The decomposition of benzenesulfonyl azide: A matrix isolation and computational study

Article abstract of DOI:10.1039/c6cp08125h

The thermal-decomposition and photo-decomposition of benzenesulfonyl azide, PhS(O)₂N₃, have been studied by combining matrix-isolation IR spectroscopy and quantum chemical calculations. Upon flash vacuum pyrolysis at 800 K, the azide splits off molecular nitrogen and exclusively furnishes phenylnitrene (PhN) and SO₂ in the gas phase. In contrast, the azide favors stepwise photodecomposition in solid Ar and Ne matrices at 2.8 K. Specifically, the UV laser photolysis (193 and 266 nm) of PhS(O)₂N₃ results in the formation of the key nitrene intermediate PhS(O)₂N in the triplet ground state, which undergoes pseudo-Curtius rearrangement into N-sulfonyl imine PhNSO₂ under subsequent visible light irradiation (380-450 nm). Further fragmentation of PhNSO₂ into SO₂ and PhN followed by ring-expansion to didehydroazepine also occurs upon visible light irradiation. The preference of the stepwise mechanism for the decomposition of PhS(O)₂N₃ is supported by quantum chemical calculations using DFT B3LYP/6-311++G(3df,3pd) and CBS-QB3 methods.

Full text of DOI:10.1039/c6cp08125h

PCCP
View Article Online
View Journal
PAPER
The decomposition of benzenesulfonyl azide:
a matrix isolation and computational study†
Cite this:DOI: 10.1039/c6cp08125h
Guohai Deng,a Xuelin Dong,a Qifan Liu,a Dingqing Li,a Hongmin Li,a Qiao Sunb and
Xiaoqing Zeng*a
The thermal-decomposition and photo-decomposition of benzenesulfonyl azide, PhS(O)2N3, have been
studied by combining matrix-isolation IR spectroscopy and quantum chemical calculations. Upon flash
vacuum pyrolysis at 800 K, the azide splits off molecular nitrogen and exclusively furnishes phenylnitrene
(PhN) and SO2 in the gas phase. In contrast, the azide favors stepwise photodecomposition in solid Ar and
Ne matrices at 2.8 K. Specifically, the UV laser photolysis (193 and 266 nm) of PhS(O)2N3 results in the
formation of the key nitrene intermediate PhS(O)2N in the triplet ground state, which undergoes
pseudo-Curtius rearrangement into N-sulfonyl imine PhNSO2 under subsequent visible light irradiation
(380–450 nm). Further fragmentation of PhNSO2 into SO2 and PhN followed by ring-expansion to
didehydroazepine also occurs upon visible light irradiation. The preference of the stepwise mechanism
for the decomposition of PhS(O)2N3 is supported by quantum chemical calculations using DFT
B3LYP/6-311++G(3df,3pd) and CBS-QB3 methods.
Received 28th November 2016,
Accepted 11th January 2017
DOI: 10.1039/c6cp08125h
www.rsc.org/pccp
7
and PhC(O)N3 ) keeps attracting attention in the field of physical
organic chemistry. By analogy, a similar rearrangement reaction,
Introduction
Covalent azides R–N3 are versatile reagents that have been known as the pseudo-Curtius rearrangement, has been extended
broadly used such as in cycloaddition reactions in chemistry to other structurally related a-oxo azides, including phosphoryl
and biology,1a synthesis of high-energy density polynitrogen azide R2P(O)N3,8 sulfinyl azide RS(O)N3,9 and sulfonyl azide
materials,1b and surface functionalization of nanomaterials.1c RS(O)2N3.10 Sulfonyl azides are useful starting materials for the
Generally, covalent azides are energetic and sensitive to heat and synthesis of biologically active enaminones, amidines, and
irradiation due to the facile elimination of molecular nitrogen amidates.11 Similar to the decomposition of carbonyl azides,
and simultaneous formation of a highly reactive nitrene inter- sulfonyl azides RS(O)2N3 may undergo pseudo-Curtius rearrange-
mediate R–N, which may undergo further inter- or intra- ment into N-sulfonyl imine RNSO2 via the sulfonyl nitrene
molecular reactions under the decomposition conditions. As an intermediate RS(O)2N. However, the early solution chemistry12
alternative route, concerted decomposition of azide into N2 and and more recent time-resolved spectroscopy studies10a,b upon
the rearrangement product may occur without the intermediacy both the thermal-decomposition and photo-decomposition of
of nitrene. The mechanism for the decomposition of covalent sulfonyl azides failed to obtain convincing evidence for the
azides and also the properties of the involved nitrene inter- formation of the final rearrangement product RNSO2, although
mediates have been extensively studied from both experimental a number of sulfonyl nitrene intermediates had been either
and theoretical points of view.2
directly detected or chemically trapped.
As one of the best known reactions in organic chemistry, the
The decomposition of arylsulfonyl azides ArS(O)2N3 in solution
Curtius-rearrangement refers to the decomposition of carbonyl was found to be rather complex. According to the previous solution
azide RC(O)N3 into isocyanate RNCO with concomitant N2 studies,13 thermolysis and photolysis of arylsulfonyl azides in
elimination.3 The mechanism for the decomposition of various solvents produce traces of SO2 and derivatives of the nitrene
carbonyl azides (e.g., HC(O)N3,4 CH3C(O)N3,5 CH3OC(O)N3,6 ArS(O)2N. Later on, the photochemistry of 4-methylbenzene-
sulfonyl azide,14 2-naphthylsulfonyl azide,10b,15 and p-bromo-
a College of Chemistry, Chemical Engineering and Materials Science,
Soochow University, 215123 Suzhou, P. R. China. E-mail: xqzeng@suda.edu.cn
b School for Radiological and Interdisciplinary Sciences, Soochow University,
215123 Suzhou, P. R. China
phenylsulfonyl and p-toylsulfonyl azides10a was explored by joint
laser flash photolysis and ultrafast spectroscopy. The corres-
ponding arylsulfonyl nitrenes have been directly detected; how-
ever, the kinetics results demonstrate that the rearrangement
proceeds in a concerted pathway via the ‘‘rearrangement in the
excited state (RIES)’’ mechanism.
† Electronic supplementary information (ESI) available: Calculated IR spectro-
scopic data, energies, and coordinates for all species discussed in the paper. See
DOI: 10.1039/c6cp08125h
This journal is ©the Owner Societies 2017
Phys. Chem. Chem. Phys.

View Article Online
Paper
PCCP
The photochemistry (265 nm, mercury lamp) of the parent Rh-plated copper block matrix support (2.8 K) under high
arylsulfonyl azide, PhS(O)2N3, has been studied by Sheridan vacuum (B10ꢀ6 Pa), using a closed-cycle helium cryostat
and Rempala in a solid N2 matrix at 16 K,16 and the pseudo- (Sumitomo Heavy Industries, SRDK-408D2-F50H) inside the
Curtius rearrangement product PhNSO2 with two IR bands at vacuum chamber. For the IR spectral measurement, the cryo-
1371 and 1323 cmꢀ1 was tentatively identified due to spectral stat was horizontally rotated (1801) to allow the matrix-isolated
congestion.
sample under the irradiation of the IR beam. The electric power
Very recently, the decomposition reactions of the simple (voltage/current) used in pyrolysis experiments was 3.5 V/2.86 A.
sulfonyl azides RS(O)2N3 (R = F,10c CH3,17 CF3,18 and CH3O10d
)
Photolysis was performed using an ArF excimer laser (Gamlaser
have been investigated by combining matrix isolation and EX5/250, 3 Hz, 193 nm), Nd3+:YAG laser (266 nm, MP-L-266, 10 mW),
quantum chemical calculations. All the corresponding sulfonyl and a high-power flashlight (Boyu T648, 20 W).
nitrenes, generated from the photolysis of the azides in cryogenic
Computational details
matrices, have been directly observed in the triplet ground state,
and the respective pseudo-Curtius rearrangement products were Geometry optimization and harmonic IR frequency (unscaled)
always accompanied only as minor photolysis products. However, calculations were performed using DFT methods (B3LYP,20
owing to the very different reactivity of the initially generated M06-2X,21 MPW1PW9122) combined with the 6-311++G(3df,3pd)
sulfonyl nitrenes, diverse secondary reactions including rearrange- basis set.23 Accurate energies of the species were further calculated
ment, dimerization, and fragmentation have been observed, by using the complete basis set (CBS-QB3).24 Time-dependent
through which novel sulfur-containing species FSO2,10c NSO2,18a (TD)25 B3LYP/6-311++G(3df,3pd) calculations were carried out
10d
and HNSO2 have been produced in the gas phase.
for the prediction of UV-vis transitions. Local minima were
Herein, the thermal-decomposition and photo-decomposition confirmed by vibrational frequency analysis, and transition
of the simplest arylsulfonyl azide PhS(O)2N3 have been studied states were further confirmed by intrinsic reaction coordinate
by combining matrix isolation IR spectroscopy and quantum (IRC) calculations.26 All calculations were performed using the
chemical calculations. A stepwise decomposition pathway via Gaussian 09 suite of programs.27
the intermediacy of PhS(O)2N and PhNSO2 into PhN, SO2, and
N2 was unambiguously established from both experimental and
theoretical points of view.
Results and discussion
Characterization of PhS(O)2N3
Despite the fact that the first synthesis of PhS(O)2N3 was
reported in 1955,28 it was only partially characterized by 1H
and 13C NMR and IR spectroscopy.19 We obtained the matrix-
isolation (Ne and Ar) IR and Raman spectra of the azide (Fig. 1
with molecular structure), and the vibrational data (Table 1)
were analyzed with the aid of quantum chemical calculations
by using the B3LYP and M06-2X methods.
Experimental section
Caution! Covalent azides are hazardous and explosive! Although
no explosions occurred during this work, safety precautions
(face shields, leather gloves, and protective leather clothing) are
strongly recommended, particularly in the case of handling pure
PhS(O)2N3 in the condensed phase.
Generally, the experimentally observed IR spectra (Fig. 1B) of
PhS(O)2N3 are in agreement with the DFT calculations (Fig. 1A).
Sample preparation
Benzenesulfonyl azide PhS(O)2N3 was prepared by the reaction of
benzenesulfonyl chloride PhS(O)2Cl with sodium azide (NaN3) in
acetone according to the reported procedure.19 The quality of the
sample was checked by 1H, 13C NMR spectroscopy (Bruker 400 MHz)
and Raman spectroscopy (Horiba JY HR800).
Matrix IR spectroscopy
Matrix IR spectra were recorded on a FT-IR spectrometer
(Bruker 70 V) in a reflectance mode using a transfer optic.
A KBr beam splitter and a liquid-nitrogen-cooled MCT detector
were used in the mid-IR region (4000–600 cmꢀ1). For each
spectrum, 200 scans at a resolution of 0.5 cmꢀ1 were co-added.
The gaseous sample was mixed by passing a flow of matrix
gas (Ne and Ar) through a cold U-trap (3–6 1C) containing
ca. 20 mg of the azide. The azide/noble gas mixture (1 : 1000
estimated) was passed through Al2O3 (o.d. 2.0 mm, i.d. 1.0 mm),
which could be heated over a length of approximately 25 mm by
a tantalum wire (o.d. 0.4 mm, resistance 0.4 O). Then, the
Fig. 1 (A) Calculated IR spectrum of PhS(O)2N3 at the B3LYP/
6-311++G(3df,3pd) level; (B) the observed IR spectrum of PhS(O)2N3 in
the solid Ne matrix at 2.8 K (absorbance A, resolution 0.5 cmꢀ1); (C) Raman
mixture was immediately deposited (2 mmol hꢀ1) onto the spectrum of liquid PhS(O)2N3 at 298 K (Raman intensity I, resolution 2 cmꢀ1).
Phys. Chem. Chem. Phys.
This journal is ©the Owner Societies 2017

View Article Online
PCCP
Paper
Table 1 Observed and calculated IR frequencies (cmꢀ1) for PhS(O)2N3
a
Observedb
Calculatedc
B3LYP
IR (Ne matrix)
3281.9 vw
IR (Ar matrix)
3280.5 vw
IR (liquid)e
3276 w
Raman (liquid)
M06-2X
Approximate assignmentd
3215 (5)
3209 (6)
n(CH)
n(CH)
n(CH)
n(CH)
n(CH)
3166 w
3075 m
3214 (o1)
3205 (o1)
3198 (o1)
3192 (4)
3181 (o1)
2348 (515)
1661 (2)
3084.9 vw
3080.6 vw
3069 w
3199 (6)
3190 (7)
3178 (o1)
2251 (554)
1622 (2)
2131.7 vs
1591.8 w
2129.0 vs
1589.7 w
2129 s
1584 w
2131 m
1588 m
nas(N3)
n(CC)
r(CH)
r(CH)
r(CH)
1621 (o1)
1658 (o1)
1479.2 w
1454.9 w
1405.2 s
1341.9 vw
1314.4 vw
1298.6 w
1194.5 s
1182.3 s
1478.0 w
1452.7 w
1397.2 s
1341.2 vw
1312.5 vw
1297.8 w
1194.1 s
1180.5 s
1477 w
1450 m
1372 s
1511 (4)
1519 (3)
1449 vw
1371 w
1479 (21)
1404 (152)
1351 (12)
1330 (2)
1269 (244)
1204 (4)
1189 (143)
1188 (o1)
1104 (5)
1094 (31)
1041 (1)
1024 (o1)
1006 (o1)
1000 (2)
958 (o1)
865 (o1)
773 (15)
742 (122)
717 (11)
695 (32)
1488 (29)
1450 (164)
1342 (9)
1326 (o1)
1279 (267)
1231 (268)
1200 (7)
1177 (o1)
1124 (36)
1110 (5)
n
as(SO2)
r(CH)
1313 w
1298 w
r(CH)
ns(N3)
r(CH)
ns(SO2)
r(CH)
r(CH)
n(SC) + r(CH)
Ring distortion
Ring distortion
o(CH)
Ring breathing
o(CH)
1185 w, sh
1162 s
1170 s
1118.3 vw
1093.4 w
1025.5 vw
1119.4 vw
1091.9 w
1024.6 vw
1088 m
1023 w
1087 w
1023 m
1056 (1)
1038 (o1)
1016 (o1)
1005 (1)
967 (o1)
874 (o1)
803 (121)
778 (25)
735 (44)
703 (39)
635 (183)
624 (36)
623 (7)
1002.4 vw
1000.9 vw
1000 w
998 vs
o(CH)
765.2 vw
753.4 s
716.9 vw
688.1 w
764.0 vw
751.4 s
715.1 vw
686.5 w
761 w
713 m
612 m
563 w
n(SN)
751 m
685 m
n(SC)
d(CH)
d(CH)
d(CH)
n(SN) + d(SO2)
do.o.p(N3)
627 (o1)
606 (259)
586 (8)
607.9 s
603.9 s
601 s
564 s
566 (86)
581 (93)
n(SN) + d(SO2)
a
b
A full list of the calculated harmonic IR frequencies (unscaled) is given in Table S1 in the ESI. Band intensities: vs, very strong; s, strong; m,
medium strong; w, weak; vw, very weak; sh, shoulder. Calculated IR frequencies and intensities (km molꢀ1) in parentheses with the
6-311++G(3df,3pd) basis set. Tentative assignment based on the calculated vibrational displacement vectors. Taken from ref. 19.
c
d
e
The band positions of PhS(O)2N3 observed in cryogenic matrices are weak band for the C–H rocking mode (r(CH)) occurs as a strong
very close to those in the liquid, except for the two SO2 stretching IR band at 1194.5 cmꢀ1. The discrepancies are probably caused
vibrational modes. In a solid Ne matrix, the asymmetric and by anharmonic vibrational coupling between these two modes.
symmetric stretching modes appear at 1405.2 and 1182.3 cmꢀ1
,
The C–S stretching vibration appears at 753.4 cmꢀ1, which is
whereas in the liquid these two shift to 1372 and 1170 cmꢀ1 in close to those in CH3S(O)2N3 (731.7 cmꢀ1 17
)
and CF3S(O)2N3
the IR and 1371 and 1162 cmꢀ1 in the Raman spectra, respec- (756.6 cmꢀ1).29
tively. Very similar changes have also been observed for the
Flash vacuum pyrolysis of PhS(O)2N3
same vibrational modes in CH3S(O)2N3 (Ar matrix IR: nasym(SO2)
1391.5 cmꢀ1, nsym(SO2) 1180.0 cmꢀ1; solid Raman: nasym(SO2)
1357 cmꢀ1, nsym(SO2) 1151 cmꢀ1).17 The red-shifts indicate
intermolecular hydrogen bonding interactions with the SO2
moiety in the condensed phase as have been observed in
CH3S(O)2N3 (H2C–Hꢁ ꢁ ꢁOSO) by X-ray crystallography.17
The pyrolysis of PhS(O)2N3 (Scheme 1) was carried out by passing the
azide/noble gas mixture (1: 1000) through a heated Al2O3 furnace
(ca. 800 K). The resulting mixture was immediately deposited
onto the matrix support (2.8 K) for the IR spectroscopy study.
The most characteristic asymmetric stretching vibration
mode of the N3 group in PhS(O)2N3 locates at 2131.7 cmꢀ1
(Ne matrix). It is slightly red-shifted in comparison to those of
other sulfonyl azides CH3S(O)2N3 (2138.6 cmꢀ1),17 FS(O)2N3
(2162.3 cmꢀ1) and CF3S(O)2N3 (2151.6 cmꢀ1).29 According to
the calculated vibrational displacement vectors, the symmetric
N3 stretching mode should appear as a strong IR band due to
the large IR intensity (244 km molꢀ1, B3LYP/6-311++G(3df,3pd)).
However, only weak bands at 1298.6 (Ne matrix) and 1298 cmꢀ1
Scheme 1 Thermal decomposition of PhS(O)2N3 and photochemistry
(liquid) are found in the IR spectra. In contrast, the predicted of PhN.
This journal is ©the Owner Societies 2017
Phys. Chem. Chem. Phys.

View Article Online
Paper
PCCP
in the IR spectrum of the FVP products of PhS(O)2N3 renders this
route unlikely.
As an alternative pathway, the decomposition of PhS(O)2N3
may proceed by spitting off N2 and forming the nitrene inter-
mediate PhS(O)2N, which can either isomerize into PhNSO2
followed by N–S bond cleavage into PhN and SO2, or directly
dissociate into PhN and SO2 concertedly. This assumption is
partially supported by the observation of the trapping products
of arylsulfonyl nitrenes during the thermolysis of arylsulfonyl
azides in aprotic solvents.13a However, the nitrene PhS(O)2N
was not experimentally observed among the pyrolysis products.
Thus, it is very likely that the initially generated nitrene PhS(O)2N
in the gas phase is unstable against fragmentation under the
pyrolysis conditions.
Photodecomposition of PhS(O)2N3
To capture the decomposition intermediate, the photochemistry
of PhS(O)2N3 (Scheme 2) in noble gas matrices (Ar, Ne) at 2.8 K
was studied by utilizing an ArF excimer laser (193 nm), which
was proved to be powerful in producing various missing a-oxo
nitrenes in cryogenic matrices through photolysis of the azide
precursors.33
Fig. 2 (A) IR spectrum of PhS(O)2N3 in the solid Ne matrix at 2.8 K. (B) IR
spectrum of the flash vacuum pyrolysis of PhS(O)2N3 in the solid Ne matrix
at 2.8 K; (C) IR difference spectrum showing the change of the matrix
containing flash vacuum pyrolysis products of PhS(O)2N3 upon visible light
(l = 380–450 nm, 20 W, 5 min) irradiation, bands for the depleted species
point downwards and bands for the formed species point upwards. The IR
bands of PhS(O)2N3 (a), PhN (d), SO2 (e), cyclic C6H5N (f), CO2 (g), and H2O
(h) are labeled.
The IR difference spectrum reflecting the change of the
matrix-isolated azide upon irradiation is shown in Fig. 3B.
Unlike the aforementioned FVP of PhS(O)2N3, neither PhN nor
SO2 but a new species with one set of IR bands at 1588.7, 1452.5,
1348.9, 1168.2, 1090.8, 1000.2, 766.7, 736.7, and 683.7 cmꢀ1 was
generated upon irradiation. To aid the assignment, theoretical
calculations on the IR frequencies for the most likely candidate
species, benzenesulfonyl nitrene (PhS(O)2N), in both singlet and
triplet states were performed, and the results are summarized in
Table 2. Clearly, the calculated spectrum for the lower-energy
triplet species (Fig. 3D) shows better agreement with the experi-
mental data. The two most intense bands at 1348.9 and 1168.2 cmꢀ1
belong to the asymmetric and symmetric SO2 stretching modes,
respectively, which are close to those in PhS(O)2N3 (1405.2 and
1182.3 cmꢀ1), CF3S(O)2N (1389.8 and 1179.8 cmꢀ1),18b and
CH3S(O)2N (1349.8 and 1156.6 cmꢀ1).17 The S–N stretching
mode appears at 736.7 cmꢀ1, which is also close to that in
CF3S(O)2N (723.2 cmꢀ1).18b
A typical IR spectrum of the pyrolysis products in the solid
Ne matrix at 2.8 K is shown in Fig. 2B.
Compared to the IR spectrum of the starting material
PhS(O)2N3 (Fig. 2A), almost 90% of the azide decomposes at
800 K. As a result, SO2 (e, 1358.6 cmꢀ1 and 1151.0 cmꢀ1 30 and
)
PhN (d, 1683.6, 1552.6, 1525.6, 1428.1, 1408.5, 1312.3, 1280.8,
1249.3, 1146.1, 1078.5, 1008.2, 963.3, 885.9, 819.1, 747.5,
655.4 cmꢀ1 31
were identified by IR spectroscopy as the domi-
)
nant pyrolysis products. To further prove the identification of
PhN, subsequent irradiation of the matrix with visible light
(l = 380–450 nm) was applied. The corresponding IR difference
spectrum (Fig. 2C) shows the exclusive depletion of PhN (d) and
the formation of didehydroazepine (f, cyclic C6H5N). The same
photo-induced ring expansion for PhN has been usually
observed during the photolysis of phenyl azide in noble gas
matrices.31
Therefore, the thermal decomposition of PhS(O)2N3 is
very different from those of FS(O)2N3,10c CF3S(O)2N3,18a and
CH3OS(O)2N3.10d In addition to the common IR inactive product
N2, mainly the thermally persistent sulfonyl nitrene FS(O)2N was
obtained in the FVP of FS(O)2N3. As for CF3S(O)2N3, the thermo-
lysis products were found to be CF3 and NSO2. Whereas CH2O
and HNSO2 were identified by IR spectroscopy for CH3OS(O)2N3.
Two possible pathways leading to the formation of PhN, SO2,
and N2 from PhS(O)2N3 might be involved. One is the concerted
migration of the N3 group and elimination of the SO2 moiety.
Then, as the key intermediate species, phenylazide (PhN3)
should be generated followed by its complete decomposition into
PhN and N2 under the pyrolysis conditions. However, according to
the latest study of the flash vacuum pyrolysis of PhN3, some of the
According to the TD-B3LYP/6-311++G(3df,3pd) calculation
(Table S3 in the ESI†), triplet PhS(O)2N is expected to absorb
visible light with predicted vertical transition at 425 nm
( f = 0.054). Therefore, the matrix containing the 193 nm laser
photolysis products of PhS(O)2N3 was further subjected to
Scheme 2 Photodecomposition of PhS(O)2N3 and photochemistry of
azide can survive at 800 K.32 The absence of the IR bands for PhN3 PhS(O)2N.
Phys. Chem. Chem. Phys.
This journal is ©the Owner Societies 2017

View Article Online
PCCP
Paper
Table 2 Observed and calculated IR frequencies (cmꢀ1) of PhS(O)2Na
Observedb
Ne matrix
Calculatedc
B3LYP
Approximate
assignmentd
Ar matrix
M06-2X
3215 (7)
3220 (7)
n(CH)
n(CH)
n(CH)
n(CH)
n(CH)
n(CC)
r(CH)
r(CH)
r(CH)
r(CH)
nas(SO2)
r(CH)
r(CH)
r(CH)
3214 (o1)
3218 (o1)
3208 (o1)
3203 (4)
3199 (5)
3191 (7)
3178 (o1)
1618 (28)
1611 (o1)
1507 (o1)
1477 (18)
1353 (2)
1336 (148)
1331 (o1)
1204 (o1)
1189 (o1)
1163 (206)
1104 (4)
1090 (16)
1040 (o1)
1026 (o1)
1007 (o1)
998 (3)
963 (o1)
864 (o1)
777 (30)
3189 (o1)
1658 (o1)
1657 (15)
1519 (o1)
1492 (21)
1427 (169)
1347 (14)
1334 (o1)
1230 (210)
1205 (18)
1185 (o1)
1123 (28)
1114 (3)
1057 (o1)
1043 (o1)
1024 (o1)
1005 (o1)
977 (o1)
882 (o1)
797 (12)
1588.7 m
1586.6 m
1452.5 m
1348.9 s
1450.4 m
1336.3 s
1168.2 vs
1090.8 w
1160.6 vs
1087.6 w
ns(SO2)
r(CH)
n(SC) + r(CH)
Ring distortion
Ring distortion
o(CH)
Ring breathing
o(CH)
Fig. 3 (A) IR spectrum of PhS(O)2N3 in the solid Ne matrix at 2.8 K; (B) IR
difference spectrum showing the change of PhS(O)2N3 in the solid Ne
matrix upon the 193 nm laser photolysis (3 Hz, 5 mJ, 10 min); (C) IR
difference spectrum showing the change of the matrix upon subsequent
visible light irradiation (l = 380–450 nm, 20 W, 5 min), bands for the
depleted species point downwards and bands for the formed species point
upwards; (D) calculated IR spectrum of triplet PhS(O)2N at the UB3LYP/
6-311++G(3df,3pd) level. The IR bands of PhS(O)2N3 (a), PhS(O)2N (b),
PhNSO2 (c), SO2 (e), cyclic C6H5N (f), and H2O (h) are labeled.
1000.2 vw
999.9 vw
o(CH)
766.7 m
736.7 w
765.9 m
737.3 m
n(SC)
734 (16)
697 (3)
763 (60)
721 (11)
n(SN)
d(CH)
683.7 m
682.5 m
690 (30)
625 (o1)
703 (30)
622 (o1)
d(CH)
d(CH)
a
A full list of the calculated harmonic IR frequencies (unscaled) is
b
given in Table S2 of the ESI. Band positions and intensities: vs, very
c
strong; s, strong; m, medium strong; w, weak; vw, very weak. Calculated
visible light irradiation (380–450 nm). All the IR bands of
PhS(O)2N (b) vanished completely, and SO2 (e) and cyclic
C6H5N (f) were produced (Fig. 3C). Additionally, a new carrier
with a weak but distinguishable band at 1372.7 cmꢀ1 (c) was
formed after irradiation. The band position is very close to the
predicted strongest IR band at 1376 cmꢀ1 for the SO2 asymmetric
stretching mode in the expected pseudo-Curtius rearrangement
product PhNSO2 (Table S4 in the ESI†). However, the attempt to
increase the abundance of the carrier for this band upon pro-
longed 193 nm laser irradiation of the azide was unsuccessful.
Considering the existence of two UV absorption bands at 270
and 200 nm for PhS(O)2N3, a 266 nm laser was also used for the
photolysis of the azide. The IR difference spectra showing the
changes of the azide in a solid Ar matrix after the successive
266 nm laser and visible light (380–450 nm) irradiation are Fig. 4B
and C, respectively. For comparison, the theoretically predicted IR
spectrum of PhNSO2 at the B3LYP/6-311++G(3df,3pd) level is also
depicted (Fig. 4D).
with the 6-311++G(3df,3pd) basis set, IR intensities (km molꢀ1) in
d
parentheses. Tentative assignment based on the calculated vibrational
displacement vectors.
Similar to the 193 nm laser irradiation, the 266 nm laser
irradiation caused efficient depletion of the azide and forma-
tion of the sulfonyl nitrene PhS(O)2N (b, 1586.6, 1450.4, 1336.3,
1160.6, 1087.6, 999.9, 765.9, 737.3, and 682.5 cmꢀ1). In contrast,
the subsequent visible light irradiation enables much more
Fig. 4 (A) IR spectrum of PhS(O)2N3 in the solid Ar matrix at 2.8 K; (B) IR
difference spectrum showing the change of PhS(O)2N3 in the solid Ar
confident identification of the IR bands for PhNSO2 (c). In matrix upon 266 nm laser photolysis (10 mW, 10 min); (C) IR difference
addition to the band at 1367.1 cmꢀ1 (1372.7 cmꢀ1 in Ne matrix),
spectrum showing the change of the matrix upon subsequent visible light
irradiation (l = 380–450 nm, 20 W, 5 min), bands for the depleted species
point downwards and bands for the formed species point upwards;
(D) calculated IR spectrum of PhNSO2 at the B3LYP/6-311++G(3df,3pd) level.
several weaker bands at 1484.6, 1328.6, 1280.3, 1190.4, 1005.8,
758.0, and 686.9 cmꢀ1 were found to be associated with the same
carrier (Fig. 4C). The band positions are in good agreement with
The IR bands of PhS(O)2N3 (a), PhS(O)2N (b), PhNSO2 (c), cyclic C6H5N (f),
the calculated values at 1520, 1376, 1349, 1310, 1189, 1013, 775, and H2O (h) are labeled.
This journal is ©the Owner Societies 2017
Phys. Chem. Chem. Phys.

View Article Online
Paper
PCCP
and 689 cmꢀ1 using the B3LYP/6-311++G(3df,3pd) method related 2-napththylsulfonyl nitrene10b and p-toylsufonyl nitrene10a
(Table S4 in the ESI†). Based on the calculated vibrational were found to quickly relax from the singlet state into the
displacement vectors, the SO2 asymmetric stretching vibration longer-lived triplet state in CCl4 with lifetimes of 700 ꢂ 300 and
couples heavily with the ring breathing vibration and appears at 660 ꢂ 100 ps, respectively, as determined by time-resolved
1005.8 cmꢀ1. It should be noted that the IR bands for PhNSO2 infrared spectroscopy.
(1367.1 and 1328.6 cmꢀ1) in the Ar matrix are quite close to the
previously observed two bands at 1371 and 1323 cmꢀ1 in the N2 ment from PhS(O)2N to PhNSO2, which needs to surmount a
matrix during the 265 nm photolysis of PhS(O)2N3.16 moderate barrier of 103 kJ molꢀ1. More importantly, the latter
The second process refers to the intramolecular rearrange-
The unambiguous identification of the IR bands for PhS(O)2N is expected to be a thermally persistent species due to the
and PhNSO2 clearly demonstrates that the stepwise Curtius extremely large C–N bond dissociation energy (BDE) of 336 kJ molꢀ1
rearrangement occurs during the photolysis of PhS(O)2N3 in (Ph–NSO2 - Ph + O2SN). The absence of PhNSO2 among the
cryogenic matrices (2.8 K). The absence of these two inter- pyrolysis products confirms that no rearrangement occurred
mediates among the pyrolysis products suggests that they are for the initially generated singlet nitrene even at 800 K.
probably not stable enough to survive under the pyrolysis condi-
tions (800 K).
The last possible process for singlet PhS(O)2N is the direct
fragmentation of the C–S bond into the radical pair Ph and
O2SN. Although the BDE (82 kJ molꢀ1) is even lower than the
activation barrier for the rearrangement (103 kJ molꢀ1), none of
Quantum chemical calculation
To account for the mechanism, the potential energy surface these two species was experimentally observed in the decom-
(PES) for the decomposition of PhS(O)2N3 is computationally position of the azide. Thus, it can be concluded that the
explored at the CBS-QB3 and B3LYP/6-311++G(3df,3pd) levels, prevailing channel for singlet PhS(O)2N is the spin relaxation
and the results are summarized in Fig. 5.
into the triplet state, which is fully consistent with the observa-
Similar to the PESs of carbonyl azides,4a,5 two distinct path- tion of the triplet nitrene as the dominant photolysis products
ways (stepwise versus concerted) for the pseudo-Curtius rearrange- in the cryogenic matrices.
ment of PhS(O)2N3 into PhNSO2 and N2 were computationally
Despite the fact that the triplet nitrene can be trapped in
located. The activation barrier for the stepwise route to the nitrene cryogenic matrices, its metastability in the gas phase can be
PhS(O)2N and N2 is 145 kJ molꢀ1 at the CBS-QB3 level, which is evidenced by the activation barrier of 94 kJ molꢀ1 for its rearran-
lower in energy than that for the concerted by 14 kJ molꢀ1. As gement into triplet PhNSO2 (Fig. 5). Unlike singlet PhNSO2, the
for the initially generated nitrene in the singlet state (1A), there triplet species is highly unstable because of the rather weak C–N
are three competing processes for its disappearance. One is the bond with the BDE of merely 13 kJ molꢀ1. Therefore, the complete
intersystem crossing (ISC) into the lower-energy triplet state dissociation fragments SO2, triplet PhN, and N2 are expected to be
(3A). The singlet–triplet energy gap (DEST) is estimated to be obtained as the pyrolysis products of PhS(O)2N3, which is fully
30 kJ molꢀ1, which is very close to that of CH3S(O)2N (34 kJ molꢀ1
and CF3S(O)2N (30 kJ molꢀ1) at the same theoretical level.17 In fact,
)
consistent with the experimental observation.
To gain more insight into the stability of the key nitrene
the ISC for PhS(O)2N can be extremely fast since the closely intermediate PhS(O)2N, the molecular structures in the singlet
Fig. 5 Calculated potential energy surface for the decomposition of PhS(O)2N3 in the singlet state at the CBS-QB3 and B3LYP/6-311++G(3df,3pd)
(in parentheses) levels of theory. The decomposition pathway of PhS(O)2N in the triplet state (in red colour) is also given. The relative energies are given
in kJ molꢀ1. The calculated molecular structures for the transition states are depicted. The symbols for nitrogen (blue), oxygen (red), and sulfur (yellow)
atoms are shown, and the symbols for carbon (grey) and hydrogen (turquoise) atoms are omitted for clarity.
Phys. Chem. Chem. Phys.
This journal is ©the Owner Societies 2017

View Article Online
PCCP
Paper
decomposition products PhN and SO2. The successful isolation
of singlet PhNSO2 from the photo-induced rearrangement of
triplet PhS(O)2N in solid noble gas matrices suggests the possible
involvement of even short-lived excited state species under the
photolysis conditions.
Acknowledgements
Fig. 6 Optimized structures of PhS(O)2N in the singlet (1A) and triplet (3A00)
states at the B3LYP/6-311++G(3df, 3pd) level of theory. Selected bond
lengths [Å] and angles [1, in italics] are shown. The molecular symmetries
are given in parentheses.
This work was supported by the National Natural Science Foun-
dation of China (21422304, 21372173 and 21673147), the Priority
Academic Program Development of Jiangsu Higher Education
Institutions (PAPD) and the China Postdoctoral Science Founda-
tion (2016M600436). The authors greatly acknowledge Professor
Robert S. Sheridan for sharing the unpublished results.
(1A) and triplet (3A) states were optimized at the B3LYP/
6-311++G(3df,3pd) level of theory. The key structural para-
meters are shown in Fig. 6. The most characteristic difference
between the two spin states of PhS(O)2N is the much narrower
O–S–N angle in the singlet state (70.01) than that in the triplet
state (106.71). Very similar differences have been found not only
for other sulfonyl nitrenes like CF3S(O)2N (singlet: 72.21; triplet:
107.91)18b and CH3S(O)2N (singlet: 70.71; triplet: 106.61)17 and
also for other a-oxo nitrenes RC(O)N,5 R2P(O)N,8 and RS(O)N.9b
The significant structural difference can be rationalized by
an intramolecular interaction between the electron-deficient
nitrene center and the adjacent oxygen atom (Nꢁ ꢁ ꢁO), which
stabilizes the closed-shell singlet by forming an oxazirine-like
structure with the Nꢁ ꢁ ꢁO distance of 1.767 Å. In sharp contrast,
the Nꢁ ꢁ ꢁO distance in the triplet state is 2.503 Å. Nevertheless,
the triplet state remains as the ground state for PhS(O)2N,
whereas the closely related carbonyl analogue PhC(O)N prefers
singlet multiplicity.7
References
¨
1 (a) S. Brase and K. Banert, Organic Azides: Syntheses and
Applications, John Wiley Sons, Chichester, 2010;
&
(b) P. Portius and M. Davis, Coord. Chem. Rev., 2013,
257, 1011; (c) L. H. Liu and M. D. Yan, Acc. Chem. Res.,
2010, 43, 1434.
2 (a) D. E. Falvery and A. D. Gudmundsdottir, Nitrenes and
Nitrenium Ions, Wiley Series of Reactive Intermediates in
Chemistry and Biology, John Wiley & Sons, Hoboken, 2013;
(b) N. P. Gritsan, M. S. Platz and W. T. Borden, The Study of
Nitrenes by Theoretical Methods, Taylor and Francis, Boca
Raton, 2005; (c) R. A. Moss, M. S. Platz and M. Jones, Jr.,
Reactive Intermediate Chemistry, Wiley-Interscience, Hoboken,
2004.
3 T. Curtius, Ber. Dtsch. Chem. Ges., 1890, 23, 3023.
4 (a) K. Banert, C. Berndt, M. Hagedorn, H. Liu, T. Anacker,
J. Friedrich and G. Rauhut, Angew. Chem., Int. Ed., 2012,
51, 4718; (b) X. Q. Zeng, E. Bernhardt, H. Beckers, K. Banert,
M. Hagedorn and H. Liu, Angew. Chem., Int. Ed., 2013,
52, 3503, and references therein.
Conclusions
Both thermal and photolytic decomposition reactions of the
parent arylsulfonyl azide PhS(O)2N3 have been studied by
combining matrix isolation IR spectroscopy and quantum
chemical calculations. Upon flash vacuum pyrolysis at 800 K,
the azide eliminates N2 and exclusively furnishes SO2 and PhN,
which provides a new method for generating the important
phenylnitrene in the gas phase. In contrast, laser photolysis
(193 and 266 nm) of PhS(O)2N3 in cryogenic matrices enables
the identification of the key nitrene intermediate PhS(O)2N
in the triplet ground state, which undergoes rearrangement
into N-sulfonyl imine PhNSO2 under visible light irradiation
(380–450 nm). The identification of these two intermediates firmly
supports the feasibility of stepwise Curtius rearrangement for
arylsulfonyl azides via the corresponding nitrene intermediate.
Computationally, the potential energy surface (PES) for the
decomposition of PhS(O)2N3 reveals the preference of the step-
wise Curtius rearrangement via the intermediacy of PhS(O)2N.
However, the initially generated nitrene in the singlet state favors
intersystem crossing to the triplet ground state, which may
rearrange into PhNSO2 on the triplet PES followed by immediate
dissociation into the experimentally observed final thermal
5 J. Liu, S. Mandel, C. M. Hadad and M. S. Platz, J. Org. Chem.,
2004, 69, 8583.
6 (a) J. M. Dyke, G. Levita, A. Morris, J. S. Ogden, A. A. Dias,
M. Algarra, J. P. Santos, M. L. Costa, P. Rodrigues, M. M.
Andrade and M. T. Barros, Chem. – Eur. J., 2005, 11, 1665;
(b) R. E. Wilde, T. K. K. Srinivasan and W. Lwowski, J. Am.
Chem. Soc., 1971, 93, 860; (c) J. H. Teles and G. Maier, Chem.
Ber., 1989, 122, 745.
7 (a) E. A. Pritchina, N. P. Gritsan, A. Maltsev, T. Bally,
T. Autrey, Y. L. Liu, Y. H. Wang and J. P. Toscano, Phys.
Chem. Chem. Phys., 2003, 5, 1010; (b) C. Wentrup and
H. Bornemann, Eur. J. Org. Chem., 2005, 4521.
8 (a) S. Vyas, S. Muthukrishnan, J. Kubicki, R. D. McCulla,
G. Burdzinski, M. Sliwa, M. S. Platz and C. M. Hadad, J. Am.
Chem. Soc., 2010, 132, 16796; (b) R. D. McCulla, G. A. Gohar,
C. M. Hadad and M. S. Platz, J. Org. Chem., 2007, 72, 9426.
9 (a) T. J. Maricich and V. L. Hoffman, J. Am. Chem. Soc., 1974,
96, 7770; (b) Z. Wu, D. Q. Li, H. M. Li, B. F. Zhu, H. L. Sun,
J. S. Francisco and X. Q. Zeng, Angew. Chem., Int. Ed., 2016,
This journal is ©the Owner Societies 2017
Phys. Chem. Chem. Phys.

View Article Online
Paper
PCCP
55, 1507; (c) T. J. Maricich and C. N. Angeletakis, J. Org. 22 C. Adamo and V. Barone, J. Chem. Phys., 1998, 108, 664.
Chem., 1984, 49, 1931.
23 (a) T. Clark, J. Chandrasekhar, G. W. Spitznagel and
P. V. R. Schleyer, J. Comput. Chem., 2004, 4, 294; (b) M. J.
Frisch, J. A. Pople and J. S. Binkley, J. Chem. Phys., 1984,
80, 3265.
10 (a) A. V. Kuzmin, C. Neumann, L. J. G. W. van Wilderen,
B. A. Shainyan and J. Bredenbeck, Phys. Chem. Chem. Phys.,
2016, 18, 8662; (b) J. Kubicki, H. L. Luk, Y. Zhang, S. Vyas,
H. L. Peng, C. M. Hadad and M. S. Platz, J. Am. Chem. Soc., 24 J. A. Montgomery, M. J. Frisch, J. W. Ochterski and
2012, 134, 7036; (c) X. Q. Zeng, H. Beckers and H. Willner, G. A. Petersson, J. Chem. Phys., 2000, 112, 6532.
J. Am. Chem. Soc., 2013, 135, 2096; (d) G. H. Deng, Z. Wu, 25 (a) R. E. Stratmann, G. E. Scuseria and M. J. Frisch, J. Chem.
D. Q. Li, R. Linguerri, J. S. Francisco and X. Q. Zeng, J. Am.
Chem. Soc., 2016, 138, 11509.
Phys., 1998, 109, 8218; (b) J. B. Foresman, M. Head-Gordon,
J. A. Pople and M. J. Frisch, J. Phys. Chem., 1992, 96, 135.
11 (a) T. Miura, Y. Funakoshi, M. Morimoto, T. Biyajima and 26 (a) C. Gonzalez and H. B. Schlegel, J. Chem. Phys., 1989,
M. Murakami, J. Am. Chem. Soc., 2012, 134, 17440;
(b) H. D. Xu, Z. H. Jia, K. Xu, M. Han, S. N. Jiang, J. Cao,
90, 2154; (b) C. Gonzalez and H. B. Schlegel, J. Phys. Chem.,
1990, 94, 5523.
J. C. Wang and M. H. Shen, Angew. Chem., Int. Ed., 2014, 27 M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria,
53, 9284; (c) Y. D. Li, Y. D. Feng, L. H. Xu, L. H. Wang and
X. L. Cui, Org. Lett., 2016, 18, 4924; (d) X. S. Xiao, C. Hou,
Z. H. Zhang, Z. F. Ke, J. Y. Lan, H. F. Jiang and W. Zeng,
Angew. Chem., Int. Ed., 2016, 55, 11897.
M. A. Robb, J. R. Cheeseman, G. Scalmani, V. Barone,
B. Mennucci, G. A. Petersson, H. Nakatsuji, M. Caricato,
X. Li, H. P. Hratchian, A. F. Izmaylov, J. Bloino, G. Zheng,
J. L. Sonnenberg, M. Hada, M. Ehara, K. Toyota, R. Fukuda,
J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O. Kitao,
H. Nakai, T. Vreven, J. J. A. Montgomery, J. E. Peralta,
F. Ogliaro, M. Bearpark, J. J. Heyd, E. Brothers, K. N. Kudin,
V. N. Staroverov, T. Keith, R. Kobayashi, J. Normand,
K. Raghavachari, A. Rendell, J. C. Burant, S. S. Iyengar,
J. Tomasi, M. Cossi, N. Rega, J. M. Millam, M. Klene,
J. E. Knox, J. B. Cross, V. Bakken, C. Adamo, J. Jaramillo,
R. Gomperts, R. E. Stratmann, O. Yazyev, A. J. Austin,
R. Cammi, C. Pomelli, J. W. Ochterski, R. L. Martin,
K. Morokuma, V. G. Zakrzewski, G. A. Voth, P. Salvador,
12 (a) M. T. Reagan and A. Nickon, J. Am. Chem. Soc., 1968,
90, 4096; (b) R. A. Abramovitch and R. G. Sutherland,
Top. Curr. Chem., 1970, 16, 1; (c) R. A. Abramovitch and
W. D. Holcomb, Chem. Commun., 1969, 1298.
13 (a) W. Lwowski and E. Scheiffele, J. Am. Chem. Soc., 1965,
87, 4359; (b) R. A. Abramovitch, W. D. Holcomb and
S. Wake, J. Am. Chem. Soc., 1981, 103, 1525; (c) D. S.
Breslow, E. I. Edwards, E. C. Linsay and H. Omura, J. Am.
Chem. Soc., 1976, 98, 4268.
14 J. C. Garay, V. Maloney, M. Marlow and P. Small, J. Phys.
Chem., 1996, 100, 5788.
¨
J. J. Dannenberg, S. Dapprich, A. D. Daniels, O. Farkas, J. B.
15 J. Kubicki, Y. L. Zhang, J. D. Xue, H. L. Luk and M. S. Platz,
Phys. Chem. Chem. Phys., 2012, 14, 10377.
Foresman, J. V. Ortiz, J. Cioslowski and D. J. Fox, Gaussian 09,
Revision A.2, Gaussian, Inc., Wallingford CT, 2009.
16 R. S. Sheridan and P. Rempala, Books of Abstracts, 28 O. C. Dermer and M. T. Edmison, J. Am. Chem. Soc., 1955,
217th ACS National Meeting, Anaheim, CA, American
Chemical Society, Washington, DC, 1999; ORGN-287.
17 G. H. Deng, D. Q. Li, Z. Wu, H. M. Li, E. Bernhardt and
X. Q. Zeng, J. Phys. Chem. A, 2016, 120, 5590.
77, 70.
29 X. Q. Zeng, M. Gerken, H. Beckers and H. Willner, J. Phys.
Chem. A, 2010, 114, 7624.
30 X. Q. Zeng, H. Beckers, P. Neuhaus, D. Grote and W. Sander,
Z. Anorg. Allg. Chem., 2012, 638, 526.
18 (a) X. Q. Zeng, H. Beckers and H. Willner, Angew. Chem.,
Int. Ed., 2013, 52, 7981; (b) X. Q. Zeng, H. Beckers, 31 J. C. Hayes and R. S. Sheridan, J. Am. Chem. Soc., 1990,
H. Willner, P. Neuhaus, D. Grote and W. Sander, J. Phys.
Chem. A, 2015, 119, 2281.
112, 5879.
´
32 J. Mierse-Perez, E. Mendez-Vega, K. Velappan and
19 J. Waser, B. Gaspar, H. Nambu and E. M. Carreira, J. Am.
Chem. Soc., 2006, 128, 11693.
20 A. D. Becke, J. Chem. Phys., 1993, 98, 5648.
21 Y. Zhao and D. G. Truhlar, Theor. Chem. Acc., 2008, 120, 215.
W. Sander, J. Org. Chem., 2015, 80, 11926.
´
´
33 H. M. Li, H. B. Wan, Z. Wu, D. Q. Li, D. Begue, C. Wentrup
and X. Q. Zeng, Chem. – Eur. J., 2016, 22, 7856, and
references therein.
Phys. Chem. Chem. Phys.
This journal is ©the Owner Societies 2017