The iminyl radical O2SN

Article abstract of DOI:10.1002/anie.201302968

The novel iminyl radical O₂SN, a SO₃ analogue, was produced by flash vacuum pyrolysis of gaseous alkyl sulfonyl azides RSO ₂N₃ (R=CF₃, CH₃) mixed with argon. Photoisomerization (λ>360 nm) of O₂SN into planar syn-OSNO and anti-OSNO (see picture) was observed in solid noble-gas matrices. Copyright

Full text of DOI:10.1002/anie.201302968

Angewandte
Chemie
DOI: 10.1002/anie.201302968
Small Molecules
The Iminyl Radical O2SN**
Xiaoqing Zeng,* Helmut Beckers,* and Helge Willner
Dedicated to Prof. Werner Uhl on the occasion of his 60th birthday
Sulfonylazides, RSO2N3, are widely used diazo[1] and azide[2]
transfer reagents, and in particular the photolytic or thermal
elimination of N2 from these azides have been well-studied.[3]
Although the latter routes are generally used for the in situ
generation of sulfonyl nitrenes, RSO2N, both were found to be
complicated owing to secondary product formation, such as
ble carriers of the new IR bands, the deposit was subjected to
light with wavelength l > 360 nm and radiation from a 365 nm
LED source, which both were found to efficiently deplete the
same set of the new bands. The mid-IR difference spectrum
obtained from the first experiment is shown in Figure 1.
SO2 + RN, as well as the pseudo Curtius rearrangement
[3,4]
=
product RN SO2.
Recently it became apparent that initially formed reactive
singlet sulfonyl nitrenes might be rapidly quenched by
efficient intersystem crossing (ISC) to yield the lower-
energy triplet nitrenes of sluggish reactivity.[4] Thermally
persistent triplet FSO2N was produced by flash pyrolysis (ca.
8008C) of FSO2N3 in yields up to 66%.[5] Only traces of SO2
and FSO2 were found as byproducts, and the lifetime of triplet
FSO2N in the gas phase was found to be dominated by
precursor concentration-dependent triplet nitrene dimeriza-
tion.[5] Triplet nitrenes, produced by flash-pyrolysis of selected
azides, have also offered unique access to synthetically
[6]
challenging molecules, such as OCN NCO, cyclo-N2CO,[7]
ꢀ
OPN/ONP,[8] SPN/SNP/cyclo-PSN,[9] and the FSO2 radical.[5]
Herein, we extend our flash vacuum pyrolysis studies to
the simple alkyl sulfonyl azides CF3SO2N3 and CH3SO2N3,
and present the detection and the photochemistry of the novel
O2SN (sulfonyliminyl) radical.
Figure 1. IR spectral changes in the region of 1500–900 cmꢀ1 obtained
after photolysis (l>360 nm) of Ar matrix-isolated (16 K) pyrolysis
products of CF3SO2N3. Bands assigned to O2SN (a) decreased, and
bands that are due to SO2, syn (b) and anti OSNO (c) appeared.
Residual bands of the CF3 radical (marked by asterisks) occurred in
the difference spectrum owing to matrix site changes upon photolysis.
Flash vacuum pyrolysis of CF3SO2N3 highly diluted in
argon (1:500) was performed by passing the gas mixture
through a hot quartz furnace (ca. 8008C, inside diameter
1.0 mm, length 30 mm). The reaction mixture was immedi-
ately deposited onto the cold matrix support (16 K) in a high
vacuum (see the Supporting Information for details). The IR
spectrum of the deposit reveals almost complete decomposi-
tion of the azide (Supporting Information, Figure S1). Along
with the known IR bands of SO2, SO3, CF3, and C2F6, there are
a number of new IR bands showing distinct 15N isotopic shifts
in experiments using a mixture of CF3SO215NaNN and
CF3SO2NN15Ng (1:1). To distinguish between different possi-
Based on their same photolytic behavior, four IR bands at
1358.6, 1328.3, 1216.9, and 966.9 cmꢀ1 in the mid-IR spectrum
(Figure 1, a) are assigned to species a. Its IR spectrum was
completed by recording three additional bands in the far-IR
region (Figure 2) at 474.5, 413.2, and 367.6 cmꢀ1. Each band
exhibits a satellite that is due to matrix sites, and along with
the complete set of 14/15N isotopic shifts, bands associated with
the naturally abundant 34S isotopologue were also observed.
The experimental vibrational data are compared in Table 1
with predicted values for the planar O2SN (C2v) radical
obtained at the DFT B3LYP/6-311 + G(3df) level. With the
exception of the band at 1328.3 cmꢀ1, the agreement is very
good. This band is attributed to the combination 966.9 (a1) +
367.6 (b2) = 1334.5 cmꢀ1 (b2), which takes intensity through
[*] Prof. Dr. X.-Q. Zeng
College of Chemistry, Chemical Engineering and Materials Science
Soochow University, 215123 Suzhou (China)
E-mail: xqzeng@suda.edu.cn
Dr. H. Beckers, Prof. Dr. H. Willner
FB C-Anorganische Chemie, Bergische Universitꢀt Wuppertal
Gaussstrasse 20, 42119 Wuppertal (Germany)
E-mail: beckers@uni-wuppertal.de
a
Fermi resonance from the strong fundamental at
1358.6 cmꢀ1 (b2). Apart from the six fundamentals, two
combination and two overtone bands of O2SN were observed
and assigned.
[**] Support from the Deutsche Forschungsgemeinschaft (DFG) (WI
663/26-1) for this work, and in particular for funding a research stay
of X.-Q.Z. in Wuppertal, is gratefully acknowledged.
The assignment of the IR spectrum of O2SN is given in
Table 1. The two strongest bands at 1358.6 and 1216.9 cmꢀ1
are readily attributed to antisymmetric (n5) and symmetric
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.201302968.
Angew. Chem. Int. Ed. 2013, 52, 7981 –7984
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
7981

.
Angewandte
Communications
The identification of the photolysis products of O2SN is
straightforward. Traces of SO2 formed during UV irradiation
(l > 360 nm) of Ar matrix-isolated O2SN (Figure 1) indicate
its photodecomposition into SO2 and nitrogen atoms. Apart
from the known SO2 bands, strong bands appeared at 1450.3
and 1152.9 cmꢀ1 (Figure 1, b) and weaker bands at 1459.2 and
1182.9 cmꢀ1 (Figure 1, c). These band positions are close to
reported NO and SO stretching frequencies for the open-
chain molecules syn OSNO (1450.8 and 1154.9 cmꢀ1) and anti
OSNO (1456.0 and 1181.2 cmꢀ1) isolated in argon matrices.[10]
Large 15N isotope shifts obtained for the NO stretches (syn
24.9 cmꢀ1, anti 24.8 cmꢀ1) support their assignment, whereas
the SO stretches are much less affected by 15N isotopic
substitution (Supporting Information, Figure S3).
In previous work, the bent OSNO isomers were obtained
in low yield by laser photolysis (248 nm) of matrices contain-
ing OCS and NO2.[10] Only their two strongest IR bands were
observed, and the presence of a CO molecules along with the
OSNO radicals in the matrix cage are responsible for the
differences between reported and our band positions. IR
radiation from the globar source was found to convert anti
OSNO to the more stable syn isomer.[10] We also observed this
conversion using 365 nm LED radiation.
The efficient photoisomerization of O2SN (C2v) into the
chain-like OSNO (CS) radicals allows for a more complete
analysis of their IR spectra (Figure 2). Their vibrational data
are summarized in Table 2 and Table 3, respectively. For the
syn isomer the weak SN stretching band (n3) appeared at
Figure 2. IR spectral changes in the far-IR region of 640–180 cmꢀ1
obtained after irradiation (l>360 nm) of the Ar matrix-isolated (16 K)
pyrolysis products of CF3SO2N3. Bands assigned to O2SN (a)
decreased, and bands due to SO2, syn (b) and anti OSNO (c)
appeared.
Table 1: Observed and calculated IR frequencies n, isotope shifts Dn
(cmꢀ1) and band assignments for the O2SN (2B2, C2v) radical.
n
Dn (14/15N)
Dn (32/34S) Mode
expt (Ar)[a]
calcd[b]
expt calcd expt calcd ni[c]
2700.8 (1)
2561.0 (1)
2424.5 (2)
1358.6 (100) 1336 (156)
1328.3 (18)
1216.9 (37)
966.9 (15)
474.5 (26)
413.2 (38)
367.6 (22)
<0.5
8.6
13.0
1.7
21.9
6.5
17.8
1.4
2n5
n1 +n5
2n1
0.1 16.5
6.4 13.5
17.4 n5, nas(SO2)
n3 +n6
Table 2: Observed and calculated IR frequencies n and isotope shifts Dn
(cmꢀ1) for syn OSNO (2A’, CS).
1201 (46)
954 (9)
462 (19)
414 (20)
358 (20)
14.0 n1, ns(SO2)
1.1 n2, n(SN)
2.5 n3, d(SO2)
7.5 n4, do.o.p.
1.8 n6, d(OSN)
n
Dn (14/15N)
expt calcd
Dn (32/34S) Mode
expt calcd
19.2
1.2
2.4
7.1
1.8
expt (Ar)[a]
calcd[b]
1.5
1.9
6.0
1.8
6.0
2866.7 (6)
1450.3 (100) 1461 (188)
48.3
24.9
2.3
5.6
10.9
7.5
<0.5
25.9 <0.5
2n1
0.2 n1, n(NO)
11.8 n2, n(SO)
8.9 n3, n(SN)
2.6 n4, d(OSN)
2.0 n5, d(ONS)
0.3 n6, t
1152.9 (87)
600.2 (4)
521.2 (15)
375.0 (11)
202.2 (4)
1144 (183)
614 (3)
520 (18)
393 (9)
0.3
7.0
11.2
8.1
11.8
8.4
2.3
1.8
[a] Band positions of the most intense matrix sites in solid argon at 16 K,
and relative intensities (in parenthesis) based on integrated areas of all
matrix sites. [b] Harmonic frequencies calculated at the DFT B3LYP/6-
311+G(3df) level of theory and scaled by a factor of 0.9679. Calculated
IR intensities (km molꢀ1) are given in parentheses. [c] Numbering of
fundamentals ni (i=1–6) according to their symmetry (a1: i=1–3, b1: 4,
b2: 5 and 6) and frequency.
200 (3)
<0.5
0.4 <0.5
[a] Band position of the most intense matrix site in solid argon at 16 K,
relative intensities (in parenthesis) based on the integrated areas of all
matrix sites. [b] Harmonic frequencies calculated at the B3LYP/6-
311+G(3df) level of theory and scaled by a factor of 0.9679; calculated
IR intensities [kmmolꢀ1] are given in parentheses.
(n1) stretches of a SO2 moiety, respectively, and the assign-
ment of the SN stretch (n2) to the band at 966.9 cmꢀ1 is
supported by its large 15N isotope shift of 17.8 cmꢀ1. A
significant 34S isotope shift for the band at 413.2 cmꢀ1 is
consistent with its assignment to an out-of-plane mode n4 of
the planar radical. Thus, the band at 474.5 cmꢀ1 should be
attributed to the SO2 bending mode (n3), which is lower in
frequency than that of free SO2 (516.7 cmꢀ1, Ar matrix) owing
to vibrational coupling with the SN stretch in O2SN, and the
lowest energy vibration at 367.6 cmꢀ1 to the OSN bending
mode n6. We note that the O2SN radical was also detected
among the flash pyrolysis products of CH3SO2N3 (Supporting
Information, Figure S2). However, the main products were
SO2 and CH2NH, and the yield of O2SN was significantly
lower compared to that found by pyrolysis of CF3SO2N3.
600.2 cmꢀ1 (Figure 2, b) along with three additional deforma-
tions in the far-IR region. Owing to a lower abundance of the
less-stable anti conformer, the assignment of its weak SN
stretch is difficult. A weak overtone band 2n1 was however
observed for both isomers.
Upon 266 nm laser irradiation, syn OSNO rearranged into
the less stable anti isomer. Using mercury arc radiation (l >
255 nm), traces of SNO radicals were also detected by means
of two weak IR bands at 1593.3 (2n2, Dn (14/15N) = 28.3 cmꢀ1)
and 1522.6 cmꢀ1 (n1, Dn (14/15N) = 24.7 cmꢀ1)[11] (Supporting
Information, Figure S4).
7982 www.angewandte.org
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2013, 52, 7981 –7984

Angewandte
Chemie
Table 3: Observed and calculated IR frequencies n and isotope shifts Dn
(cmꢀ1) for anti OSNO (2A’, CS).
n
Dn (14/15N)
Mode
expt (Ar)[a]
calcd[b]
expt
calcd
2887.3 (2)
1459.2 (67)
1182.9 (100)
49.8
24.8
1.0
2n1
1471 (253)
1162 (206)
582 (2)
451 (5)
279 (0.3)
261 (8)
26.7
0.1
10.9
4.0
5.3
2.1
n1, n(NO)
n2, n(SO)
n3, n(SN)
n4, d(OSN)
n6, t
452.0 (2)
272.6 (<1)
264.4 (3)
3.8
4.7
2.1
n5, d(ONS)
Figure 3. Calculated planar molecular structures (bond lengths in ꢁ) of
[a] Band position of the most intense matrix site in solid argon at 16 K,
relative intensities (in parenthesis) based on the integrated areas of all
matrix sites. [b] Harmonic frequencies calculated at the B3LYP/6-
311+G(3df) level of theory and scaled by a factor of 0.9679; the
calculated IR intensities [kmmolꢀ1] are given in parentheses.
O2SN (2B2, ffOSO=121.78), O2SNꢀ (1A1, ffOSO=110.18), syn OSNO
(2A’, ffOSN=115.38, ffSNO=134.98), and anti OSNO (2A’,
ffOSN=111.98, ffSNO=132.58) at the B3LYP/6-311+G(3df) level of
theory.
In the initially obtained IR spectrum of the pyrolysis
products of CF3SO2N3, some bands remained unassigned.
They also revealed distinct 15N isotope shifts (Supporting
Information, Figure S5), but remained almost unaffected by
any of the aforementioned irradiations. These bands
appeared at 1413.1, 1313.6, 1243.8, 1213.1, 1196.3, and
1045.4 cmꢀ1, indicating the presence of SO2 (1413.1,
1045.4 cmꢀ1) and CF3 (1243.8, 1213.1, 1196.3 cmꢀ1) group
vibrations. A large 15N isotope shifts (19.3 cmꢀ1) for the strong
energy by, respectively, 17.4 (13.6) kJmolꢀ1 and 15.2
(17.3) kJmolꢀ1 at the CCSD(T)//MP2 (CBS-QB3) levels of
theory. We note the existence of another weakly interacting
syn OS···NO minimum (SN 2.047 ꢀ; Supporting Information,
Table S2), having the unpaired electron in a bonding a’’
orbital; however, in the following we will focus on the
experimentally confirmed isomers only.
DFT calculations predict a planar molecular structure of
C2v symmetry for the O2SN radical (Figure 3). A comparison
band at 1313.6 cmꢀ1 suggests an S N stretching band. IR
with the related structure of the N-sulfonyl imine CF3N SO2,
=
=
frequencies calculated for the most likely carrier, N-sulfonyl
calculated at the same level of theory (see Supporting
=
=
imine CF3N SO2, fit very well with the observations (Sup-
Information), revealed similar S O bond lengths and OSO
=
porting Information, Table S1). Three weaker far-IR bands at
585.3, 471.9, and 406.0 cmꢀ1 were also assigned to this species.
The identification of intermediates from the flash pyrol-
ysis of sulfonyl azides RSO2N3 (R = CH3, CF3) sheds light on
their thermal decomposition mechanism and the formation of
O2SN. Upon pyrolysis, both azides are expected to eliminate
N2 furnishing the short-lived singlet nitrene RSO2N. Initially
formed singlet RSO2N may undergo: 1) ISC to the triplet
ground state; 2) Curtius-type rearrangement to RN = SO2; or
angles (O2SN: 1.429 ꢀ (S O), ffOSO = 121.78; CF3N = SO2:
=
1.427 and 1.422 ꢀ (S O), ffOSO = 121.78) for these two
=
related molecules. The S N bond in O2SN (1.517 ꢀ) is slightly
longer than that of CF3NSO2 (1.507 ꢀ).
Although O2SN (2B2) is isostructural and isoelectronic to
the known FCO2 (2B2) radical,[14] the electronic structures of
these two radicals are very different. The electron spin density
of FCO2 is equally shared between the two oxygen atoms,[14]
but localized at the nitrogen atom in O2SN. Calculations for
the closed-shell O2SNꢀ anion (Figure 3)[15] reveal a signifi-
ꢀ
3) R S bond fission to yield the radical pair O2SN and R. The
=
=
favorable formation of O2SN by flash pyrolysis of CF3SO2N3
according to (3) is consistent with the weak F3Cd+-Sd+ bond in
the nitrene, CF3SO2N, for which a dissociation energy of
66 kJmolꢀ1 (singlet CF3SO2N) is predicted by preliminary
calculations using the complete basis set CBS-QB3 method.[12]
cantly shortened S N (1.458 ꢀ) associated with elongated S
O bond lengths (1.477 ꢀ), indicating that the unpaired
electron in O2SN resides in a p(NS) bonding (b2) molecular
orbital. The O2SNꢀ anion is isoelectronic to recently detected
F2PN.[16] According to natural bond orbital (NBO) analysis,
two doubly occupied atomic p-type orbitals of nitrogen in
these molecules are engaged in p bonding to the central atom,
whereas in the O2SN radical the corresponding in-plane (b2)
HOMO is only half-filled (Supporting Information, Fig-
ure S6, Table S3). The lowest-energy transition between the
highest doubly occupied b1 and the singly occupied b2 orbitals
in O2SN is forbidden by symmetry (A2). However, time-
dependent (TD) DFT calculations predict reasonably strong
UV bands at 322 (A1, f = 0.0094), 224 (B1, 0.0025) and 217 nm
(B2, 0.0074). These predictions are consistent with the
observed photosensitivity of O2SN and the presence of two
structured UV absorptions centered at lmax ꢁ 350 nm and
below 250 nm (Supporting Information, Figure S7). The latter
absorption shows a regular vibrational spacing of about
490 cmꢀ1. These absorptions are associated with O2SN, as they
ꢀ
For the methyl analogue, the C S bond energies of singlet
CH3SO2N is considerably higher (117 kJmolꢀ1), suggesting
(2) and (3) are competing processes in the flash pyrolysis of
CH3SO2N3. The Curtius-type rearrangement (2) eventually
yields fragments (SO2 and CH2NH) of CH3NSO2, although
their formation by radical reactions between CH3 and NSO2,
formed under the pyrolysis conditions, cannot be excluded.
In recent studies on NO2S species,[10,13] the planar O2SN
radical was not considered at all. We have calculated the
molecular structures and energies of various planar NO2S
isomers on the doublet potential energy surface using differ-
ent theoretical methods (Supporting Information, Table S2).
At all of the levels applied, syn OSNO (2A’, Figure 3) was
found to be the lowest-energy isomer, O2SN (2B2) and anti
OSNO (2A’, Figure 3) were found to be slightly higher in
Angew. Chem. Int. Ed. 2013, 52, 7981 –7984
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
7983

.
Angewandte
Communications
vanished completely after irradiation with light of l > 360 nm
(Supporting Information, Figures S8 and S9).
Keywords: azides · flash pyrolysis · IR spectroscopy · nitrenes ·
rearrangement
.
The SN bond lengths predicted for the planar bent OSNO
isomers of 1.62–1.63 ꢀ (Figure 3) are between the expected
values for single (1.74 ꢀ) and double SN bonds (1.54 ꢀ).[17] In
a MO scheme, bonding interactions between the diatomic
fragments OS and NO may arise from a combinations of their
partially filled p* fragment orbitals (for the frontier molecular
orbitals, see the Supporting Information, Figure S6). For the
bent OSNOꢀ anion (24 valence electrons), the corresponding
in-plane (a’) and out-of-plane (a’’) (p*p*) bonding interac-
[1] J. Raushel, S. M. Pitram, V. V. Fokin, Org. Lett. 2008, 10, 3385 –
3388, and references therein.
[2] P. Panchaud, L. Chadaud, Y. Landais, C. Ollivier, P. Renaud, S.
Zigmantas, Chem. Eur. J. 2004, 10, 3606 – 3614, and references
therein.
[3] For examples, see: a) W. Lwowski, E. Scheiffele, J. Am. Chem.
Soc. 1965, 87, 4359 – 4365; b) R. A. Abramovitch, C. I. Azogu,
I. T. McMaster, J. Am. Chem. Soc. 1969, 91, 1219 – 1220; c) R. A.
Abramovitch, G. N. Knaus, V. Uma, J. Am. Chem. Soc. 1969, 91,
7532 – 7533; d) D. S. Breslow, E. I. Edwards, E. C. Linsay, H.
Omura, J. Am. Chem. Soc. 1976, 98, 4268 – 4275; e) J.-C. Garay,
V. Maloney, M. Marlow, P. Small, J. Phys. Chem. 1996, 100,
5788 – 5793.
[4] a) S. Vyas, J. Kubicki, H. L. Luk, Y. Zhang, N. P. Gritsan, C. M.
Hadad, M. S. Platz, J. Phys. Org. Chem. 2012, 25, 693 – 703; b) J.
Kubicki, Y. Zhang, S. Vyas, G. Burdzinski, H. L. Luk, J. Wang, J.
Xue, H. L. Peng, E. A. Pritchina, M. Sliwa, G. Buntinx, N. P.
Gritsan, C. M. Hadad, M. S. Platz, J. Am. Chem. Soc. 2011, 133,
9751 – 9761, and references therein.
[5] X. Q. Zeng, H. Beckers, H. Willner, J. Am. Chem. Soc. 2013, 135,
2096 – 2099.
[6] G. Maier, M. Naumann, H. P. Reisenauer, J. Eckwert, Angew.
Chem. 1996, 108, 1800 – 1801; Angew. Chem. Int. Ed. Engl. 1996,
35, 1696 – 1697.
[7] a) X. Q. Zeng, H. Beckers, H. Willner, J. F. Stanton, Angew.
Chem. 2011, 123, 1758 – 1761; Angew. Chem. Int. Ed. 2011, 50,
1720 – 1723; b) X. Q. Zeng, H. Beckers, H. Willner, J. F. Stanton,
Eur. J. Inorg. Chem. 2012, 3403 – 3409.
[8] X. Q. Zeng, H. Beckers, H. Willner, J. Am. Chem. Soc. 2011, 133,
20696 – 20699.
[9] X. Q. Zeng, H. Beckers, H. Willner, J. S. Francisco, Angew.
Chem. 2012, 124, 3390 – 3395; Angew. Chem. Int. Ed. 2012, 51,
3334 – 3339.
[10] M. Bahou, Y. P. Lee, J. Chem. Phys. 2001, 115, 10694 – 10700.
[11] M. Hawkins, A. J. Downs, J. Phys. Chem. 1984, 88, 3042 – 3047.
[12] J. A. Montgomery, M. J. Frisch, J. W. Ochterski, G. A. Petersson,
J. Chem. Phys. 1999, 110, 2822 – 2827.
[13] W.-C. Chen, C.-H. Yu, J. Chem. Phys. 2001, 115, 7495 – 7502.
[14] H. Beckers, H. Willner, D. Grote, W. Sander, J. Geier, J. Chem.
Phys. 2008, 128, 224301 – 224307.
[15] a) H. W. Roesky, W. Schmieder, W. Isenberg, D. Bçhler, G. M.
Sheldrick, Angew. Chem. 1982, 94, 143; Angew. Chem. Int. Ed.
Engl. 1982, 21, 153; b) N. H. Morgon, H. V. Linnert, J. M.
Riveros, J. Phys. Chem. 1995, 99, 11667 – 11672; c) R. Mews, P. G.
Watson, E. Lork, Coord. Chem. Rev. 1997, 158, 233 – 273.
[16] X. Q. Zeng, H. Beckers, H. Willner, Angew. Chem. 2009, 121,
4922 – 4925; Angew. Chem. Int. Ed. 2009, 48, 4828 – 4831.
[17] P. Pyykkç, M. Atsumi, Chem. Eur. J. 2009, 15, 12770 – 12779.
[18] M. P. Grubb, M. L. Warter, H. Xiao, S. Maeda, K. Morokuma,
S. W. North, Science 2012, 335, 1075 – 1078.
=
tions affords a formal S N double bond (Supporting Infor-
mation, Table S4). For OSNO (23 valence electrons), the
unpaired electron resides in the a’ bonding combination
(Supporting Information, Figure S6). However, the spin
densities for syn and anti OSNO were calculated to be
localized mainly at the NO moiety.
A strong UV absorption at lmax ꢁ 285 nm with a regular
vibrational spacing of about 1000 cmꢀ1 occurred in the
experimental spectrum after UV photolysis of O2SN (Sup-
porting Information, Figure S7), which is most likely associ-
ated with the bent OSNO (2A’) radicals. This assignment is
supported by TD DFT calculations, which predict strong UV
transitions (f > 0.09) for syn and anti OSNO at 266 and
269 nm, respectively (Supporting Information, Figure S7).
The anti to syn interconversion was predicted to proceed by
ꢀ
a low-energy N-inversion process rather than by S N bond
rotation, and for both isomers low barriers of less than
40 kJmolꢀ1 were reported for their dissociation into the
diatomic fragments OS + NO.[13] In contrast, the thermally
robust nature of the O2SN radical is confirmed by the absence
of SO and NO in the flash pyrolysis products. On the other
hand, photoisomerization from O2SN to bent OSNO radicals
may proceed either via a cyclic transition state or by
photodissociation and recombination of the photofragments
OSN + O in a matrix cage. The dissociative mechanism has
been suggested for the rearrangement of the related mole-
cules NO3 (!ON + O2)[18] and SO3 (!OSOO).[19]
In summary, flash vacuum pyrolysis of sulfonyl azides,
CF3SO2N3 and CH3SO2N3, yields the planar O2SN (2B2)
radical. IR and UV spectra of the novel radical have been
assigned and its photorearrangement into the planar bent
isomers, syn and anti OSNO (2A’), has been observed in solid
noble-gas matrices. The facile generation of O2SN in the gas
phase from two easily accessible azides will stimulate further
studies on its spectroscopy, structure, and chemistry.
Received: April 10, 2013
Published online: June 20, 2013
[19] S. H. Jou, M. Y. Shen, C. H. Yu, Y. P. Lee, J. Chem. Phys. 1996,
104, 5745 – 5753.
7984 www.angewandte.org
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2013, 52, 7981 –7984