UV-laser photochemistry of isoxazole isolated in a low-temperature matrix

Article abstract of DOI:10.1021/jo301699z

The photochemistry of matrix-isolated isoxazole, induced by narrowband tunable UV-light, was investigated by infrared spectroscopy, with the aid of MP2/6-311++G(d,p) calculations. The isoxazole photoreaction starts to occur upon irradiation at λ = 240 nm, with the dominant pathway involving decomposition to ketene and hydrogen cyanide. However, upon irradiation at λ = 221 nm, in addition to this decomposition, isoxazole was also found to isomerize into several products: 2-formyl-2H-azirine, 3-formylketenimine, 3-hydroxypropenenitrile, imidoylketene, and 3-oxopropanenitrile. The structural and spectroscopic assignment of the different photoisomerization products was achieved by additional irradiation of the λ = 221 nm photolyzed matrix, using UV-light with λ ≥ 240 nm: (i) irradiation in the 330 ≥ λ ≥ 340 nm range induced direct transformation of 2-formyl-2H-azirine into 3-formylketenimine; (ii) irradiation with 310 ≥ λ ≥ 318 nm light induced the hitherto unobserved transformation of 3-formylketenimine into 3-hydroxypropenenitrile and imidoylketene; (iii) irradiation with λ = 280 nm light permits direct identification of 3-oxopropanenitrile; (iv) under λ = 240 nm irradiation, tautomerization of 3-hydroxypropenenitrile to 3-oxopropanenitrile is observed. On the basis of these findings, a detailed mechanistic proposal for isoxazole photochemistry is presented.

Full text of DOI:10.1021/jo301699z

Article
pubs.acs.org/joc
UV-Laser Photochemistry of Isoxazole Isolated in a Low-Temperature
Matrix
Claudio M. Nunes,* Igor Reva,* Teresa M. V. D. Pinho e Melo, and Rui Fausto
́
Department of Chemistry, University of Coimbra, P-3004-535 Coimbra, Portugal.
S
* Supporting Information
ABSTRACT: The photochemistry of matrix-isolated isoxa-
zole, induced by narrowband tunable UV-light, was inves-
tigated by infrared spectroscopy, with the aid of MP2/6-311+
+G(d,p) calculations. The isoxazole photoreaction starts to
occur upon irradiation at λ = 240 nm, with the dominant
pathway involving decomposition to ketene and hydrogen
cyanide. However, upon irradiation at λ = 221 nm, in addition
to this decomposition, isoxazole was also found to isomerize
into several products: 2-formyl-2H-azirine, 3-formylketen-
imine, 3-hydroxypropenenitrile, imidoylketene, and 3-oxopro-
panenitrile. The structural and spectroscopic assignment of the
different photoisomerization products was achieved by additional irradiation of the λ = 221 nm photolyzed matrix, using UV-light
with λ ≥ 240 nm: (i) irradiation in the 330 ≤ λ ≤ 340 nm range induced direct transformation of 2-formyl-2H-azirine into
3‑formylketenimine; (ii) irradiation with 310 ≤ λ ≤ 318 nm light induced the hitherto unobserved transformation of
3‑formylketenimine into 3-hydroxypropenenitrile and imidoylketene; (iii) irradiation with λ = 280 nm light permits direct
identification of 3-oxopropanenitrile; (iv) under λ = 240 nm irradiation, tautomerization of 3-hydroxypropenenitrile to
3‑oxopropanenitrile is observed. On the basis of these findings, a detailed mechanistic proposal for isoxazole photochemistry is
presented.
photoisomerization reaction to 2,5-diphenyloxazole C. At λ =
245 nm the reaction occurs through an initial ring contraction
to yield the isolable 2-benzoyl-3-phenyl-2H-azirine B, which
then undergoes ring expansion to C. On the other hand,
irradiation of 2H-azirine B with longer wavelength (λ > 300
nm) leads to photoreversion to isoxazole A.
The observed wavelength dependence was interpreted as
selective excitation of one of two chromophores in 2H-azirine
B.13−15 The n,π* excitation of the carbonyl chromophore that
occurs at longer wavelengths should lead to the cleavage of the
C−N bond, and formation of a diradical intermediate
(vinylnitrene type), which recloses to isoxazole A. In contrast,
the n,π* excitation of the imine chromophore, promoted at
shorter wavelengths, should lead to the cleavage of the C−C
bond and formation of a zwitterionic intermediate (nitrile ylide
type), which recloses to oxazole C.13−15
Studies on the photoisomerization of isoxazole A (and its
aryl derivatives), using sensitizers, have not fully elucidated if
the reported reaction involves triplet or singlet excited
states.13−16 Nevertheless, calculations on the potential energy
surface (PES) of the parent isoxazole, at the MO-CI (STO-3G)
ab initio level, indicate that the rearrangement to 2H-azirine
occurs in a concerted manner, through the lowest singlet
excited state (S1) of isoxazole.17 The reverse reaction was
1. INTRODUCTION
Isoxazoles are an important class of five-membered unsaturated
heterocyclic compounds. They show several applications in
diverse areas such as pharmaceuticals, agrochemistry, and
industry.1−5 Isoxazoles are also found in natural sources
showing insecticidal, plant growth regulation, and pigment
functions.1−3
A structural feature that distinguishes isoxazoles from other
heterocycles is the fact that its ring has properties of an
aromatic system and, at the same time, has a weak N−O bond
that plays a central role in their chemistry.1−3 Indeed, under
thermal or photochemical conditions, or even during the
metabolism of several isoxazole-containing drugs, the N−O
bond cleavage is described as the first step for their
transformation.1−12
The earliest photochemical studies of isoxazoles were
reported for 3,5-diphenylisoxazole A by Ullman and Singh in
1966 (Scheme 1).13 This molecule shows an interesting
Scheme 1. Photoisomerization of 3,5-Diphenylisoxazole A to
2,5-Diphenyloxazole C, via the 2-Benzoyl-3-phenyl-2H-
azirine B Intermediate13
Received: August 9, 2012
Published: September 6, 2012
© 2012 American Chemical Society
8723
dx.doi.org/10.1021/jo301699z | J. Org. Chem. 2012, 77, 8723−8732

The Journal of Organic Chemistry
Article
Figure 1. (a) Experimental IR spectrum of isoxazole 1 isolated in an argon matrix at 15 K. (b) Experimental difference spectrum; the spectrum
obtained after UV-laser irradiation at λ = 238 nm (70 min) of isoxazole 1 in argon matrix “minus” the spectrum of the sample before irradiation (as
●
○
deposited). (c) Simulated IR spectra of ketene 2 ( ) and hydrogen cyanide 3 ( ) at the MP2/6-311++G(d,p) level, in ratio 1:1.
suggested to occur through the T1 triplet state, upon
intersystem crossing from the S1 state of 2H-azirine. On the
other hand, upon excitation of 2H-azirine to the S2 singlet
excited state, C−C bond cleavage should occur, leading to
oxazole formation.17
photochemistry of the studied compound. Here we reported
previously unobserved UV-induced isoxazole ring-cleavage
reactions and isomerizations, including various intramolecular
hydrogen-atom shifts. The characterization of the photo-
products generated in these reactions allowed us to shed
light, for the first time, on the complex photochemistry of the
parent isoxazole.
Further photochemical studies on several different sub-
stituted isoxazoles have established that isomerization to
2H‑azirines, and then to oxazoles, is the most common
pathway.18−32 However, depending on the ring substitution
pattern, or even on the solvent used, in some cases competitive
reactions to nitriles or other species have also been
observed.22−33 In spite of these investigations, little is still
known about the intermediates involved in the photochemical
reactions of isoxazoles. Furthermore, no experimental results
concerning the photochemical behavior of the simplest member
of this family have been reported hitherto. Indeed, the
photochemistry mechanism of isoxazoles is still far from
being understood.
In this paper, we present the study of the photochemistry of
the parent isoxazole using low-temperature matrix-isolation and
infrared spectroscopy, complemented by quantum chemical
calculations. Narrowband tunable UV-irradiation of the matrix-
isolated isoxazole was undertaken in situ by means of an optical
parametric oscillator (OPO) pumped by a pulsed Nd:YAG
laser. The advantages provided by the cryogenic matrix
environment, specifically the low temperature, inertness of
the media, inhibition of molecular diffusion from the matrix
cage, and high spectral resolution, make the used approach a
powerful strategy for investigating the reaction intermediates
and characterizing the reaction mechanisms involved in the
2. RESULTS AND DISCUSSION
Photochemistry of Isoxazole at λ = 238 nm. The
photochemistry of monomeric isoxazole 1 isolated in an argon
matrix was investigated using narrowband laser irradiation. A
longer UV wavelength was first selected (300 nm) and then
gradually decreased until changes in the sample, monitored by
infrared spectroscopy, were observed. It was noticed that 1
starts to react, very slowly, upon UV-irradiation with λ = 240
nm, whereas upon irradiation at λ = 238 nm the same reaction
goes slightly faster. Irradiations at 238 nm for 70 min consumed
∼10% of 1, and simultaneously, several new bands due to
photoproducts appeared in the IR spectrum (Figure 1).
The new emerging band at 2137 cm−1 was found to be at
least 4 times more intense than the most intense band of 1
consumed. The frequency and high infrared intensity of this
band allowed its easy assignment to the characteristic
ν(CCO)as vibration of a ketene moiety.34 In fact,
observation of this very characteristic band, together with the
bands appearing at 3056 and 1376 cm−1, allows unambiguous
conclusion that this product is the parent ketene 2
[H2CCO]. In addition, hydrogen cyanide 3 [HCN],
which is expected to be formed complementary to ketene 2 if
8724
dx.doi.org/10.1021/jo301699z | J. Org. Chem. 2012, 77, 8723−8732

The Journal of Organic Chemistry
Article
photodecomposition of isoxazole 1 occurs (Scheme 2), is
identified by the observed bands at 3300 and 727/725 cm−1
regions, are due to the formation of several photoisomerization
products labeled with numbers 4 to 8 (Figure 2b). As will be
shown below, their identification was achieved with the aid of
subsequent irradiation experiments using longer wavelengths
(λ ≥ 240 nm), i.e., under conditions where the parent isoxazole
is photostable.
Scheme 2. Dominant Photochemical Reaction of Isoxazole 1
Isolated in an Argon Matrix upon UV-Laser Irradiation at
λ = 238 nm
Identification of Isoxazole Photoisomerization Pro-
ducts Generated at λ = 221 nm. The UV-laser irradiation at
λ = 221 nm was applied to isoxazole 1 isolated in an argon
matrix until a maximum yield of photoisomerization products
was achieved (which corresponded to ∼60% consumption of
1).41 Subsequently, the matrix was further irradiated using UV-
light within the 340−240 nm region, starting with longer
wavelengths, which were then gradually decreased while
changes were monitored by FTIR spectroscopy. These
irradiation experiments allowed the observation of individual
reactions of the products 4 to 8, while the isoxazole 1 remained
intact. The results presented below correspond to the most
illustrative and spectacular observed changes, which were
selected to demonstrate consumption of different photo-
isomerization products and permit their unequivocal assign-
ment. The actual number of the irradiations performed, in the
course of the reported experiments, was quite large to
guarantee that as many as possible different photoreactions,
occurring upon excitation at different UV-thresholds, are
characterized (see SI, Table S2).
(Figure 1). These bands correspond to the ν(C−H) and
δ(HCN) vibrations of HCN, respectively (see Supporting
Information (SI), Table S1).
When the current IR data are compared with the data
reported for matrix isolated monomeric 235−37 and 3,38,39
a
shift of 5 cm−1 in frequencies is observed (Table S1, SI),
suggesting that ketene 2 and hydrogen cyanide 3, produced
from isoxazole 1, form a complex in the same matrix cage. This
hypothesis is also supported by the excellent agreement
between the present results and those corresponding to the
reported UV photochemistry of acetyl cyanide [CH3COCN] in
an argon matrix, where formation of complexes of 2 and 3 was
described (Table S1, SI).40
Photochemistry of Isoxazole at λ = 221 nm. When
isoxazole 1 isolated in an argon matrix was irradiated at shorter
wavelengths, different results were obtained. Upon irradiation
at λ = 221 nm, the shortest wavelength available in our (UV-
laser + OPO) system, many additional photoproduct bands,
besides those corresponding to 2 and 3, were observed (Figure
2). In these conditions, after one minute of irradiation more
than 10% of 1 was consumed. The most intense new bands,
which emerged in the 2100−2000 cm−1 and 1800−1600 cm−1
Irradiation Experiments in the 330 ≤ λ ≤ 340 nm Range.
The irradiation within the 330 ≤ λ ≤ 340 nm range,42 of the
photoisomerization products of isoxazole 1 isolated in an argon
matrix, led to the production of 3-formylketenimine 5 and
consumption of the isomeric 2-formyl-2H-azirine 4 (Figure 3
and Scheme 3).43 The bands that increased in the IR spectrum,
observed at 2042, 1703 (also labeled with 5 in Figure 2b), and
Figure 2. (a) Experimental difference IR spectrum; the spectrum obtained after UV-laser irradiation at λ = 221 nm (1 min) of isoxazole 1 in argon
matrix “minus” the spectrum of the sample before irradiation (as deposited). Positive labeled bands are due to ketene 2 and hydrogen cyanide 3. (b)
Expanded view, corresponding to the dashed rectangle in frame (a), showing only the 2100−2000 cm−1 and 1800−1600 cm−1 regions, where the
most intense bands due to photoisomerization products (labels 4 to 8) appear (see text).
8725
dx.doi.org/10.1021/jo301699z | J. Org. Chem. 2012, 77, 8723−8732

The Journal of Organic Chemistry
Article
Figure 3. (a) Experimental difference IR spectrum; the spectrum obtained after UV-laser irradiation of photoisomerization products of isoxazole 1 in
an argon matrix, in the 330 ≤ λ ≤ 340 nm range (10 min), “minus” the spectrum before this irradiation, corresponding to the photoisomerization
products of isoxazole 1 produced at λ = 221 nm (see text). (b) Simulated difference IR spectrum at the MP2/6-311++G(d,p) level considering the
production of ketenimine anti-5 (positive bands) and the consumption of 2H-azirine anti-4 (negative bands).
Figure 2b, the lower-frequency band labeled with 4, observed at
1728 cm−1, is assigned to the syn-4 conformer (the higher-
frequency band labeled with 4, at 1735 cm−1, is due to anti-4, as
previously indicated); additional data are presented in Table S4
(SI). In the case of the syn-5 conformer, its identification is
mainly established based on the minor observed bands, labeled
with 5 in Figure 2b, that appear at 2047 cm−1 [ν(CCN)as]
and 1680/1675 cm−1 [ν(CO)] (see also Table S3, SI).
Irradiation Experiments in the 310 ≤ λ ≤ 318 nm Range.
After the consumption of 2-formyl-2H-azirine 4, further
irradiation conducted in the 318−310 nm range led to the
consumption of 3-formylketenimine 5 concomitantly with the
increase of several bands (Figure 4a). From the analysis of the
spectral changes, 3-hydroxypropenenitrile syn-(Z)-6 (the most
stable of four conformers, Figure S4, SI) could be easily
identified. For example, the bands at 3538, 2221, and 1627
cm−1 (also labeled with 6 in Figure 2b) correlate well with the
reported IR data for this species,9 namely, its ν(O−H),
ν(C≡N), and ν(CC) characteristic bands, respectively (see
Table S5, SI).
Furthermore, the formation of the imidoylketene anti-(E)-7
(the most stable of four conformers, Figure S5, SI) could also
be unambiguously established. The band observed at 2136/
2134 cm−1 appears in the characteristic region of the
ν(CCO)as vibration of imidoylketenes isolated in the
low-temperature matrix.44−53 In addition, the bands observed at
1612/1610 cm−1 (also labeled with 7 in Figure 2b) and 1020
cm−1 show an excellent correlation with the most intense bands
estimated at the MP2/6-311++G(d,p) level for anti-(E)-7,
namely, those predicted at 1612 cm−1 [ν(NC)] and 1022
cm−1 [δ(HNC) + ν(C−C)].
Scheme 3. Identification of 2-Formyl-2H-azirine 4 and
3‑Formylketenimine 5 as Photoisomerization Products of
Isoxazole 1 (λ = 221 nm), by Means of Subsequent
Irradiation Experiments in the 330 ≤ λ ≤ 340 nm Range
880 cm−1, are reliably assigned to ketenimine 5, namely, to its
characteristic ν(CCN)as, ν(CO), and δ(HNC) vibra-
tional modes, respectively (see also Table S3, SI). At the same
time, the bands that decreased in the spectrum, observed at
2824, 1735 (also labeled with 4 in Figure 2b), 1673, and 1186/
1182 cm−1, are ascribed to 2H-azirine 4, more specifically to its
ν(CHO), ν(CO), ν(CN)ring, and ν(C−C)ring vibrational
modes, respectively (see also Table S4, SI). These conclusions
are unequivocally established by the excellent correlation
between the experimental and the MP2/6-311++G(d,p)
calculated difference IR spectra (Figure 3).
Calculations on the PES of 4 and 5 show that both molecules
have two different conformations, syn and anti, with the anti
forms being more stable by 4−5 kJ mol−1 (Figures S2 and S3,
SI). As presented in Figure 3, the irradiation in the 330 ≤ λ ≤
340 nm range led mainly to the photoreaction of the anti-4
form to give the anti-5 form. However, the syn forms of
2H‑azirine 4 and ketenimine 5 could also be identified as
photoisomerization products of isoxazole 1 at λ = 221 nm. In
The MP2/6-311++G(d,p) calculated difference IR spectrum,
considering the consumption of syn-5 and anti-5 (0.1:1) and
formation of syn-(Z)-6 and anti-(E)-7 (0.6:0.5) (Figure 4b),
shows an excellent agreement with the experimental difference
8726
dx.doi.org/10.1021/jo301699z | J. Org. Chem. 2012, 77, 8723−8732

The Journal of Organic Chemistry
Article
Figure 4. (a) Experimental difference IR spectrum; the spectrum obtained after UV-laser irradiation, in the 310 ≤ λ ≤ 318 nm range (25 min),
“minus” the spectrum previous to this laser irradiation, after the consumption of 4 (see text). (b) Calculated IR spectra at the MP2/6-311++G(d,p)
□
■
level considering the production of hydroxynitrile syn-(Z)-6 (red line, ) and imidoylketene anti-(E)-7 (blue line, ) (0.6:0.5) and the consumption
of ketenimine syn-5 and anti-5 (0.1:1) (green line, negative bands).
IR spectrum (Figure 4a), strongly supporting the structural
assignments (Scheme 4).
The two most intense bands in the experimental spectrum
observed at 1760/1758/1755 cm−1 (also labeled with 8 in
Figure 2b) and at 1020/1015 cm−1 are in good agreement with
those estimated at 1729 cm−1 [ν(CO)] and at 1032 cm−1
[ν(OC−C)] for anti-8.54 In addition, the two bands observed
at 1400 and 1384 cm−1 could be assigned to the δ(CH2) and
δ(OCH) bending modes of anti-8, calculated at 1424 and 1394
cm−1, respectively.
Simultaneously with the consumption of the ketonitrile 8,
the bands at 2292 and at 2257 cm−1 were observed to increase.
These bands are characteristic of acetonitrile.55−57 Although the
appearance of the band due to CO (∼2138 cm−1)58,59 is
difficult to establish with certainty, due to the overlap with the
much more intense bands of ketenes 2 and 7, the formation of
CO together with acetonitrile could indicate a potential
photodecomposition channel for 8.
Scheme 4. Identification of 3-Hydroxypropenenitrile syn-
(Z)-6 and Imidoylketene anti-(E)-7 as Photoisomerization
Products of Isoxazole 1 (λ = 221 nm), by Means of
Subsequent Irradiation Experiments in the 310 ≤ λ ≤ 318
nm Range, That Lead to Their Formation from the
Consumption of 3-Formylketenimine 5
After consumption of ketonitrile 8 (and also imidoylketene 7,
that disappear during subsequent irradiations with 280−250
nm; see also SI, Table S2), irradiation experiments in the 248−
244 nm region led to the Z → E photoisomerization of the
hydroxynitrile 6 (Figure 6). This result is in accordance with
reported results from previous experiments performed under
similar conditions (see Table S5, SI).9 On the basis of these
data, the bands labeled with 6 in Figure 2b, observed at 1673,
1647, and 1627 cm−1, are reliably assigned to the ν(CC) of
anti-(E)-6, syn-(E)-6, and syn-(Z)-6, respectively (the hydrox-
ynitrile syn-(Z)-6 was also identified during irradiation
experiments in the 310 ≤ λ ≤ 318 nm range, as already
mentioned).
Irradiation Experiments in the 240 ≤ λ ≤ 280 nm Range.
After consumption of 2H-azirine 4 and ketenimine 5 in the
course of 300 ≤ λ ≤ 340 nm irradiation, it was observed that
further irradiation with λ = 280 nm led to the reaction of
another species, which could be identified as 3-oxopropaneni-
trile 8. Although other transformations seem also to occur,
under these conditions the main process was consumption of
the anti-8 conformer (the most stable form of 8, Figure S6, SI).
This is clearly supported by the good agreement between the
experimental difference IR spectrum and the MP2/6-311+
+G(d,p) calculated data (Figure 5).
Finally, irradiation experiments with λ = 240 nm (under
these conditions isoxazole 1 reacts negligibly slow) led to the
consumption of different conformers of hydroxynitrile 6 and to
formation of ketonitrile 8 (Figure 7). The phototautomeriza-
tion 6 → 8 is undoubtedly identified by considering, for
8727
dx.doi.org/10.1021/jo301699z | J. Org. Chem. 2012, 77, 8723−8732

The Journal of Organic Chemistry
Article
Figure 5. (a) Experimental difference IR spectrum; the spectrum obtained after UV-laser irradiation at λ = 280 nm (30 min) “minus” the spectrum
□
▲
previous to this laser irradiation, after the consumption of 4 and 5 (see text). Bands labeled with ( ) and ( ) are due to syn-(E)-6 and syn-(Z)-6,
■
and that labeled with ( ) is due to anti-(E)-7. (b) Simulated difference IR spectrum at the MP2/6-311++G(d,p) level considering the consumption
of ketonitrile anti-8.
Figure 6. (a) Experimental difference IR spectrum; the spectrum obtained after UV-laser irradiation, in the 244 ≤ λ ≤ 248 nm range (11 min),
“minus” the spectrum previous to this laser irradiation, after the decomposition of imidoylketene 7 and ketonitrile 8 (see text). (b) Simulated
□
difference IR spectrum at the MP2/6-311++G(d,p) level considering the photoisomerization of hydroxynitrile syn-(Z)-6 ( ) (negative bands) to
▲
△
syn-(E)-6 ( ) and anti-(E)-6 ( ) (positive bands) in a ratio of 0.85:0.15.
example, the decrease of the bands at 1627, 1647, and 1673
cm−1 (due to the ν(CC) of syn-(Z)-6, syn-(E)-6, and anti-
(E)-6, respectively) and the concomitant increase of the band
at 1760/1758/1755 cm−1, corresponding to ν(CO) of anti-8
(identified also during irradiation experiments at λ = 280 nm
and labeled with 8 in Figure 2b). The summary of the
irradiation experiments in the 240 ≤ λ ≤ 280 nm range is
presented in Scheme 5.
Mechanism of the Photochemistry of Isoxazole. The
summary of the experimental results together with a
mechanistic proposal for the photochemistry of parent
isoxazole 1 are presented in Scheme 6. Irradiations within the
238 ≤ λ ≤ 240 nm range (the longest wavelengths that induce
reaction of 1) led mainly to the hitherto unobserved
photodecomposition of 1 to ketene 2 and hydrogen cyanide
3. One may suppose that this reaction occurs via an isoxazole
carbene tautomer, formed by a [1,2] sigmatropic H-shift from
8728
dx.doi.org/10.1021/jo301699z | J. Org. Chem. 2012, 77, 8723−8732

The Journal of Organic Chemistry
Article
Figure 7. Experimental difference IR spectrum showing the phototautomerization 6 → 8; the spectrum obtained after UV-laser irradiation at λ = 240
nm (6 min) “minus” the spectrum previous to this laser irradiation, after the Z → E photoisomerization of hydroxynitrile 6 (see text).
hydroxynitrile 6, imidoylketene 7, and ketonitrile 8, were also
observed. Contrary to what could be expected, no photo-
isomerization to oxazole 9 was detected.60 Since the bands due
to 2H-azirine 4 and ketenimine 5 reach their maxima in the first
minutes (while the bands of the other photoisomerization
products reach their maxima somewhat later, Figure S1, SI), it
can be suggested that 4 and 5 are primary isomerization
products in the isoxazole 1 photochemistry.
Scheme 5. Identification of 3-Hydroxypropenenitrile syn-
(Z)-6, syn-(E)-6, and anti-(E)-6, and 3-Oxopropanenitrile
anti-8 as Photoisomerization Products of Isoxazole 1
(λ = 221 nm), by Means of Subsequent Irradiation
Experiments in the 240 ≤ λ ≤ 280 nm Range
Recently, we reported that vinylnitrene I, which results from
the cleavage of the N−O bond, is a very reactive minimum on
the PES of isoxazole 1.9 Although its existence remains yet to
be experimentally proven, this species is probably an
intermediate in the photochemistry of 1 that gives rise to the
formation of 2H-azirine 4 and ketenimine 5. Under this
hypothesis, it is likely that the open-shell singlet vinylnitrene I,
better described as a 1,3-diradical, collapses easily once formed
to give the 2H-azirine 4 (as reported in ref 9; this is the lowest-
energy deactivation path and requires only the movement of
the nitrogen atom out of the molecular plane and the closure of
the N−C−C angle).9,61−63 On the other hand, it can
conceivably be hypothesized that the ground-state triplet
vinylnitrene I, in which the amount of delocalization of the
nonbonding π electron from nitrogen is expected to be
modest,9,61−63 is responsible for the formation of ketenimine 5
through a [1,2] sigmatropic H-shift reaction (vide infra; in
addition it should be noted that triplet ground states of
arylnitrenes are known to be reactive as hydrogen-atom
abstracting agents64−66).
C5 to C4, which could then decompose through the cleavage of
the weakest N−O and C4−C3 bonds.9 However, since no
intermediates were detected during the experiments in argon
matrices at 15 K, the establishment of the definitive mechanism
for this new photodecomposition reaction still requires further
investigation.
When the irradiation of isoxazole 1 was performed with
shorter wavelengths, namely, at λ = 221 nm, the new
photoisomerization products, 2H-azirine 4, ketenimine 5,
Scheme 6. Summary of Experimental Observations and the Mechanistic Proposal for the UV-Induced Photochemistry Pathways
a
of Isoxazole 1 Isolated in a Low-Temperature Argon Matrix
a
Structures given in blue possess different conformers (for simplicity not shown).
8729
dx.doi.org/10.1021/jo301699z | J. Org. Chem. 2012, 77, 8723−8732

The Journal of Organic Chemistry
Article
In the irradiation experiments of photoisomerization
products of isoxazole 1, it was observed that 2-formyl-2H-
azirine 4 is converted into 3-formylketenime 5 when irradiated
in the 330 ≤ λ ≤ 340 nm range. Some examples of the unusual
photochemical cleavage of the C−N bond in 2H-azirines, to
give rise to ketenimines, have been reported.67−71 This reaction
is mainly observed when longer irradiation wavelengths are
used, but even in these apparently exceptional cases the
common C−C bond cleavage also occurs (at least with lower
irradiation wavelengths).67−71 On the basis of different
experimental data, Inui and Murata proposed that the cleavage
of the C−N bond of 2H-azirines involves the formation of a
vibrationally excited triplet vinylnitrene, which then undergoes
Curtius-like rearrangement to the corresponding keteni-
mines.69,72 These results are consistent with our hypothesis,
where it is suggested that triplet vinylnitrene I, formed from
isoxazole 1 via N−O bond cleavage (or from 2H-azirine 4), is
an intermediate in formation of ketenimine 5.
In some particular cases, the photoisomerization of isoxazoles
into ketonitriles has been previously reported.32 The results
from the current study indicate that this reaction should take
place preferentially with 5-unsubstituted isoxazoles, since the
[1,5] sigmatropic H-shift from ketenimine to hydroxynitrile is
not blocked by substituents, and the subsequent photo-
tautomerization to ketonitriles can occur. In fact, the detailed
mechanism for a ketonitrile formation upon irradiation of
isoxazole was established in this study for the first time.
3. CONCLUSION
Use of matrix isolation infrared spectroscopy and in situ
narrowband tunable UV irradiation of matrix-isolated parent
isoxazole allowed us for the first time to establish a detailed
picture of the unimolecular photochemistry of this compound.
The experimental studies received support from MP2/6-311+
+G(d,p) theoretical calculations of the structure and spectra of
the relevant species.
The nonexistence of photoisomerization to oxazole 9
indicates that during irradiation 2H-azirine 4 does not react
via C−C bond cleavage. Only the C−N bond cleavage of 4
should exclusively occur, even when lower wavelengths are
used, which made this derivative a special case in the
photochemistry of 2H-azirines (results that surely deserve
future investigations).
Isoxazole was found to be photostable upon UV irradiation
with λ > 240 nm. Upon irradiation of the matrix-isolated
compound at λ = 240−238 nm, the dominant observed
photoreaction was isoxazole decomposition to ketene and
hydrogen cyanide. On the other hand, irradiation at λ = 221 nm
was found to induce additionally a series of isomerization
processes that led to formation of several products, which could
be identified by additional photochemical experiments where
the λ = 221 nm photolyzed isoxazole matrix (containing a
mixture of isoxazole and its photoproducts) was subjected to
irradiations with λ ≥ 240 nm. Irradiation in the 330 ≤ λ ≤ 340
nm range induced direct transformation of 2-formyl-2H-azirine
into 3-formylketenimine; irradiations with 310 ≤ λ ≤ 318 nm
induced transformation of 3-formylketenimine into 3-hydrox-
ypropenenitrile and imidoylketene; and irradiations with
λ = 280 nm and λ = 240 nm permitted identification of
3‑oxopropanenitrile, in the last case at the expense of
3‑hydroxypropenenitrile tautomerization.
Several important mechanistic considerations were extracted
from the observations: (i) photodecomposition of isoxazole to
ketene and hydrogen cyanide seems to occur via an isoxazole
carbene tautomer, formed by a [1,2] sigmatropic H-shift from
isoxazole C5 to C4, and subsequent cleavage of the weakest
N−O and C4−C3 bonds; (ii) the 2-formyl-2H-azirine and
3‑formylketenime are primary isomerization products in the
isoxazole photochemistry, both species presumably resulting
from rearrangements of an initially formed vinylnitrene in its
excited open-shell singlet and ground triplet states, respectively;
(iii) the vibrationally excited triplet vinylnitrene seems also be
involved in the observed conversion of 2-formyl-2H-azirine into
3-formylketenime, via Curtius-like rearrangement, upon irradi-
ation in the 330 ≤ λ ≤ 340 nm range; (iv) on the other hand,
the nonobservation of oxazole indicates that under these
conditions the 2H-azirine does not react via C−C bond
cleavage; (v) photoisomerization of 3-formylketenimine to
hydroxynitrile and imidoylketene was observed for the first
time; and finally, (vi) observation of phototautomerization of
the hydroxynitrile to the corresponding ketonitrile, upon
irradiation at λ = 240 nm, suggests that this reaction should
take place preferentially with 5-unsubstituted isoxazoles, where
the required [1,5] sigmatropic H-shift from ketenimine to
hydroxynitrile is not blocked and the subsequent photo-
tautomerization to ketonitriles can occur.
Subsequently, irradiation experiments with 310 ≤ λ ≤ 318
nm led to the interesting photoisomerization of 3-formylkete-
nimine 5 to hydroxynitrile 6 and imidoylketene 7. This reaction
must be conformer-selective; i.e., the syn-5 leads to syn-(Z)-6
through a [1,5] sigmatropic H-shift (the only conformer
identified), whereas the anti-5 gives rise to anti-(E)-7 through a
[1,3] sigmatropic H-shift (also the only conformer identified)
(see also Scheme 4). However, the ratio of syn-5:anti-5 before
irradiation (∼0.1:1) does not match the ratio of products syn-
(Z)-6:anti-(E)-7 formed after the irradiation (∼0.6:0.5). This
fact could be due to the existence, during the irradiation
process, of facile photoisomerization between syn-5 and anti-5
(several examples of isomerizations of aldehydes of this type,
during the UV-irradiation, are known73,74). To the best of our
knowledge, the phototransformation of 3-formylketenimine is
reported for the first time herein. The [1,5] sigmatropic H-shift
reaction of ketenimines to give hydroxynitrile derivatives was
only suggested in the literature as a thermally induced reaction
to explain the results of the study of flash vacuum pyrolysis
(FVP) matrix-isolation of pyrrolidone derivatives, namely, the
conversion of the imidoylketene derivatives (primary products)
to the isomeric nitriles (final products).52 The [1,3] sigmatropic
H-shift reaction of ketenimines to give imidoylketenes (or the
corresponding reverse isomerization) is, on the other hand, a
better known thermal reaction described in several FVP matrix-
isolation studies.47−52
Finally, the formation of ketonitrile 8 was observed to occur
through phototautomerization of hydroxynitrile 6, based on the
results of irradiation experiments with λ = 240 nm of isoxazole
1 photoproducts. This observation is not completely
unexpected since the existence of thermal tautomeric
equilibrium between these species (with their ratio being
dependent on the solvent used) was already reported in the
FVP study of a Meldrum’s acid derivative.46 Furthermore, in
our previous reported study on the thermal reactivity of
isoxazoles, the possibility that this reaction takes place was also
suggested based on experimental and theoretical results.9
8730
dx.doi.org/10.1021/jo301699z | J. Org. Chem. 2012, 77, 8723−8732

The Journal of Organic Chemistry
4. EXPERIMENTAL SECTION
Sample. A commercial sample of isoxazole 1 (Aldrich, 99% purity)
was used.
Article
ACKNOWLEDGMENTS
■
These studies were partially funded by the Portuguese
“Fundacao para a Ciencia e a Tecnologia” (FCT), FEDER,
̧
̂
̃
Matrix-Isolation FTIR Spectroscopy. Prior to usage, the sample
was degassed by using the standard freeze−pump−thaw method. The
sample vapor was premixed with high purity argon (N60, supplied by
Air Liquide), in ratios from 1:1000 to 1:2000, using standard
monomeric techniques. The matrices were prepared by effusive
deposition of the premixed samples, from a 3 L glass reservoir, and the
flux was controlled by reading the drop pressure in the reservoir. A CsI
window, cooled to 15 K, was used as an optical substrate for the
matrices. The temperature of the CsI window was measured directly
by a silicone diode sensor connected to a digital controller with
accuracy of 0.1 K. In all experiments, an APD Cryogenics closed
helium refrigeration system with DE-202A expander was used.
The IR spectra were obtained using a Fourier transform infrared
spectrometer, equipped with a deuterated triglycine sulfate (DTGS)
detector and a Ge/KBr beam splitter, with 0.5 cm−1 resolution. To
avoid interference from atmospheric H2O and CO2, the optical path of
the spectrometer was continuously purged by a stream of dry air.
UV-Laser Irradiation Experiments. The matrices were irradiated
through an outer quartz window of the cryostat, using a frequency-
doubled signal beam provided by a MOPO-SL optical parametric
oscillator (fwhm ∼ 0.2 cm−1, repetition rate = 10 Hz, pulse energy ∼ 1
mJ, duration = 10 ns) pumped with a pulsed Nd:YAG laser.
Theoretical Calculations. All calculations were performed with
Gaussian 0375 at MP276 and DFT levels of theory using the standard
6-311++G(d,p) basis set.77,78 The DFT calculations were carried out
with the three-parameter density functional B3LYP,79 which includes
Becke’s gradient exchange correction,80 the Lee, Young, Parr,81 and
the Vosko, Wilk, Nusair correlation functionals.82 The geometry
optimizations were followed by harmonic frequency calculations, at the
same level of theory, which also allowed characterization of the nature
of the stationary points. To correct for the vibrational anharmonicity,
basis set truncation, and the neglected part of electron correlation, the
calculated frequencies below 3100 cm−1 were scaled down by 0.980
(B3LYP) or 0.976 (MP2) and by 0.950 if they are above 3100 cm−1.
The resulting frequencies, together with the calculated intensities, were
used to simulate the spectra shown in the figures by convoluting each
peak with a Lorentzian function with a full width at half-maximum
(fwhm) of 2 cm−1.83 Note that the peak intensities in the simulated
spectra (arbitrary units of “relative intensity”) are several times less
than the calculated intensities (in km mol−1).
and projects PTDC/QUI-QUI/111879/2009 and PTDC/
QUI-QUI/118078/2010, FCOMP-01-0124-FEDER-021082,
cofunded by QREN-COMPETE-UE. C. M. Nunes acknowl-
edges FCT for Grant No. SFRH/BD/28844/2006 and I. Reva
for Grant No. IF/00464/2012. We thank Prof. Hugh Douglas
Burrows for the helpful discussions.
REFERENCES
■
(1) Comprehensive Heterocyclic Chemistry; Lang, S. A., Lin, Y. I., Eds.;
Pergamon: Oxford, 1984; Vol. 6, Part 4B.
(2) Chemistry of Heterocyclic Compounds; Grunanger, P., Vita-Finzi,
̈
P., Eds.; John Wiley and Sons: New York, 1991; Vol. 49, Part 1.
(3) Comprehensive Heterocyclic Chemistry III; Giomi, D., Cordero, F.
M., Machetti, F., Eds.; Elsevier: Oxford, 2008; p 365.
(4) Pevarello, P.; Amici, R.; Brasca, M. G.; Villa, M.; Varasi, M.
Targets Heterocycl. Syst.: Chem. Prop. 1999, 3, 301.
(5) Pinho e Melo, T. M. V. D. Curr. Org. Chem. 2005, 9, 925.
(6) Fonseca, S. M.; Burrows, H. D.; Nunes, C. M.; Pinho e Melo, T.
M. V. D.; Gonsalves, A. M. d’A. R. Chem. Phys. Lett. 2005, 414, 98.
(7) Fonseca, S. M.; Burrows, H. D.; Nunes, C. M.; Pinho e Melo, T.
M. V. D. Chem. Phys. Lett. 2009, 474, 84.
́
(8) Lopes, S.; Nunes, C. M.; Gomez-Zavaglia, A.; Pinho e Melo, T.
M. V. D.; Fausto, R. J. Phys. Chem. A 2011, 115, 1199.
(9) Nunes, C. M.; Reva, I.; Pinho e Melo, T. M. V. D.; Fausto, R.;
̌
Solomek, T.; Bally, T. J. Am. Chem. Soc. 2011, 133, 18911.
(10) Dalvie, D. K.; Kalgutkar, A. S.; Khojasteh-Bakht, S. C.; Obach,
R. S.; O’Donnell, J. P. Chem. Res. Toxicol. 2002, 15, 269.
(11) Kalgutkar, A. S.; Nguyen, H. T.; Vaz, A. D. N.; Doan, A.; Dalvie,
D. K. Drug Metab. Dispos. 2003, 31, 1240.
(12) Yu, J.; Folmer, J. J.; Hoesch, V.; Doherty, J.; Campbell, J. B.;
Burdette, D. Drug Metab. Dispos. 2011, 39, 302.
(13) Ullman, E. F.; Singh, B. J. Am. Chem. Soc. 1966, 88, 1844.
(14) Singh, B.; Ullman, E. F. J. Am. Chem. Soc. 1967, 89, 6911.
(15) Singh, B.; Zweig, A.; Gallivan, J. B. J. Am. Chem. Soc. 1972, 94,
1199.
(16) D’Auria, M. Adv. Heterocycl. Chem. 2001, 79, 41.
(17) Tanaka, H.; Osamura, Y.; Matsushita, T.; Nishimoto, K. Bull.
Chem. Soc. Jpn. 1981, 54, 1293.
(18) Good, R. H.; Jones, G. J. Chem. Soc. C 1971, 1196.
(19) Padwa, A.; Chen, E.; Ku, A. J. Am. Chem. Soc. 1975, 97, 6484.
(20) Grellmann, K. H.; Tauer, E. J. Photochem. 1977, 6, 365.
(21) Sauers, R. R.; Hadel, L. M.; Scimone, A. A.; Stevenson, T. A. J.
Org. Chem. 1990, 55, 4011.
(22) Kurtz, D. W.; Shechter, H. Chem. Commun. 1966, 689.
(23) Ferris, J. P.; Antonucci, F. R.; Trimmer, R. W. J. Am. Chem. Soc.
1973, 95, 919.
ASSOCIATED CONTENT
■
S
* Supporting Information
Tables listing the observed IR peaks and their assignment based
on MP2 or B3LYP calculations or experimental reported data,
for ketene 2, hydrogen cyanide 3, 3-formylketenimine 5,
2‑formyl-2H-azirine 4, and 3-hydroxypropenenitrile 6. IR
spectra from laser irradiation isoxazole 1 with λ = 221 nm at
different times. PES profile for internal rotation of the C−C
bond of 2‑formyl-2H-azirine 4, 3-formyl-N-ketenimine 5, and 3-
oxopropanenitrile 8. Calculated relative energies for the
different conformations of 3-hydroxypropenenitrile 6 and
imidoylketene 7. Cartesian coordinates and frequencies (2−
8) from MP2/6-311++G(d,p) and (4−5) from B3LYP/6-311+
+G(d,p). This material is available free of charge via the
Internet at http://pubs.acs.org.
(24) Sato, T.; Yamamoto, K.; Fukui, K. Chem. Lett. 1973, 111.
(25) Ferris, J. P.; Antonucci, F. R. J. Am. Chem. Soc. 1974, 96, 2014.
(26) Heinzelmann, W.; Marky, M. Helv. Chim. Acta 1974, 57, 376.
̈
(27) Sato, T.; Saito, K. J. Chem. Soc., Chem. Commun. 1974, 781.
(28) Dietliker, K.; Gilgen, P.; Heimgartner, H.; Schmid, H. Helv.
Chim. Acta 1976, 59, 2074.
(29) Ferris, J. P.; Trimmer, R. W. J. Org. Chem. 1976, 41, 13.
(30) Sato, T.; Yamamoto, K.; Fukui, K.; Saito, K.; Hayakawa, K.;
Yoshiie, S. J. Chem. Soc., Perkin Trans. 1 1976, 783.
(31) Sauers, R. R.; Van Arnum, S. D. Tetrahedron Lett. 1987, 28,
5797.
(32) Pavlik, J. W.; St. Martin, H.; Lambert, K. A.; Lowell, J. A.;
Tsefrikas, V. M.; Eddins, C. K.; Kebede, N. J. Heterocycl. Chem. 2005,
42, 273.
(33) Adembri, G.; Donati, D.; Ponticelli, F. J. Photochem. 1981, 17,
81.
AUTHOR INFORMATION
■
Corresponding Author
*E-mail: cmnunes@qui.uc.pt; reva@qui.uc.pt.
Notes
(34) Tidwell, T. T. Spectroscopy and Physical Properties of Ketenes.
In Ketenes II, Second ed.; John Wiley and Sons: Hoboken, NJ, 2006.
The authors declare no competing financial interest.
8731
dx.doi.org/10.1021/jo301699z | J. Org. Chem. 2012, 77, 8723−8732

The Journal of Organic Chemistry
Article
(35) Moore, C. B.; Pimentel, G. C. J. Chem. Phys. 1963, 38, 2816.
(36) Romano, R. M.; Della Vedova, C. O.; Downs, A. J. J. Phys. Chem.
A 2002, 106, 7235.
(37) Breda, S.; Reva, I.; Fausto, R. J. Phys. Chem. A 2012, 116, 2131.
(38) Milligan, D. E.; Jacox, M. E. J. Chem. Phys. 1967, 47, 278.
(39) Satoshi, K.; Takayanagi, M.; Nakata, M. J. Mol. Struct. 1997, 413,
365.
(68) Inui, H.; Murata, S. Chem. Commun. 2001, 1036.
(69) Inui, H.; Murata, S. J. Am. Chem. Soc. 2005, 127, 2628.
(70) Gomez-Zavaglia, A.; Kaczor, A.; Cardoso, A. L.; Pinho e Melo,
T. M. V. D.; Fausto, R. J. Phys. Chem. A 2006, 110, 8081.
(71) Kaczor, A.; Gomez-Zavaglia, A.; Cardoso, A. L.; Pinho e Melo,
T. M. V. D.; Fausto, R. J. Phys. Chem. A 2006, 110, 10742.
(72) Inui, H.; Murata, S. Chem. Phys. Lett. 2002, 359, 267.
(73) Giuliano, B. M.; Reva, I.; Fausto, R. J. Phys. Chem. A 2010, 114,
2506.
́
́
(40) Guennoun, Z.; Couturier-Tamburelli, I.; Combes, S.; Aycard, J.
P.; Pietri, N. J. Phys. Chem. A 2005, 109, 11733.
(74) Kus,̧ N.; Reva, I.; Fausto, R. J. Phys. Chem. A 2010, 114, 12427.
(41) The analysis of successive 1 min irradiation experiments of 1
(λ = 221 nm) shows that some of the photoisomerization products
reach their maxima during the first minutes, while others reach their
maxima somewhat later (Figure S1, SI).
(42) The first irradiation experiment (λ = 320 nm, 1 min, Table S2,
SI) of photoisomerization products of 1, generated at λ = 221 nm, led
to the complete consumption of minor bands at 1915, 1657, and 868
cm−1 which correspond to an unidentified species.
(43) The photodecomposition products, ketene 2 and hydrogen
cyanide 3, do not react during irradiations with λ ≥ 240 nm.
(44) Krantz, A.; Hoppe, B. J. Am. Chem. Soc. 1975, 97, 6590.
(45) Briehl, H.; Lukosch, A.; Wentrup, C. J. Org. Chem. 1984, 49,
2772.
(75) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B. et al. GAUSSIAN
03, Revision C.02; Gaussian, Inc.: Wallingford, CT, 2004.
(76) Head-Gordon, M.; Head-Gordon, T. Chem. Phys. Lett. 1994,
220, 122.
(77) Frisch, M. J. P., J. A.; Binkley, J. S. J. Chem. Phys. 1984, 80, 3265.
(78) Clark, T. C., J.; Spitznagel, G. W.; Schleyer, P. V. R. J. Comput.
Chem. 1983, 4, 294.
(79) Becke, A. D. J. Chem. Phys. 1993, 98, 5648.
(80) Becke, A. D. Phys. Rev. A 1988, 38, 3098.
(81) Lee, C.; Yang, W.; Parr, R. G. Phys. Rev. B 1988, 37, 785.
(82) Vosko, S. H.; Wilk, L.; Nusair, M. Can. J. Phys. 1980, 58, 1200.
(83) Irikura, K. K. Program SYNSPEC; Natl. Inst. Standards and
Technol.: Gaithersburg, MD20899, USA.
(46) Wentrup, C.; Briehl, H.; Lorencak, P.; Vogelbacher, U. J.;
Winter, H. W.; Maquestiau, A.; Flammang, R. J. Am. Chem. Soc. 1988,
110, 1337.
(47) Bencheikh, A.; Chuche, J.; Manisse, N.; Pommelet, J. C.;
Netsch, K. P.; Lorencak, P.; Wentrup, C. J. Org. Chem. 1991, 56, 970.
(48) Kappe, C. O.; Kollenz, G.; Regis, L. T.; Wentrup, C. J. Chem.
Soc., Chem. Commun. 1992, 487.
(49) Fulloon, B.; Elnabi, H. A. A.; Kollenz, G.; Wentrup, C.
Tetrahedron Lett. 1995, 36, 6547.
(50) Fulloon, B. E.; Wentrup, C. J. Org. Chem. 1996, 61, 1363.
(51) Moloney, D. J. W.; Wong, M. W.; Flammang, R.; Wentrup, C. J.
Org. Chem. 1997, 62, 4240.
(52) George, L.; Bernhardt, P. V.; Netsch, K. P.; Wentrup, C. Org.
Biomol. Chem. 2004, 2, 3518.
(53) The imidoylketene 7 is reported to be generated from pyrolysis
of Meldrum’s acid, and its identification was suggested only based on
the observation of a band at 2130 cm−1 at 77 K (neat) (ref 45). This
band correlates well with the current observed band at 2136/2134
cm−1, which is attributed to anti-(E)-7 isolated in the argon matrix.
(54) The ν(C≡N) of ketonitrile anti-8 was estimated, at the MP2
level, to have an infrared absorption at least 100 times less intense than
the corresponding ν(CO), justifying in this way its nonobservation.
(55) Kim, H. S.; Kim, K. Bull. Korean Chem. Soc. 1992, 13, 520.
(56) Bernstein, M. P.; Sandford, S. A.; Allamandola, L. J. Astrophys. J.
1997, 476, 932.
(57) Sechkarev, A. V.; Fadeev, Y. A.; Reva, I. D. J. Appl. Spectrosc.
1999, 66, 708.
(58) Abe, H.; Takeo, H.; Yamada, K. M. T. Chem. Phys. Lett. 1999,
311, 153.
(59) Abe, H.; Yamada, K. M. T. Struct. Chem. 2003, 14, 211.
(60) The molecular modeling studies on the mechanism of the
isoxazole−oxazole photoisomerization (the most common reaction in
the isoxazole photochemistry) should be considered with care, if
analyzed for the simplest parent isoxazole 1, because this reaction was
not observed under present experimental conditions (as described in
this work).
(61) Karney, W. L.; Borden, W. T. J. Am. Chem. Soc. 1997, 119, 1378.
(62) Platz, M. S. Nitrenes, In Reactive Intermediate Chemistry; Moss, R.
A., Platz, M. S., Maitland, J., Eds.; Wiley-Interscience: Hoboken, NJ,
2004.
(63) Parasuk, V.; Cramer, C. J. Chem. Phys. Lett. 1996, 260, 7.
(64) Albini, A.; Bettinetti, G.; Minoli, G. J. Am. Chem. Soc. 1997, 119,
7308.
(65) Harder, T.; Stosser, R.; Wessig, P.; Bendig, J. J. Photochem.
Photobiol. A 1997, 103, 105.
(66) Murata, S.; Tsubone, Y.; Tomioka, H. Chem. Lett. 1998, 549.
(67) Inui, H.; Murata, S. Chem. Lett. 2001, 832.
8732
dx.doi.org/10.1021/jo301699z | J. Org. Chem. 2012, 77, 8723−8732