms.[2]
Acetonitrile N-sulfide, CH3CNS, has never been isolated
or detected in solutions. This molecule was identified by IR
spectroscopy among the flash vacuum pyrolysis (FVP) products
of 5-methyl-2,2-bis(trichloromethyl)-1,3,4-oxathiazole in an
argon matrix at 12 K.[7] Although the yield for CH3CNS in this
thermal process is very low, the fact that small amounts of this
nitrile sulfide survived the hot zone of the high-temperature
furnace at 300–5008C indicates a substantial strength of the N–S
Owing to the paucity of experimental investigations, theore-
tical calculations can provide information about the structure
and stability of CH3CNS. Ab initio MP2[10,11] calculations and
density functional theory using the B3LYP functional[11,12] have
predicted a linear CCNS frame. A recent theoretical study[11] of
the global potential energy surface of the [CH3,N,C,S] system at
the B3LYP, MP2, and QCISD levels has predicted that CH3CNS
is thermodynamically the third most stable isomer, following its
isothiocyanate and thiocyanate isomers, and as far as unimole-
cular isomerization is concerned, the molecule is stable at room
temperature. Calculations at the G2(MP2),[10] QCISD//B3LYP,[11]
and CCSD(T)//B3LYP[12] levels have demonstrated that direct
sulfur atom formation from CH3CNS cannot happen at low or
ambient temperatures, because the spin-allowed extrusion of a
singlet sulfur atom requires quite high energies of 244, 210, and
245 kJ molꢁ1 respectively. We note, however, that the RHF
wave function is unstable for CH3CNS (see below); thus earlier
MP2 calculations must be treated with care.
In this paper, we present an experimental study on the
generation of CH3CNS in an argon matrix from a novel
precursor and its IR and UV spectroscopy, as well as a quan-
tum-chemical study on the equilibrium structure and stability.
3,4-Dimethyl-1,2,5-thiadiazole is considered as a potential pre-
cursor based on our recent success of generating the parent
Ó CSIRO 2010
10.1071/CH10303
0004-9425/10/121686

RESEARCH FRONT
Structure, Stability, and Generation of CH3CNS
1687
thiofulminic acid[5] and its fluoro derivative[6] from the corre-
sponding thiadiazole. We note that we have addressed pre-
viously the decomposition mechanism of nitrile sulfides at the
B3LYP/6–31G** level;[12] however, possible open-shell (sing-
let diradical) structures and synchronous cycloaddition reac-
tions between nitrile sulfides were not considered. Of special
computational interest here are the structure and potential
mechanisms for the initiating step of the decomposition of
CH3CNS.
N
ΔG Њ
kJ molϪ1
C
C
0K
S
N
S
1MeCN ϩ 1S
C
C
4
3MeCNS
200
119
N
C
C
S
230
100
0
1MeCN ϩ 3S
142
176
Results and Discussion
Structure and Stability of CH3CNS
Structure
1MeCNS
1MeC(NS)
C
C
N
S
The careful selection of quantum-chemical methods to
describe the electronic and geometric structure of nitrile sulfides
is important. The number of available quantum-chemical meth-
ods has been narrowed by detecting that the HF wave function is
unstable for CH3CNS (RHF–UHF instability); e.g. MPn calcu-
lations were rejected in this work. The CCSD(T) energy using
the RHF-unstable solution has been found to be identical to that
using the UHF solution; thus CCSD(T) calculations based on
RHF wave functions were performed. To test the reliability of
the single-reference approximation, CASSCF(8,8) single-point
calculation was done. The latter indicated that apart from the
main HF configuration (weight 93%), there is no other important
configuration, and thus single-reference post-HF methods seem
to be sufficient for describing the electronic structure (calculated
weights of all other configurations are smaller than 2%). The
CCSD T1 diagnostic[13] for CH3CNS provided a value of 0.019,
below the limiting value of 0.020, thus suggesting that non-
dynamical electron correlation has minor importance and con-
firming the adequacy of single-reference post-HF methods.
The B3LYP wave function was found to be stable. Therefore,
B3LYP and CCSD(T) methods were used for calculating the
structure of CH3CNS.
N
C
C
Ϫ100
S
Fig. 1. Structure and lowest-energy unimolecular isomerization and bond
dissociation of ground-state and excited state CH3CNS. Calculated at the
CCSD(T)//B3LYP/cc-pVTZ level.
acetonitrile and ground-state triplet sulfur, S(3P), in this process
is spin-forbidden. According to our calculations, both the iso-
merization and the bond rupture require quite high energies
of 142 and 230 kJ molꢁ1; thus calculations demonstrate that
CH3CNS must be stable at low or ambient temperatures, as far as
unimolecular processes are considered. Calculations also show
that the N–S bond easily breaks if the molecule is excited into its
triplet state.
Bimolecular reactions. As the activation energy for unim-
olecular decomposition is quite high for CH3CNS, potential
bimolecular reactions have been considered. Our test calcula-
tions for the selection of the appropriate quantum-chemical
method have indicated that intermediates on the multistep
dimerization paths (ctc, ttc, ttt, tct in Fig. 2) have open-shell
singlet diradical character; thus multireference (MR) calcula-
tions are required to properly describe their electronic structure.
CASSCF(8,8) calculations for these intermediates have indi-
cated nearly equal weight for two configurations, 46–52% for
the HF configuration and 39–45% for another derived by
exciting the two HOMO electrons to LUMO. Considering the
number and size of molecules in this study, MR calculations,
including geometry optimization, however, are not feasible. To
perform geometry optimization at the lower UB3LYP level and
single-point energy calculations at the MR-AQCC level using
UB3LYP geometries seemed to be a reliable alternative. The
UB3LYP method is known to include dynamical, and also some
non-dynamical, electron correlation, and has been found to
provide a good description of geometries for both concerted
and multistep cyclization reactions.[14–18] It is fortunate in the
present study that a limited (2,2) active space seems to be
sufficient to describe the two basic configurations in MR-AQCC
calculations. Therefore, MR-AQCC(2,2)//(U)B3LYP/cc-pVTZ
calculations were performed for investigating bimolecular
reactions.
The structure of CH3CNS is shown in Fig. 1 and calculated
bond lengths and angles using various basis sets are listed in
Table S1 (Accessory Publication). There is a good agreement
between B3LYP and CCSD(T) results, although bond lengths
calculated using the CCSD(T) level are slightly longer than
those of B3LYP. Calculations predict that CH3CNS is a
symmetric-top molecule with a linear CCNS frame (CCSD(T)/
˚
˚
˚
cc-pVTZ: re(C–C) 1.462A, re(C–N) 1.166 A, re(N–S) 1.630A,
˚
re(C–H) 1.091 A, ae(HCC) 109.98, me 6.95 D, Ae 159.0590 GHz,
Be 2.3732 GHz, Ce 2.3732 GHz). The molecule has a singlet
electronic ground state, which is more stable than the triplet by
216 kJ molꢁ1 (DG8 ) at the CCSD(T)//B3LYP/cc-pVTZ level.
0 K
Stability
The stability of CH3CNS at low or room temperature can be
discussed in the context of the lowest-energy path leading to
isomerization, homolytic bond dissociation, or bimolecular
reactions.
Unimolecular reactions. The global potential energy sur-
face of the [CH3,N,C,S] system was calculated recently[11] (see
Introduction), and the relevant part of the potential energy
surface is recalculated in the current work at the CCSD(T)//
B3LYP/cc-pVTZ level. The lowest-energy isomerization path,
leading to a thiazirine derivative, and the lowest-energy homo-
lytic bond dissociation, leading to acetonitrile and singlet sulfur
atom, S(1D), are shown in Fig. 1. The formation of ground-state
After investigating all possible paths on the potential energy
surface, six distinct bimolecular reactions leading to different
products were obtained for the initiating step: three multistep
reactions to 3,4-dimethyl-1,2,5-thiadiazole-2-sulfide and 4,5-
dimethyl-1,2,3,6-dithiadiazine (Fig. 2), two distinct one-step

RESEARCH FRONT
1688
M. Krebsz et al.
ΔGЊ
0K
kJ molϪ1
100
TS1
TS3
53
TS2
TS7
TS6
TS5
tct
TS4
ttc
0
Ϫ100
Ϫ200
2 MeCNS
ctc
ttt
ss
tfur
Fig. 2. Multistep cyclodimerization of CH3CNS to 3,4-dimethyl-1,2,5-thiadiazole-2-sulfide (tfur) and 4,5-
dimethyl-1,2,3,6-dithiadiazine (ss). Gibbs free energies are relative to that of the two monomers. Calculated at the
MR-AQCC(2,2)//(U)B3LYP/cc-pVTZ level.
ΔG 0Њ K
kJ molϪ1
ΔG0Њ K
kJ molϪ1
100
0
100
TSs2
TSs1
TS-t
55, 57
2 MeCNS
50
0
Ϫ100
Ϫ200
2 MeCNS
Ϫ100
Ϫ200
2 MeCN ϩ 1S2
SP2
SP1
Fig. 3. Concerted [3þ2] and [3þ3] cyclodimerization of CH3CNS to 3,5-
dimethyl-1,2,4-thiadiazole-4-sulfide (SP1) and to 3,6-dimethyl-1,4,2,5-
dithiadiazine (SP2). Gibbs free energies are relative to that of the two
monomers. Calculated at the MR-AQCC(2,2)//(U)B3LYP/cc-pVTZ level.
Fig. 4. Bimolecular decomposition of CH3CNS to acetonitrile and S2.
Gibbs free energies are relative to that of the two monomers. Calculated at
the MR-AQCC(2,2)//(U)B3LYP/cc-pVTZ level.
singlet diradical ctc intermediate. The kinetic energy barrier
concerted reactions leading to 3,5-dimethyl-1,2,4-thiadiazole-
4-sulfide and 3,6-dimethyl-1,4,2,5-dithiadiazine (Fig. 3), and
one decomposition reaction leading to acetonitrile and singlet S2
(Fig. 4). Total energies and relative Gibbs free energies (DG8 )
0 K
of all minima and transition states are given in Table S2
(Accessory Publication).
The three multistep dimerization paths are shown in Fig. 2.
Three letters, e.g. ctc (cis–trans–cis), are used to define the
relative conformation of the two NS groups with respect to the
(S)N–C, C–C, and C–N(S) bonds. The rate-determining step for
dimerization is the initial C–C bond formation at TS1, forming a
at TS1 is DG8 ¼ 53 kJ molꢁ1 (DG8
¼ 92 kJ molꢁ1). Once
0 K
298 K
ctc is formed, three diradical pathways are possible with
relatively similar energetics, implying that all would be feasible.
The pathways are identified below, for convenience, by the final
transition state (TS) (TS4, TS5, or TS7). The path via TS4
(TS1 - ctc - TS4 - ss) following formation of the ctc inter-
mediate involves rotation around the established C–C bond
over a relatively low barrier to form the ss six-membered ring
4,5-dimethyl-1,2,3,6-dithiadiazine. The TS5 pathway (TS1 -
ctc - TS2 - ttc - TS5 - tfur) leads to 3,4-dimethyl-1,2,5-
thiadiazole-2-sulfide, after only one inversion at the nitrogen

RESEARCH FRONT
Structure, Stability, and Generation of CH3CNS
1689
A
atom and consecutive rotation around the C–C bond over two
relatively low barriers at TS2 and TS5. The TS7 path (TS1 -
ctc - TS2 - ttc - TS3 - ttt - TS6 - tct - TS7 - tfur)
involves, once ctc is formed, sequential inversion at the
nitrogen atoms to form the ttc (via TS2) and ttt (via TS3)
intermediates, the latter then rotating around the C–C bond
(via TS6) to give the tct intermediate. The final step via
TS7 is then nitrogen inversion followed by completion of
the reaction to 3,4-dimethyl-1,2,5-thiadiazole-2-sulfide.
We obtained two concerted head-to-tail reaction paths lead-
ing to CH3CNS dimers, one of them a [3þ2] dimerization giving
3,5-dimethyl-1,2,4-thiadiazole-4-sulfide SP1 and the other a
[3þ3] process to 3,6-dimethyl-1,4,2,5-dithiadiazine SP2 (Fig. 3).
These molecules, including TSs, have closed-shell electronic
structures. The TSs for both processes suggest a charge-controlled
1.5
3,4-Dimethyl-1,2,5-thiadiazole
1.0
0.5
0.0
1.5
After 254-nm photolysis
1.0
0.5
mechanism with barriers of DG8 ¼ 55kJ molꢁ1 (DG8
¼
0K
298 K
100 kJmolꢁ1) for the thiadiazole-4-sulfide (TSs1) and DG8
¼
0K
0.0
1.5
57kJ molꢁ1 (DG8
(TSs2).
¼101 kJ molꢁ1
)
for the dithiadiazine
298 K
After broad-band UV photolysis
A bimolecular decomposition reaction via S–S bond forma-
tion was also obtained (Fig. 4), which is kinetically more
favoured than the cycloadditions above. In this reaction,
two molecules of CH3CNS produce two molecules of
nitrile (CH3CN) and singlet S2 directly in a ‘tail-to-tail’
1.0
0.5
0.0
(CH3CNSꢂꢂꢂSNCCH3) reaction after passing over
a
¼
kinetic energy barrier of DG8 ¼ 50 kJ molꢁ1 (DG8
0 K
298 K
200
250
300
350
400
77 kJ molꢁ1). There is no intermediate in this reaction, and the
two N–S bonds simultaneously break, releasing S2 and the
nitrile. We note that S2 and nitrile formation via dimerization
of nitrile sulfides has been suggested before.[2]
Wavelength [nm]
Fig. 5. UV spectrum of the 3,4-dimethyl-1,2,5-thiadiazole precursor in
Ar matrix (top panel), the UV spectrum recorded after 450-min photolysis
at 254 nm (middle panel), and the UV spectrum observed after 120-min
consecutive broad-band UV photolysis (bottom panel). A, absorbance.
The kinetic energy barriers for all six initial bimolecular
reactions are relatively low (DG8 ¼ 50–57 kJ molꢁ1) and
0 K
energy differences in barriers are relatively small, which
suggests competing reactions and spontaneous dimerization
at room or higher temperatures. The whole decomposition
mechanism, however, is expected to be much more complicated,
because all products of initial bimolecular reactions, in princi-
ple, can react further and can also take part in the desulfuration
process of CH3CNS. The investigation of such bimolecular
reactions is beyond the scope of this work. Analogous reactions
for the computationally more tractable hydrogen derivative,
HCNS, are discussed in our previous paper.[12]
was co-condensed with Ar at 8 K and this matrix was irradiated
with 254-nm UV light in an analogous way as described
previously for HCNS and FCNS.[5,6] A new band emerged at
209 nm at the expense of the precursor, which was bleached on
broad-band UV irradiation. According to our IR investigations
(see below), this band belongs to CH3CNS. We note that this
band cannot be due to the formation of a sulfur atom, because
the latter has no absorption at this wavelength,[21] an atomic
absorption or emission line should be narrower, and the band
disappears on broad-band UV irradiation.
According to IR investigations, the 254-nm UV irradiation of
the precursor evidently led to the formation of CH3CNS (Fig. 6).
Fig. 6 shows the difference spectrum obtained by subtracting the
spectrum measured before the photolysis from that measured
after the photolysis, as well as the computed IR spectra of
CH3CNS and CH3CN monomers. There is a good agreement
between experimental and computed vibrational frequencies
and relative IR intensities (Table 1), which clearly indicates
that most of the spectral lines that appeared after the 254-nm
photolysis can be assigned to the fundamental vibrations of
these two products, and the most intense lines belong to
CH3CNS. Table 1 lists experimental and calculated vibrational
frequencies and infrared intensities for CH3CNS.
Generation, Identification, and IR and UV Spectra
of CH3CNS
As calculations have predicted that CH3CNS is stable under
isolated conditions, namely bimolecular reactions are hindered,
a possible route to CH3CNS seemed to be photolytic CH3CN
elimination from 3,4-dimethyl-1,2,5-thiadiazole in an inert low-
temperature matrix. There is no chemical reaction with primary
bond formation between CH3CNS and CH3CN at low tem-
perature; the barrier for the lowest-energy cycloaddition
reaction is DG8 ¼ 57 kJ molꢁ1 (DG8
¼ 99 kJ molꢁ1),
0 K
298 K
according to MR-AQCC(2,2)//B3LYP/cc-pVTZ calculations.
Acetonitrile is an ideal side product in the sense that it has only
low-intensity IR bands in the mid-IR[19] and does not absorb in
the UV region above 183 nm.[20] However, one has to keep in
mind that weak interactions between CH3CNS and CH3CN may
be established, because they form in the same cage of the noble
gas matrix.
The strongest IR band appearing after 254-nm photolysis
at 2237 cmꢁ1, together with a lower-intensity sub-band at
2240 cmꢁ1, matches very well with the calculated, 2230 cmꢁ1
,
anharmonic frequency of the n2 (CꢀN stretching) mode of
CH3CNS (see Table 1). This is also in good agreement with
the previous observations of Wentrup and coworkers,[7]
The UV spectrum of 3,4-dimethyl-1,2,5-thiadiazole in an
Ar matrix is shown in Fig. 5, which justifies that photolysis at
254 nm is feasible. Therefore, 3,4-dimethyl-1,2,5-thiadiazole

RESEARCH FRONT
1690
M. Krebsz et al.
suggesting that the IR band emerging at 2230 cmꢁ1 after matrix
isolation of pyrolysis products of 5-methyl-2,2-bis(trichloro-
methyl)-1,3,4-oxathiazole in a 12-K Ar matrix belong to
CH3CNS. The 7 cmꢁ1 (and 10 cmꢁ1) difference between the
former and present observations can be explained by different
molecular environments; both CH3CNS and CH3CN are formed
in the same matrix cage in our experiment; thus weak interaction
should exist between them, whereas CH3CNS was isolated in a
separate Ar cage by Wentrup et al.[7] The appearance of a sub-
band can be explained by different van der Waals complex
formations or by different matrix sites, i.e. a site-splitting effect.
Complex formation has been observed between FCN and FCNS,
generated in a similar manner from difluoro-1,2,5-thiadiazole,
in our previous work.[6]
A
The relative intensity of the three most intense IR bands,
observed at 2240/2237, 1005/1001, and 565 cmꢁ1, as well as
that of the weak band at 2926 cmꢁ1 changes exactly at the same
rate during the 254-nm UV photolysis, suggesting that they
belong to the same species. Band positions and relative IR
intensities are in good agreement with calculated anharmonic
frequencies and IR intensities of CH3CNS (Table 1 and Fig. 6);
thus, their assignment to n1, n2, n4, and n5 vibrational modes
of CH3CNS is straightforward. In addition, two very-weak-
intensity IR bands are detected at 1346 and 1413 cmꢁ1, which
might belong to n3 and n7 modes of CH3CNS; calculated
anharmonic frequencies are 1382 and 1444 cmꢁ1 respectively.
According to our calculations, n6, n8, n9, and n10 fundamental
vibrational modes have low intensity; furthermore, n10 is out of
the investigated region, and thus they are not observed in the Ar
matrix spectrum.
The side photolysis product, CH3CN, has very low IR
transition intensities (Fig. 6). As it forms together with other
photolysis product(s) in the same matrix cage, its infrared
frequencies are perturbed owing to weak intermolecular inter-
actions. Thus, not surprisingly, the most intense IR transitions
of the CH3CN dimer[19] were identified among the very weak
bands at 2256, 1448, 1041, and 916 cmꢁ1. These bands formed
parallel with the formation of CH3CNS (or CH3CNꢂCH3CNS
complex) on 254-nm photolysis. There are several other weak
bands, and weak bands at 2263, 2253, 1048, 1045, and 929 cmꢁ1
may be tentatively assigned to the fundamental vibrations of
CH3CN in the CH3CNꢂCH3CNS complex (see Figs S1 and S2,
Accessory Publication).
Ar matrix
CH3CNS computed
CH3CN computed
3000
2500
2000
1500
1000
500
Wavenumber [cmϪ1
]
Fig. 6. Computed IR spectra (CCSD(T)/cc-pVTZ harmonic frequencies
corrected with B3LYP/cc-pVTZ anharmonic contributions; intensities
obtained at the B3LYP/cc-pVTZ level) of CH3CNS and CH3CN monomers
(bottom panel), and the experimental spectrum (top panel) obtained as the
difference of the spectra of 3,4-dimethyl-1,2,5-thiadiazole in Ar matrix after
and before 254-nm photolysis.
Interestingly, a medium-intensity IR band appeared at
2218 cmꢁ1 on 254-nm UV irradiation of the precursor. This
Table 1. Experimental and calculated anharmonic vibrational frequencies (~v) and infrared intensities (I) of CH3CNS
Vibrational mode Calculated
Experimental (Ar matrix)C
A
~v ½cmꢁ1
ꢃ
I [km molꢁ1 B
]
n1(a1) CH3 s str
n2(a1) CꢀN str
n3(a1) CH3 s def
n4(a1) CC strD
n5(a1) NS strD
n6(e) CH3 as str
n7(e) CH3 as def
n8(e) CH3 rock
n9(e) CNS def
2940 (3050)
2230 (2271)
1382 (1413)
1006 (1020)
580 (574)
12 [18]
383 [295]
8 [9]
2926 vw
2237, 2240E vs
(1346 vw)F
1005, 1001E m
565 w
78 [64]
38 [29]
2 [4]
2984 (3132)
1444 (1482)
1030 (1046)
413 (406)
Not observed
(1413 vw)F
Not observed
Not observed
Not observed
18 [19]
3 [5]
1 [1]
n10(e) CCN def
162 (141)
2 [3]
AAnharmonic frequencies are calculated as the sum of harmonic frozen-core CCSD(T)/cc-pVTZ frequencies (given in brackets) and B3LYP/cc-pVTZ
anharmonic corrections.
BCalculated at the all-electron CCSD(T)/cc-pCVTZ [B3LYP/cc-pVTZ] level using the harmonic approximation.
COwing to the presence of an acetonitrile molecule in the same site, the data presented might be slightly shifted compared with that of the free isolated CH3CNS
molecule.
DCC and NS stretching modes are strongly mixed.
ESite splitting or splitting due to different complex formations.
FVery weak bands, tentative assignment.

RESEARCH FRONT
Structure, Stability, and Generation of CH3CNS
1691
band is relatively more intense at the beginning of the photo-
lysis, and its relative intensity decreases compared with the
assigned bands of CH3CNS (or more precisely CH3CNꢂ
CH3CNS) on prolonged 254-nm UV irradiation. The assign-
ment of this band is not clear, but it may belong to a higher-
energy CH3CNꢂCH3CNS complex.
The observation of blue light also confirms that CH3CNS
decomposes to CH3CN and an S atom on broad-band UV
photolysis. We also note that our similar unpublished experi-
ments with the parent 1,2,5-thiadiazole and its fluoro derivative,
namely the generation and photodecomposition of HCNS and
FCNS, produced the same blue chemiluminescence.
We note that there is a group of bands at 2210, 2165, 2159,
2152, 2145, 2143, and 2136 cmꢁ1 (see Fig. S1, Accessory
Publication), whose assignment is not clear. These bands have
very weak relative intensity after the 254-nm photolysis, but
their intensity increases after bleaching CH3CNS with broad-
band UV irradiation (see below). These bands may belong to
reaction products formed owing to the presence of moisture in
the matrix.
Conclusion
According to our computational results, CH3CNS is a sym-
metric-top molecule with a linear CCNS frame, and it is unstable
owing to bimolecular reactions. Six possible bimolecular
pathways with almost identical activation energies of DG8
0 K
50–57 kJ molꢁ1 were identified for the initiating step of
decomposition, leading to various products.
¼
CH3CNSin the matrix cannot be photodecomposed by 254-nm
UV light. Broad-band UV irradiation, however, destroyed the
molecule, and absorption bands of CH3CN increased, which
suggests the formation of a sulfur atom. All fundamental
vibrational bands of the CH3CN dimer could be detected at
3003, 2946, 2256, 1448, 1375, 1041, and 916 cmꢁ1. According
to our calculations (Fig. 1), the excited-state singlet sulfur atom
reacts in a barrierless process with acetonitrile, but the ground-
state triplet sulfur atom does not react at low temperature. The
triplet sulfur atom may form either by singlet–triplet transition
of the singlet sulfur atom, following singlet sulfur extrusion, or
by N–S bond rupture of an excited triplet CH3CNS, formed from
excited singlet species after intersystem crossing. The ground-
state triplet sulfur atom is thus the final product of the photolysis,
and an experimental proof of its formation is provided below.
It was very interesting to observe a very bright, blue chemi-
luminescence after warming the matrix after broad-band UV
irradiation. Light emission started above ,20 K, with increasing
temperature increasing the intensity of the emitted light. The
spectrum of this light recorded at ,20–25 K is shown in Fig. 7.
After a perusal of the relevant literature, we found that this
spectrum is identical with the chemiluminescence spectrum of
triplet sulfur atoms in an Ar matrix observed by Wurfel and
Pimentel[22] after warming the matrix of the photolysis products
CH3CNS was generated experimentally in an Ar matrix from
3,4-dimethyl-1,2,5-thiadiazole by photoelimination of aceto-
nitrile, and its IR and UV spectra have been recorded for the
first time. Considering that CH3CNS and CH3CN are generated
in the same cage of the matrix, the formation of a weak van der
Waals complex between these species, CH3CNꢂCH3CNS, is
very likely. Similar complex formation has been detected earlier
for HCNS[5] and FCNS[6] after generating them from the
corresponding thiadiazole derivatives. It is thus expected that
the presence of CH3CN slightly shifts the spectral bands of
CH3CNS compared with the free isolated CH3CNS.
The observed CꢀN stretching frequency of CH3CNS,
2240 cmꢁ1, is higher than those of HCNS, 2035 cmꢁ1 [5]
and
,
FCNS, 1978 cmꢁ1 [6]
generated under similar conditions. The
,
opposite trend is observed for the N–S stretching, namely 565,
707, and 737 cmꢁ1 for CH3CNS, HCNS, and FCNS, respec-
tively. This difference may be explained by different coupling
of the normal modes and/or different bond strength; however,
the interpretation is difficult owing to the complex formation
between nitrile sulfides and nitriles. We note that our calcula-
tions indicate that C–C and N–S stretching modes are strongly
mixed in CH3CNS.
CH3CNS has a characteristic absorption band at 209 nm in
the near-UV region. This molecule decomposes to acetonitrile
and a triplet sulfur atom on broad-band UV irradiation.
00 3
ꢁ
~
~
of OCS. The blue emission was assigned to the B Pu ! XSg
emission of S2.[22] This indicates that triplet sulfur atoms react
with each other, and not with any other species, and S2 species
form in the electronically excited state on warming the matrix.
Experimental and Computational Methods
The 3,4-dimethyl-1,2,5-thiadiazole precursor was synthesized
from dimethylglyoxime and disulfur dichloride according to a
literature method.[23] Starting materials were commercial pro-
ducts (Aldrich). The purity of the precursor was checked by
NMR and IR spectroscopy.
For matrix isolation experiments, the vapour of the degassed
sample was mixed with argon (Messer, purity 99.9997%) in a
1:1000 mol ratio in a glass vacuum line. The gas mixture was
deposited onto an 8-K CsI window for IR, and onto a 12-K
quartz window for UV-visible spectroscopic measurements.
These windows were mounted on a Janis CCS-350R cold head
cooled by a CTI Cryogenics 22 closed-cycle refrigerator unit
coupled to a CTI 8200 compressor. The flow rate was regulated
with an MKS flow controller system. The flow rate was ,1.2
and 0.3 cm3 minꢁ1 for IR and UV measurements, respectively.
The temperature of the copper cold head was set with a
Lake Shore 321 thermostat equipped with a silicon diode
thermometer.
300
350
400
450
500
550
600
Wavelength [nm]
In the photolysis experiments, the matrix was irradiated with
a Cathodeon HPK 125-W high-pressure mercury lamp (the most
Fig. 7. The emission spectrum observed on warming up the matrix to
,20–25 K after broad-band UV photolysis.

RESEARCH FRONT
1692
M. Krebsz et al.
intense lines are at ,580, 578, 548, 440, 405, 365, 313, and
254 nm) through a quartz window mounted on the vacuum
chamber. Melles Griot interference filters (full width at the
half-height of 10 nm) were used to select a single line of the
mercury lamp.
References
[1] R. M. Paton, Chem. Soc. Rev. 1989, 18, 33. doi:10.1039/
CS9891800033
[2] C. Wentrup, P. Kambouris, Chem. Rev. 1991, 91, 363. doi:10.1021/
CR00003A004
[3] S. Kanemasa, Sci. Synthesis 2004, 19, 17.
IR spectra were recorded by a Bruker IFS 55 Fourier trans-
form infrared (FT-IR) spectrometer equipped with a KBr
beamsplitter and a deuterated triglycine sulphate detector. A
total of 250–1000 scans were accumulated at 1 cmꢁ1 resolution
in the 400–4000 cmꢁ1 spectral window. The Happ-Genzel
apodization function, Mertz phase correction using a phase
resolution of 32 cmꢁ1, and a zero filling factor of 4 were applied.
Absorption UV-visible spectra were recorded with a Varian
Cary3E spectrometer, using a 5-nm minꢁ1 scan rate, 0.333-nm
step size, and 1-nm spectral band width. The data were collected
in the 190–400 nm spectral region.
[4] M. Krebsz, T. Pasinszki, Curr. Org. Chem. 2010, in press.
[5] T. Pasinszki, M. Krebsz, G. Bazso´, G. Tarczay, Chemistry 2009, 15,
6100. doi:10.1002/CHEM.200900408
[6] T. Pasinszki, G. Bazso´, M. Krebsz, G. Tarczay, Phys. Chem. Chem.
Phys. 2009, 11, 9458. doi:10.1039/B913204J
[7] P. Kambouris, M. Plisnier, R. Flammang, J. K. Terlouw, C. Wentrup,
Tetrahedron Lett. 1991, 32, 1487. doi:10.1016/0040-4039(91)
80365-D
[8] N. Harrit, A. Holm, I. R. Dunkin, M. Poliakoff, J. J. Turner, J. Chem.
Soc, Perkin Trans. 2 1987, II, 1227. doi:10.1039/P29870001227
[9] R. Flammang, P. Gerbaux, E. H. Morkved, M. W. Wong, C. Wentrup,
J. Phys. Chem. 1996, 100, 17452. doi:10.1021/JP9618800
[10] P. Gerbaux, Y. V. Haverbeke, R. Flammang, M. W. Wong,
C. Wentrup, J. Phys. Chem. A 1997, 101, 6970. doi:10.1021/
JP971170þ
Emission UV-visible spectra were recorded with an Ocean
Optics HR2000 fibre optic spectrometer. The spectrum accu-
mulation time was 3 s.
Quantum chemical calculations were performed with
the ACES2,[24,25] Gaussian03,[26] and MOLPRO[27] program
packages. Equilibrium geometries, transition structures,
homolytic dissociation energies, harmonic frequencies, and
anharmonic vibrational frequencies within the framework of
second-order vibrational perturbational theory[28,29] were cal-
culated by coupled cluster and density functional methods.
Stability checks of wave functions were performed for all
calculated structures. The connecting lowest-energy paths
between minima were obtained using an intrinsic reaction
coordinate (IRC)[30,31] approach. Owing to the complexity of
the surface, the IRC was also manually checked, by proceeding
along the given reaction coordinate and simultaneously relaxing
all other bond lengths and angles. Unimolecular properties
were calculated at the B3LYP[32,33] and CCSD(T)[34] levels.
Bimolecular reactions were calculated at the MR-AQCC(2,2)//
(U)B3LYP level; geometry optimizations were performed at the
UB3LYP and RB3LYP levels for open-shell singlet and closed-
shell species, respectively, and single-point energy calculations
were done using (U)B3LYP geometries at the MR-AQCC
(2,2)[35] level to obtain the total energies. Total energies of
CH3CNS and CH3CN were calculated at the SR-AQCC//
B3LYP level. Zero-point vibrational energies and thermal
corrections were calculated at the (U)B3LYP level. Three
common correlation consistent one-electron basis sets, cc-
pVTZ, aug-cc-pVTZ, and cc-pCVTZ, were used for all mole-
cules, the latter two only for calculating the geometry and
vibrational frequencies of CH3CNS. Unless noted otherwise,
the results obtained using the cc-pVTZ basis set are presented
in the Discussion section. Test calculations were performed
using the (U)B3LYP/cc-pVTZ geometries of selected test
molecules. Core electrons were kept frozen (fc) in post-HF
calculations, except for the cc-pCVTZ basis set.
[11] Z. Fu, X. Pan, Z. Li, C. Sun, R. Wang, Chem. Phys. Lett. 2006, 430, 13.
doi:10.1016/J.CPLETT.2006.07.103
[12] T. Pasinszki, T. Ka´rpa´ti, N. P. C. Westwood, J. Phys. Chem. A 2001,
105, 6258. doi:10.1021/JP010830E
[13] T. J. Lee, P. R. Taylor, Int. J. Quant. Chem. Symp. 1989, 23, 199.
[14] Zh.-X. Yu, P. Caramella, K. N. Houk, J. Am. Chem. Soc. 2003, 125,
15420. doi:10.1021/JA037325A
[15] E. Goldstein, B. Beno, K. N. Houk, J. Am. Chem. Soc. 1996, 118, 6036.
doi:10.1021/JA9601494
[16] J. Gra¨fenstein, A. M. Hjerpe, E. Kraka, D. Cremer, J. Phys. Chem. A
2000, 104, 1748. doi:10.1021/JP993122Q
[17] G. Orlova, J. D. Goddard, J. Chem. Phys. 2000, 112, 10085.
doi:10.1063/1.481650
[18] T. Pasinszki, B. Hajgato´, B. Havasi, N. P. C. Westwood, Phys. Chem.
Chem. Phys. 2009, 11, 5263. doi:10.1039/B823406J
[19] H. S. Kim, K. Kim, Bull. Korean Chem. Soc. 1992, 13, 520.
[20] G. Herzberg, Molecular Spectra and Molecular Structure III. Electro-
nic Spectra and Electronic Structure of Polyatomic Molecules 1966
(van Nostrand Reinhold: New York, NY).
[21] NIST Atomic Spectra Database Lines Form. Available online at:
http://physics.nist.gov/cgi-bin/Elements/ASD_lines.pl?el¼S [verified
October 2010].
[22] B. E. Wurfel, G. C. Pimentel, Chem. Phys. Lett. 1994, 223, 301.
doi:10.1016/0009-2614(94)00461-7
[23] L. M. Weinstock, P. Davies, B. Handelsman, R. Tull, J. Org. Chem.
1967, 32, 2823. doi:10.1021/JO01284A040
[24] ACES II: Mainz-Austin-Budapest version: J. F. Stanton, J. Gauss, J. D.
Watts, P. G. Szalay, R. J. Bartlett with contributions from A. A. Auer,
D. E. Bernholdt, O. Christiansen, M. E. Harding, M. Heckert, O. Heun,
C. Huber, D. Jonsson, J. Juse´lius, W. J. Lauderdale, T. Metzroth,
C. Michauk, D. R. Price, K. Ruud, F. Schiffmann, A. Tajti, M. E.
Varner, J. Va´zquez and the integral packages: MOLECULE (J. Almlo¨f
and P. R. Taylor), PROPS (P. R. Taylor), and ABACUS (T. Helgaker,
H. J. Aa. Jensen, P. Jørgensen, and J. Olsen). See also J. F. Stanton,
J. Gauss, J. D. Watts, W. J. Lauderdale, R. J. Bartlett, Int. J. Quantum
Chem. Symp. 1992, 26, 879. Current version see: http://www.aces2.de
[verified October 2010].
Accessory Publication
[25] J. F. Stanton, J. Gauss, J. D. Watts, W. J. Lauderdale, R. J. Bartlett, Int.
J. Quantum Chem., Quantum Chem. Symp. 1992, 26, 897.
[26] M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria, M. A. Robb,
J. R. Cheeseman, J. A. Montgomery, Jr., T. Vreven, K. N. Kudin,
J. C. Burant, J. M. Millam, S. S. Iyengar, J. Tomasi, V. Barone,
B. Mennucci, M. Cossi, G. Scalmani, N. Rega, G. A. Petersson,
H. Nakatsuji, M. Hada, M. Ehara, K. Toyota, R. Fukuda, J. Hasegawa,
M. Ishida, T. Nakajima, Y. Honda, O. Kitao, H. Nakai, M. Klene, X. Li,
J. E. Knox, H. P. Hratchian, J. B. Cross, C. Adamo, J. Jaramillo,
R. Gomperts, R. E. Stratmann, O. Yazyev, A. J. Austin, R. Cammi,
The calculated equilibrium structure of CH3CNS, energetics
of the bimolecular reaction routes of CH3CNS, and enlarged
IR spectrum of the photolysis products of 3,4-dimethyl-1,2,5-
thiadiazole are available from the Journal’s website.
Acknowledgements
This work was funded by the Hungarian Scientific Research Fund (grant no.
OTKA K75877).

RESEARCH FRONT
Structure, Stability, and Generation of CH3CNS
1693
C. Pomelli, J. W. Ochterski, P. Y. Ayala, K. Morokuma, G. A. Voth,
P. Salvador, J. J. Dannenberg, V. G. Zakrzewski, S. Dapprich, A. D.
Daniels, M. C. Strain, O. Farkas, D. K. Malick, A. D. Rabuck,
K. Raghavachari, J. B. Foresman, J. V. Ortiz, Q. Cui, A. G. Baboul,
S. Clifford, J. Cioslowski, B. B. Stefanov, G. Liu, A. Liashenko,
P. Piskorz, I. Komaromi, R. L. Martin, D. J. Fox, T. Keith, M. A.
Al-Laham, C. Y. Peng, A. Nanayakkara, M. Challacombe, P. M. W.
Gill, B. Johnson, W. Chen, M. W. Wong, C. Gonzalez, J. A. Pople,
Gaussian 03, Revision C.02 2004 (Gaussian, Inc.: Wallingford, CT).
[27] MOLPRO 2010.1: MOLPRO is a package of ab initio programs
written by H.-J. Werner, P. J. Knowles, F. R. Manby, M. Schu¨tz,
P. Celani, G. Knizia, T. Korona, R. Lindh, A. Mitrushenkov,
G. Rauhut, T. B. Adler, R. D. Amos, A. Bernhardsson, A. Berning,
D. L. Cooper, M. J. O. Deegan, A. J. Dobbyn, F. Eckert, E. Goll,
C. Hampel, A. Hesselmann, G. Hetzer, T. Hrenar, G. Jansen, C. Ko¨ppl,
Y. Liu, A. W. Lloyd, R. A. Mata, A. J. May, S. J. McNicholas,
W. Meyer, M. E. Mura, A. Nicklaß, P. Palmieri, K. Pflu¨ger, R. Pitzer,
M. Reiher, T. Shiozaki, H. Stoll, A. J. Stone, R. Tarroni,
T. Thorsteinsson, M. Wang, A. Wolf, see: http://www.molpro.net
[verified October 2010]; Integral evaluation (SEWARD) R. Lindh,
U. Ryu, B. Liu, J. Chem. Phys. 1991, 95, 5889; Coupled-Cluster
treatments: C. Hampel, K. Peterson, H.-J. Werner, Chem. Phys. Lett.
1992, 190, 1 and references therein. The program to compute the
perturbative triples corrections has been developed by M. J. O.
Deegan, P. J. Knowles, Chem. Phys. Lett. 1994, 227, 321.
[28] H. H. Nielsen, Rev. Mod. Phys. 1951, 23, 90. doi:10.1103/
REVMODPHYS.23.90
[29] W. D. Allen, Y. Yamaguchi, A. G. Csa´sza´r, D. A. Clabo, Jr, R. B.
Remington, H. F. Schaefer, III, Chem. Phys. 1990, 145, 427.
doi:10.1016/0301-0104(90)87051-C
[30] C. Gonzalez, H. B. Schlegel, J. Chem. Phys. 1989, 90, 2154.
doi:10.1063/1.456010
[31] C. Gonzalez, H. B. Schlegel, J. Phys. Chem. 1990, 94, 5523.
doi:10.1021/J100377A021
[32] A. D. Becke, J. Chem. Phys. 1993, 98, 5648. doi:10.1063/1.464913
[33] C. Lee, W. Yang, R. G. Parr, Phys. Rev. B 1988, 37, 785. doi:10.1103/
PHYSREVB.37.785
[34] K. Raghavachari, G. W. Trucks, J. A. Pople, M. Head-Gordon, Chem.
Phys. Lett. 1989, 157, 479. doi:10.1016/S0009-2614(89)87395-6
[35] P. G. Szalay, R. J. Bartlett, Chem. Phys. Lett. 1993, 214, 481.
doi:10.1016/0009-2614(93)85670-J