Reaction of (4-nitorophenyl)nitrene with molecular oxygen in low-temperature matrices: First IR detection and photochemistry of aryl nitroso oxide

Article abstract of DOI:10.1246/cl.2005.478

The thermal reaction of (4-nitrophenyl)nitrene with oxygen in low-temperature matrices produced a new molecular species, which was identified as 4-nitrophenylnitroso oxide on the basis of its IR spectrum in combination with isotopic labeling with ^18<

Full text of DOI:10.1246/cl.2005.478

478
Chemistry Letters Vol.34, No.4 (2005)
Reaction of (4-Nitorophenyl)nitrene with Molecular Oxygen in Low-temperature Matrices:
First IR Detection and Photochemistry of Aryl Nitroso Oxide
Hiroshi Inui,ꢀ Masatoshi Irisawa, and Shigero Oishi
Department of Chemistry, School of Science, Kitasato University, Kitasato, Sagamihara 228-8555
(Received December 22, 2004; CL-041585)
The thermal reaction of (4-nitrophenyl)nitrene with oxygen
(c)
in low-temperature matrices produced a new molecular species,
which was identified as 4-nitrophenylnitroso oxide on the basis
of its IR spectrum in combination with isotopic labeling with
18O2 and vibration analyses by the DFT method. This is the first
products
O3
report for the infrared detection of an aryl nitroso oxide. We
could monitor also the wavelength-dependent N–O and O–O
bond cleavage of the nitroso oxide.
(b)
products
(a)
X
N
N
N
O3
It has been assumed that the photooxidation of aryl azide in
solution afforded aryl nitroso oxide as an intermediate, which
was thought to convert to the corresponding nitro compound ei-
ther by unimolecular rearrangement via the dioxaziridine or by
intermolecular O-transfer.1 While a little attempt to observe
the intermediates involved was made by means of the laser flash
photolysis and the matrix isolation technique, these studies were
limited to the detection of UV–vis2 and ESR2b–d spectra. We
have monitored the photo-oxidation process by means of FT-
IR with using cryosystems to obtain more clear information.
We describe here definitive evidences for aryl nitroso oxide
and its photochemistry.
reactant
X
(d)
We used 4-nitrophenyl azide (1) instead of simple phenyl
azide as a precursor of aryl nitrene, because the latter azide
yielded not phenylnitrene but the didehydroazepine having the
seven membered ring, under our experimental conditions. In
the case of 1,3 only triplet (4-nitrophenyl)nitrene (2; 1557,
1497, 1338, 1321, 845, 833, and 743 cmꢁ1) could be accumulat-
ed in 5% O2-doped Xe matrices at 10 K under irradiation with
light of 365 nm from a 250-W deep-UV lamp with using a
band-pass filter. Although the photoisomerization of 2 into
5-nitro-1,2-didehydroazepine (7; 1887, 1568, 1511, 1331, and
970 cmꢁ1) was induced by the irradiation at 313 or >400 nm,
2 could be easily regenerated with light of 365 nm.
(e)
1700
1500
1300
Wavenumber /cm-1
1100
900
700
Figure 1. Difference IR spectra recorded after irradiation of X
(down bands) matrix-isolated in Xe doped with 16O2 (a) and
18O2 (b) at 10 K with light of >415 nm. In product bands, bands
labeled solid circles, N, and O3 assigned 5, 2, and ozone, respec-
tively. (b) shows 18O2-isotopic effects for the bands of X in the
region 900–1200 cmꢁ1. (c) IR spectrum of 4 matrix-isolated in
O2-doped Xe at 10 K. (d) and (e) UB3LYP/6-31+Gꢀ calculated
IR spectra for cis- and trans-3, respectively (scaled by 0.9614).6
Upon warming a Xe matrix containing 2 to 50 K in the ab-
sence of oxygen, small IR peaks remained, which suggested
the sole generation of dimerized product, namely, 4,40-dinitro-
azobenzene (871, 863 cmꢁ1). In a 5% O2-doped Xe matrix, how-
ever, at heating to 50 K, the bands due to 2 almost disappeared
and the bands from new molecular species4 (designated as X) ap-
peared, together with weak intensity peaks of the azobenzene.
When the same reaction was carried out with 18O2 (99% doubly
labeled) instead of 16O2, most of the IR bands of X shifted.4 In
caused the rapid disappearance of X and afforded 1,4-dinitroben-
zene (4; 1558, 1339, 875, 836, and 709 cmꢁ1) along with 4-nitro-
nitrosobenzene (5; 1545, 1518, 1414 1349, 1167, 1111, 806, and
745 cmꢁ1) and ozone (16O3; 1029, 18O3; 977 cmꢁ1). IR peaks of
these product species could be readily assigned by comparison
with the IR spectra of authentic samples isolated in O2-doped
Xe matrices at 10 K.5 From these experimental results, we
thought that the new species X might be the nitroso oxide.
In order to confirm this idea, the theoretical analyses for the
nitroso oxide were carried out using the DFT calculations at the
unrestricted B3LYP/6-31+Gꢀ level of theory. Geometry opti-
mization of the nitroso oxide revealed that two conformers cis-
and trans-3 existed as true energy minimum structures in their
singlet states and that cis-form is slightly more stable than
trans-form by 2.5 kJ molꢁ1. On the basis of the spin and charge
densities calculated for 3, both isomers are thought to have
particular, the medium intensity bands at 1131 and 1017 cmꢁ1
,
which appeared in the O–O stretching region of the peroxide,
showed the large isotopic shifts of ꢁ62 and ꢁ49 cmꢁ1, respec-
tively. In conjuncture with this, the photoreaction of X was also
examined. When the matrix containing X was irradiated with
light of >590 nm at 10 K, the precursor nitrene was regenerated
slowly and no other reaction could be observed. On the other
hand, as shown in Figure 1, irradiation of X with >415 nm
Copyright Ó 2005 The Chemical Society of Japan

Chemistry Letters Vol.34, No.4 (2005)
479
1.325
-0081
1.307
O
O2 ,50 K
O
365 nm
O
N
O
N
-0.400
(-0.221)
+
-0.319
(-0.306)
Ar N3
Ar
N
-0.084
-N2
Ar
cis-3
Ar
>590 nm / -O2
(0.071)
O
O
(-0.038)
O
N
O
N
triplet-2
1.288
1
1.283
trans-3
0.346
(-0.002)
Ar
Ar
0.281
365 nm
(0.001)
cis-3
O
N
O
313 or
>400 nm
>415 nm
N
trans-3
Ar
6
O2N
Figure 2. Spin densities, NPA atomic charges (in parentheses),
ꢀ
O (3P)
7
NO
+
NO2
and bond lengths (underlines, A) of 3 calculated at the
UB3LYP/6-31+Gꢀ level of theory. (Ar = 4-nitrophenyl).
Ar
Ar
5
4
O2
O3
diradical characters (Figure 2). Triplet-singlet energy differences
for cis- and trans-3 were calculated as 40.3 and 33.6 kJ molꢁ1
,
Ar =
O2N
respectively. These values indicate that the nitroso oxide popu-
late in their singlet states at 50 K. The vibration frequencies cal-
culated for cis- and trans-3 were compared with experimental
values after scaled by a factor of 0.9614 (Figure 1).6 IR spectrum
observed for X seems to be similar to that calculated for more
stable isomer cis-3. If the experimental spectrum consists of pure
cis-3, band at 1131 cmꢁ1 must be assigned to the C–N–O sym-
metrical stretching mode of the theoretically calculated spec-
trum. The isotopic shift for the C–N–O stretching was calculated
as ꢁ15 cmꢁ1, which is inconsistent with the large isotopic shift
(ꢁ62 cmꢁ1) observed. While, considering that X is a cis/trans
mixture of the nitroso oxide 3, the calculated IR data fairly re-
produce the observed data. Furthermore, theoretical isotopic
shifts for O–O stretching modes of cis-3 (ꢁ54 cmꢁ1) and
trans-3 (ꢁ62 cmꢁ1) are in fair agreement with experimental re-
sults described above. Thus it was concluded that the IR spec-
trum obtained by warming the matrix containing the nitrene 2
and oxygen molecules to 50 K corresponded to cis- and
trans-4-nitronitrosobenzene O-oxide (3). We believe that this
represents the first IR detection of aryl nitroso oxide, although
Laursen et al. observed trans-HNOO formed in the reaction of
NH and O2 in a Xe matrix.7
Although the formation of singlet 3 from triplet 2 and O2 is
spin-allowed and calculated as exothermic of 22.0 kJ molꢁ1, no 3
appeared in O2-doped Ar or Xe matrices at 40 K where O2 mole-
cules could be rapidly diffuse.8 At 50 K, however, 2 reacted with
O2 to give the nitroso oxide 3 as shown above, indicating that
this reaction needs some activation energy. Kinetic control
may explain the presence of trans-3 in spite of the energy differ-
ence (2.5 kJ molꢁ1) between trans-3 and more stable cis-3.
Next, the possible mechanism for the photochemical rear-
rangement of 3 into 4 is discussed. As mentioned in an introduc-
tory part, it was thought that the dioxaziridine may participate in
this type of reaction. Assuming that the dioxaziridine 6 exists as
an intermediate in this reaction process, the energy diagram cal-
culated with the DFT method shows that the energy minimum
structure of 6 (53.5 kJ molꢁ1 above cis-3) lies in a deep potential
well with activation barriers for the paths to cis-3 (109.7 kJ
molꢁ1) and to 4 (96.9 kJ molꢁ1). However, the theoretically cal-
culated bands for 6, which include the cyclic NO2 deformation
band at about 826 cmꢁ1 (18O-isotopic shiht; ꢁ22 cmꢁ1), could
not be detected in the photoreactions of 3, regardless of the ex-
citation at a variety of wavelengths (ꢀ > 590, > 540, > 460,
> 415, or 365 nm). Therefore, we concluded that 6 is not an in-
termediate in the rearrangement of 3 into 4 in the matrix at 10 K.
From the fact that 5 and ozone are observed in this photoreac-
tion, it is predicted that 3 generates an oxygen atom photochemi-
cally. The formation of 4 could be achieved by the attack of
Scheme 1.
atomic oxygen on the nitroso moiety of 5 in matrix cage, though
some of the atomic oxygen were captured by molecular oxygen
to give ozone (Scheme 1). This mechanism could be also sup-
ported by the following experiment. A Xe matrix containing 5
and a large amount of ozone was irradiated (>520 nm) at
10 K, as the photolysis of ozone was known to give atomic oxy-
gen (3P).9 It was observed that 5 was cleanly converted to 4 with-
out any by-products.
Thus, it was found that the photoreaction of 3 with the long-
wavelength light (>590 nm) caused the N–O bond cleavage to
yield 2, and that 4 was produced by the reaction of 5 with
O(3P), which are derived from the photochemical O–O bond
cleavage of 3 with light of >415 nm. At present, a systematic
study for the photoreactions of aryl nitroso oxides using a variety
of aryl azides is in progress.
This work was supported by Grant-in-Aid for Scientific
Research from the Ministry of Education, Culture, Sports,
Science and Technology Japan (No. 15750040) and by the
Sasakawa Scientific Research Grant from The Japan Science
Society (No. 15-107).
References and Notes
1
a) K. Ishiguro and Y. Sawaki, Bull. Chem. Soc. Jpn., 73, 535 (2000).
b) S. Ishikawa, T. Nojima, and Y. Sawaki, J. Chem. Soc., Perkin
Trans. 2, 1996, 127. c) Y. Sawaki and S. Ishikawa, J. Am. Chem.
Soc., 109, 584 (1987). d) T.-Y. Liang and G. B. Schuster, J. Am.
Chem. Soc., 109, 7803 (1987).
2
a) T. Harder, P. Wessig, J. Bendig, and R. Stosser, J. Am. Chem. Soc.,
¨
121, 6580 (1999). b) N. P. Gritsan and E. S. Pritchina, J. Inf. Rec.
Mater., 17, 391 (1989). c) E. A. Pritchina and N. P. Gritsan, J.
Photochem. Photobiol., A, 43, 165 (1988). d) J. S. Brinen and B.
Singh, J. Am. Chem. Soc., 93, 6623 (1971).
IR bands of 1 (5% O2-doped Xe, 10 K) 2133m, 2123s, 2102m, 2090m,
1608m, 1596m, 1593m, 1528s, 1492m, 1341m, 1304s, 1289s, 1176w,
3
4
5
1129w, 1105w, 860w, 847w, 748w cmꢁ1
.
IR bands of X (18O-isotopic shifts in cmꢁ1) 1602m (ꢁ1), 1571w (0)
1534s (ꢁ1), 1358m (ꢁ3), 1335s (ꢁ1), 1215w (ꢁ1), 1131m (ꢁ62),
1108w (ꢁ6), 1017m (ꢁ49), 856s (ꢁ1), 844s (ꢁ1), 751m (0) cmꢁ1
.
5 was synthesized by the published method.10 Ozone was prepared
with help of an ozonizer (Willbe ozonizer OZ-2, Ozone Shi-Nine inc.)
A. P. Scott and L. Radom, J. Phys. Chem., 100, 16502 (1996).
S. L. Laursen, J. E. Grace, R. L. DeKock, and S. A. Spronk, J. Am.
Chem. Soc., 120, 12583 (1998).
6
7
8
9
W. W. Sander, J. Org. Chem., 53, 2091 (1988).
J. K. Parker and S. R. Davis, J. Am. Chem. Soc., 121, 4271 (1999).
10 I. D. Entwistle, T. Gilkerson, R. A. W. Johnstone, and R. P. Telford,
Tetrahedron, 34, 213 (1978).
Published on the web (Advance View) February 26, 2005; DOI 10.1246/cl.2005.478