Formation of thiazyl radicals by the thermolysis and photolysis of sulfur-nitrogen bicycles RCN5S3

Full text of DOI:10.1016/j.mencom.2007.06.005

Mendeleev
Communications
Mendeleev Commun., 2007, 17, 204–206
Formation of thiazyl radicals by the thermolysis
and photolysis of sulfur–nitrogen bicycles RCN5S3
Nina P. Gritsan,*a,b Konstantin V. Shuvaev,c,d Stanislav N. Kim,a,d Carsten Knapp,e
Rüdiger Mews,*e Victor A. Bagryanskya and Andrey V. Zibarev*b,c
a Institute of Chemical Kinetics and Combustion, Siberian Branch of the Russian Academy of Sciences,
630090 Novosibirsk, Russian Federation. Fax: +7 383 330 7350; e-mail: gritsan@ns.kinetics.nsc.ru
b Department of Physics, Novosibirsk State University, 630090 Novosibirsk, Russian Federation
c N. N. Vorozhtsov Novosibirsk Institute of Organic Chemistry, Siberian Branch of the Russian Academy of
Sciences, 630090 Novosibirsk, Russian Federation. Fax: +7 383 330 9752; e-mail: zibarev@nioch.nsc.ru
d Department of Natural Sciences, Novosibirsk State University, 630090 Novosibirsk, Russian Federation
e Institute for Inorganic and Physical Chemistry, University of Bremen, 28359 Bremen, Germany.
Fax: +49 421 218 4267; e-mail: mews@chemie.uni-bremen.de
DOI: 10.1016/j.mencom.2007.06.005
4-R-1,2,3,5-dithiadiazolyl radicals are identified by EPR spectroscopy upon thermal and photochemical decomposition of
7-R-1,3,5-trithia-2,4,6,8,9-pentaazabicyclo[3.3.1]nona-1(9),2,3,5,7-pentaenes in hydrocarbon solutions.
Thiazyl radicals are of fundamental interest, but they are
N5
N
S1
S2
N
S
S
especially fascinating due to their unusual physical properties
and potential applications as molecular magnets, molecular
conductors and magnetic switches.1 Recently, we demonstrated
that the thermal and photochemical degradation of sulfur–nitro-
gen heterocycles (isomeric 1,2,4,3,5- and 1,3,5,2,4-benzotrithia-
diazepines and 1,3,2,4-benzodithiadiazines) under mild condi-
tions afforded thiazyl radicals in nearly quantitative yields.2
Previously, the only route to these species was the reduction of
corresponding salts,1 whereas our findings indicated that stable
thiazyl radicals can be generated from neutral precursors. To
explore the generality of this phenomenon, we explored a series
of bicyclic compounds RCN5S3 (1)3 as possible precursors for
the formation of thiazyl radicals (Scheme 1).†
kT
4
R
R
N
N
– [SN· + N2]
3
N
S
N
1a–g
2a–g
Scheme 2
the EPR spectrum of 2b would only contain a five line pattern
from the coupling of an unpaired electron with two equivalent
14N nuclei (I = 1). A simulation of the entire EPR spectrum
[Figure 1(b)] indicated that the mixture actually represents a
superposition of resonance lines from two different kinds of
thiazyl radicals: a major radical with two equivalent nitrogen
nuclei (2b), and a minor one with three equivalent nitrogen
nuclei (3b).† However, the spectral features of minor radical 3b
are not so evident since the line-widths of the higher field
resonance components of both radical species are much larger
than those at a low field. This is explained by the very high
N
1e R = 2,6-F2C6H3
1f R = CF3
1g R = NMe2
1a R = Ph
N
S
S
1b R = 4-MeC6H4
1c R = 4-FC6H4
1d R = 4-PhC6H4
R
N
N
N
S
Compounds 1a–g were prepared as described previously.3(a)–(c)
‡
Scheme 1
All the experiments were performed in dodecane and squalane
(2,6,10,15,19,23-hexamethyltetracosane), photolysis was also performed
in CS2 solutions (c = 10–4–10–3 mol dm–3) degassed by five freeze-pump-
thaw cycles, in quartz tubes fitted with valves. Compounds 1a–e have
low solubility in hydrocarbons; therefore, the samples were prepared
under mild heating (~40 °C) in an ultrasonic bath.
A previous study showed compounds 1 to be prone to
thermal degradation in an acetonitrile solution, although only
benzonitrile and S4N4 were identified as decomposition products
in the case of 1a.4 Herein, we report that this process is more
complex than that described hitherto, and the mild thermolysis
of compounds 1 in hydrocarbon solutions results in the forma-
tion of thiazyl radicals, which were monitored by in situ EPR
spectroscopy.‡
The thermolysis of 1a–g was carried out using a glycerol thermostat in
the temperature range 363–403 K with an accuracy of 1 K.
The UV-visible spectra of aryl-substituted compounds 1 are charac-
terised by a long-wavelength shoulder at ~400 nm,3 those of aryl-sub-
stituted compounds 4, by a peak at ~410 nm,7 while CS2 is only
transparent to visible light (> 400 nm). Therefore, photolysis of 1b was
performed using a selected 436 nm line of a DRSh-500 mercury lamp
equipped with a water filter and a combination of glass filters.
The EPR spectra were recorded on a Bruker EMX spectrometer (MW
power of 0.64 mW, modulation frequency of 100 kHz and modulation
amplitude of 0.1 G). Spectral integration and simulation were performed
using the Win-EPR and Win-Sim programs, respectively. The radical
yields were determined by comparison of the integral intensities of
radicals with the integral intensity of a CuCl2·2H2O standard with an
accuracy of 15%.
Figure 1(a) shows, as a representative example, the EPR
spectrum of derivative 1b recorded after heating in squalane
at 363 K. A priori, one would expect the exclusive formation
†
·
of radicals 2 accompanied by loss of SN and N2 from the
RCN5S3 molecule (Scheme 2). Note that a series of radicals 2
obtained by other methods has previously been isolated and
revealed interesting magnetic and electrical properties.1
However, the complex nature of the spectrum (Figure 1)
suggested the presence of more than one radical species, since
†
Compounds names and numbering. 1: 7-R-1,3,5-trithia-2,4,6,8,9-pentaaza-
bicyclo[3.3.1]nona-1(9),2,3,5,7-pentaenes. 2: 4-R-1,2,3,5-dithiadiazolyls.
3: 3,5-R2-1,2,4,6-thiatriazinyls. 4: 3,7-R2-1,5,2,4,6-dithiatetrazocines.
Quantum-chemical calculations at the UB3LYP/6-31G(d) level of theory
were performed using the GAUSSIAN 98 suite of programs.9
– 204 –

Mendeleev Commun., 2007, 17, 204–206
(b)
1.0
0.5
0.0
(a)
3
2
1
0
0
3340
3350
3360
3370
3340
3350
3360
3370
H /10–4 T
H /10–4
T
0
100
200
300
Figure 1 (a) EPR spectra recorded after (1) 5, (2) 40 and (3) 320 min
thermolysis of 1b in squalane (c = 0.9×10–3 mol dm–3) at 363 K. (b) Experi-
mental EPR spectrum after 320 min thermolysis (dots) and its simulation
assuming the presence of two types of radicals (solid line): a radical with 2 N
atoms (94.2%, aN = 5.21 G) and a radical with 3 N atoms (∆g = –0.0043,
aN = 3.91 G).
Time/min
Figure 3 Kinetics of formation of radicals 2b (circles) and 3b (asterisks)
upon the thermolysis of 1b in squalane at 363 K and fitting of the data by
exponential time dependence (solid curve). A dashed line is given to
emphasize a constant amount of radical 3b.
viscosity of squalane solvent resulting in too slow molecular
tumbling to average out the partial anisotropy.6
363 K. It is evident from the kinetic data (Figure 3) that minor
radical 3b is formed much faster than major radical 2b and
its amount remains constant on further thermolysis. The only
explanation is the formation of 3b from an impurity. Unfortu-
nately, additional recrystallization of 1a from acetonitrile did
not lead to the reduction of radical 3a formation on thermolysis.
The overall yield of radicals 2a (and 2b) was close to
quantitative (to within 15%). Small amounts (< 5%) of radicals
3 were also detected upon thermolysis of derivatives 1c–e,
while only radicals 2f and 2g were detected for 1f and 1g.†
The irradiation of hydrocarbon solutions of 1a–g at ambient
temperature with UV (313 nm) or visible (436 nm) light pro-
duced EPR spectra similar to those obtained from the thermo-
lysis experiments. However, the solubility of bicycles RCN5S3
was too poor in hydrocarbons for quantitative estimations of the
radical yield. Therefore, we studied the photochemistry of 1b
in CS2.
In less viscous dodecane, the line broadening is not so
substantial and the spectrum detected after 5 min of heating 1a
very clearly demonstrates formation of the second radical with
three equivalent nitrogen atoms [Figure 2(a), 7 lines marked
by asterisks]. The g-factor and hyperfine coupling constant
(g = 2.0104, aN = 5.2 G) for radical 2a are almost identical to
those published for this radical (g = 2.0102, aN = 5.17 G).5(a)
The g-factor and hyperfine coupling constant obtained for the
second radical (g = 2.0060, aN = 3.96 G) coincide well with the
data reported for 3a† (g = 2.0059, aN = 3.97 G; Scheme 3).5(b)
Ar
3
N4
Ar
5
2 N
N6
3a Ar = Ph
3b Ar = 4-MeC6H4
S1
Scheme 3
Figure 4(a) shows the EPR spectra recorded upon the pro-
gressive photolysis of 1b in CS2. These spectra are very similar
Nitrogen hyperfine coupling constants for 2a (aN3 = aN5
=
= 6.1 G) and 3a (aN2 = aN6 = 4.8 G; aN4 = 4.9 G) predicted at
the UB3LYP/6-31G(d) level of theory are in a good agreement
with the experimental data. The calculated non-equivalence of
nitrogen atoms in radical 3a is not observed given the similarity
of their aN values, and hence yielding the experimental seven
line EPR pattern.5(b)
(a)
3
2
0
The lowest-field components of the EPR spectra of major
radicals 2 do not overlap with the components of minor radicals
3. Therefore, the kinetics of formation of radicals 2 could be
monitored by measuring the intensity of these components. The
relative concentrations of radicals 2 and 3 in the sample were
estimated by the spectral simulation [Figures 1(b) and 2(b)].
Figure 3 displays the typical kinetics of formation of radicals 2b
and 3b upon thermolysis of 1b. The formation of radical 2b is
well fitted by the exponential time dependence (Figure 3) yielding
the first-order-reaction rate constant k = (2.7 0.5)×10–4 s–1 at
1
3410
3420
3430
3440
H/G
(b)
0.4
0.2
0.0
*
*
*
*
*
(a)
(b)
*
*
0
40
80
120
Time/min
Figure 2 (a) Experimental EPR spectrum after heating 1a in dodecane
(c = 5×10–3 mol dm–3) for 5 min at T = 388 K. The resonance lines of 3a
are designated by asterisks. (b) Simulated spectrum assuming the presence
of radicals 2a (g = 2.0104, aN = 5.15 G) and 3a (g = 2.0060, aN = 3.96).
Figure 4 (a) EPR spectra recorded after (1) 5, (2) 40 and (3) 120 min
irradiation of 1b in CS2 (c = 2.1×10–3 mol dm–3) at ambient temperature by
436 nm light. (b) Kinetics of formation of radicals 2b (circles) and 3b
(asterisks).
– 205 –

Mendeleev Commun., 2007, 17, 204–206
2 (a) I. V. Vlasiuk, V. A. Bagryansky, N. P. Gritsan, Yu. N. Molin, A. Yu.
to those recorded upon thermolysis in dodecane (Figure 2) and
were simulated as a superposition of resonance lines from
radicals 2b and 3b, as discussed above. The same hyperfine
coupling constants (aN = 5.2 G and aN = 3.96 G) were extracted
for radicals 2b and 3b. Unlike the situation in squalane and
dodecane, the spectra of individual radicals were symmetrical
due to the very low viscosity of CS2. Peak-to-peak widths
(∆Hpp) for radicals 2b and 3b [Figure 4(a)] are equal to 0.168
and 0.052 mT, respectively [Figure 4(a)]. The difference in
∆Hpp is due to the much smaller hyperfine coupling constants
with protons of aryl substituents in 3b (~0.01–0.03 mT) com-
paring to 2b (0.03–0.07 mT). The aH values were taken from
UB3LYP/6-31G(d) calculations.
Makarov, Yu. V. Gatilov, V. V. Shcherbukhin and A. V. Zibarev, Phys.
Chem. Chem. Phys., 2001, 3, 409; (b) N. P. Gritsan, V. A. Bagryansky,
I. V. Vlasyuk, Yu. N. Molin, A. Yu. Makarov, M. S. Platz and A. V.
Zibarev, Izv. Akad. Nauk, Ser. Khim., 2001, 1973 (Russ. Chem. Bull., Int.
Ed., 2001, 50, 2064); (c) K. V. Shuvaev, V. A. Bagryansky, N. P. Gritsan,
A. Yu. Makarov, Yu. N. Molin and A. V. Zibarev, Mendeleev Commun.,
2003, 178; (d) N. P. Gritsan, S. N. Kim, A. Yu. Makarov, E. N. Chesnokov
and A. V. Zibarev, Photochem. Photobiol. Sci., 2006, 5, 95; (e) N. P.
Gritsan, E. A. Pritchina, T. Bally, A. Yu. Makarov and A. V. Zibarev,
J. Phys. Chem. A, 2007, 111, 817.
3 (a) T. Chivers, J. F. Richardson and N. R. M. Smith, Inorg. Chem., 1986,
25, 272; (b) R. T. Boere, J. Fait, K. Larsen and J. Yip, Inorg. Chem.,
1992, 31, 1417; (c) C. Knapp, E. Lork, T. Borrmann, W.-D. Stohrer and
R. Mews, Eur. J. Inorg. Chem., 2003, 3211; (d) C. Knapp, E. Lork, R. Mews
and A. V. Zibarev, Eur. J. Inorg. Chem., 2004, 2446; (e) C. Knapp,
E. Lork, R. Maggiulli, P. G. Watson, R. Mews, T. Borrmann, W.-D. Stohrer
and U. Behrens, Z. Anorg. Allg. Chem., 2004, 630, 1235; (f) C. Knapp
and R. Mews, Eur. J. Inorg. Chem., 2005, 3536.
Figure 4 shows the formation of radical 3b in the first
minutes of photolysis (its amount does not depend on further
irradiation, while the amount of radical 2b continues to grow).
Therefore, it is clear from kinetic data (Figures 3 and 4) that
radicals 3 are formed upon the thermolysis or photolysis of
corresponding minor impurities in 1. Typical (and practically
non-removable) impurities in 1a–e3 are correspondingly sub-
4 K. T. Bestari, R. T. Boere and R. T. Oakley, J. Am. Chem. Soc., 1989,
111, 1579.
5 (a) S. A. Fairhurst, K. M. Johnson, L. H. Sutcliffe, K. F. Preston, A. J.
Banister, Z. V. Hauptman and J. Passmore, J. Chem. Soc., Dalton Trans.,
1986, 1465; (b) P. J. Hayes, R. T. Oakley, A. W. Cordes and W. T.
Pennington, J. Am. Chem. Soc., 1985, 107, 1346.
stituted compounds 4,7 which can formally be transformed into
†
·
radicals 3 by loss of a SN fragment (Scheme 4). It should be
6 (a) A. Carrington and A. D. MacLachlan, Introduction to Magnetic
Resonance, Harper and Row, New York, Evanston, London, 1967;
(b) P. W. Atkins and D. Kivelson, J. Chem. Soc., 1966, 44, 169.
7 (a) A. D. Bond, D. A. Haynes and J. M. Rawson, Can. J. Chem., 2002, 80,
1507 and references therein; (b) I. Ernst, W. Holick, G. Rihs, D. Schomburg,
D. Shoham, D. Wenkert and R. B. Woodward, J. Am. Chem. Soc., 1981,
103, 1540.
8 C. Knapp, Doctoral Thesis, Universität Bremen, Bremen, Germany, 2002.
9 M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria, M. A.
Robb, J. R. Cheeseman, V. G. Zakrzewski, J. A. Montgomery, Jr.,
R. E. Stratmann, J. C. Burant, S. Dapprich, J. M. Millam, A. D. Daniels,
K. N. Kudin, M. C. Strain, O. Farkas, J. Tomasi, V. Barone, M. Cossi,
R. Cammi, B. Mennucci, C. Pomelli, C. Adamo, S. Clifford, J. Ochterski,
G. A. Petersson, P. Y. Ayala, Q. Cui, K. Morokuma, D. K. Malick,
A. D. Rabuck, K. Raghavachari, J. B. Foresman, J. Cioslowski, J. V.
Ortiz, B. B. Stefanov, G. Liu, A. Liashenko, P. Piskorz, I. Komaromi,
R. Gomperts, R. L. Martin, D. J. Fox, T. Keith, M. A. Al-Laham, C. Y.
Peng, A. Nanayakkara, C. Gonzalez, M. Challacombe, P. M. W. Gill,
B. Johnson, W. Chen, M. W. Wong, J. L. Andres, C. Gonzalez, M. Head-
Gordon, E. S. Replogle and J. A. Pople, Gaussian 98, Revision A.6,
Gaussian Inc., Pittsburgh, PA, 1998.
noted that radicals 3 are generated much faster than radicals 2
on both thermolysis and photolysis [Figures 3 and 4(b)].
S
Ar
N4
Ar
N
N
N
N
3
kT or hv
5
Ar
Ar
2 N
N6
– [SN]
S1
S
4
3
Scheme 4
Interestingly, corresponding cations 2a–e+ and 3a–e+ were
detected by mass spectrometry (EI, 70 eV) of 1a–e,3(c) and salts
of 3a+ and 3c+ were identified by XRD as by-products of the
reaction between 1a,c and [M(SO2)x][AsF6]2 (M = Co, Hg).8
To the best of our knowledge, however, the formation of
radicals 3 from precursors 4 has never been observed. This
reaction will be the topic of further research.
This work was supported by the Russian Foundation for
Basic Research (project nos. 04-03-32259 and 07-03-00467),
Deutsche Forschungsgemeinschaft (project no. 436 RUS 113/
486/0-3 R), the Siberian Branch of the Russian Academy of
Sciences (interdisciplinary project no. 25). C.K. acknowledges
the support of the BFK NaWi, University of Bremen.
References
1 (a) K. Awaga, T. Tanaka, T. Shirai, Y. Umezono and W. Fujita, C. R.
Chimie, 2007, 10, 52; (b) K. Awaga, T. Tanaka, T. Shirai, M. Fujimori,
Y. Suzuki, H. Yoshikawa and W. Fujita, Bull. Chem. Soc. Jpn., 2006, 79,
25; (c) J. M. Rawson, A. Alberola and A. Whalley, J. Mater. Chem.,
2006, 16, 2560; (d) J. M. Rawson and F. Palacio, Struct. Bond., 2001,
100, 93.
Received: 26th December 2006; Com. 06/2851
– 206 –