Comparison of the photochemistry of 3-methyl-2-phenyl-2h-azirine and 2-methyl-3-phenyl-2h-azirine

Article abstract of DOI:10.1021/jo402443w

Photolysis of 3-methyl-2-phenyl-2H-azirine (1a) in argon-saturated acetonitrile does not yield any new products, whereas photolysis in oxygen-saturated acetonitrile yields benzaldehyde (2) by interception of vinylnitrene 5 with oxygen. Similarly, photolysis of 1a in the presence of bromoform allows the trapping of vinylnitrene 5, leading to the formation of 1-bromo-1-phenylpropan-2-one (4). Laser flash photolysis of 1a in argon-saturated acetonitrile (? = 308 nm) results in a transient absorption with ?max at 440 nm due to the formation of triplet vinylnitrene 5. Likewise, irradiation of 1a in cryogenic argon matrixes through a Pyrex filter results in the formation of ketene imine 11, presumably through vinylnitrene 5. In contrast, photolysis of 2- methyl-3-phenyl-2H-azirine (1b) in acetonitrile yields heterocycles 6 and 7. Laser flash photolysis of 1b in acetonitrile shows a transient absorption with a maximum at 320 nm due to the formation of ylide 8, which has a lifetime on the order of several milliseconds. Similarly, photolysis of 1b in cryogenic argon matrixes results in ylide 8. Density functional theory calculations were performed to support the proposed mechanism for the photoreactivity of 1a and 1b and to aid in the characterization of the intermediates formed upon irradiation.

Full text of DOI:10.1021/jo402443w

Article
pubs.acs.org/joc
Comparison of the Photochemistry of 3‑Methyl-2-phenyl‑2H‑azirine
and 2‑Methyl-3-phenyl‑2H‑azirine
Xiaoming Zhang, Sujan K. Sarkar, Geethika K. Weragoda, Sridhar Rajam, Bruce S. Ault,
and Anna D. Gudmundsdottir*
Department of Chemistry, University of Cincinnati, Cincinnati, Ohio 45221-0172, United States
S
* Supporting Information
ABSTRACT: Photolysis of 3-methyl-2-phenyl-2H-azirine (1a) in
argon-saturated acetonitrile does not yield any new products, whereas
photolysis in oxygen-saturated acetonitrile yields benzaldehyde (2) by
interception of vinylnitrene 5 with oxygen. Similarly, photolysis of 1a in
the presence of bromoform allows the trapping of vinylnitrene 5, leading
to the formation of 1-bromo-1-phenylpropan-2-one (4). Laser flash
photolysis of 1a in argon-saturated acetonitrile (λ = 308 nm) results in a
transient absorption with λmax at ∼440 nm due to the formation of
triplet vinylnitrene 5. Likewise, irradiation of 1a in cryogenic argon
matrixes through a Pyrex filter results in the formation of ketene imine
11, presumably through vinylnitrene 5. In contrast, photolysis of 2-
methyl-3-phenyl-2H-azirine (1b) in acetonitrile yields heterocycles 6 and 7. Laser flash photolysis of 1b in acetonitrile shows a
transient absorption with a maximum at 320 nm due to the formation of ylide 8, which has a lifetime on the order of several
milliseconds. Similarly, photolysis of 1b in cryogenic argon matrixes results in ylide 8. Density functional theory calculations were
performed to support the proposed mechanism for the photoreactivity of 1a and 1b and to aid in the characterization of the
intermediates formed upon irradiation.
Scheme 1
1. INTRODUCTION
Triplet nitrenes are electron-deficient reactive intermediates.1
Electron-donating groups such as alkyl and aryl substituents
render the triplet nitrenes more stable, and thus, they do not
react with the solvent but rather decay by dimerization.2−7 In
contrast, electron-withdrawing groups destabilize nitrenes,
making them more reactive. For example, triplet ester nitrenes
react efficiently by abstracting a H atom from the solvent.8−10
Interestingly, triplet vinylnitrenes that are stabilized by an
electron-donating vinyl substituent react differently from alkyl-
and arylnitrenes, as they decay by intersystem crossing.11−13
Solution photolysis of azirine derivatives can lead to the
formation of triplet vinylnitrenes by breaking the C−N bond in
the azirine.13,14 Simple vinylnitrenes must decay by intersystem
crossing to reform their starting materials as no new
photoproducts were observed,15,16 thus giving the impression
that azirine derivatives are not photoreactive, whereas vinyl-
nitrenes that have a carbonyl group at the γ-position can form
isoxazole products in addition to reforming the starting material
(Scheme 1).11−14,17 However, triplet vinylnitrenes have been
trapped with oxygen to form stable products in both solution
and cryogenic matrixes.11,15,16 Furthermore, the rate constant
for the reaction of vinylnitrenes with oxygen in solution has
been estimated to be between 7 × 108 and 2 × 109 M−1 s−1.11,13
Because of the significant 1,3-biradical character of vinyl-
nitrenes, they react efficiently with oxygen to form a new C−O
bond, whereas phenyl- and alkylnitrenes react with oxygen at a
lower rate to form the corresponding nitro compounds.3,5,18,19
Photolysis of azirine derivatives can also lead to ylide
formation by breaking the C−C bond of the azirine moiety
(Scheme 2).20−22 The ylide formation takes place on the singlet
surface of the azirine, and the ylides are long-lived
intermediates that can be trapped in 1,3-dipolar additions.23,24
In this work, we compared the photoreactivity of two 2H-
azirine isomers to gain better insight into when the azirine
Scheme 2
Received: November 2, 2013
Published: December 23, 2013
© 2013 American Chemical Society
653
dx.doi.org/10.1021/jo402443w | J. Org. Chem. 2014, 79, 653−663

The Journal of Organic Chemistry
Article
derivatives exhibit singlet versus triplet reactivity. We showed
that upon irradiation with light above 300 nm, 3-methyl-2-
phenyl-2H-azirine (1a) forms triplet vinylnitrene intermediate
5, which can be detected directly with transient spectroscopy.
Furthermore, in solution the triplet vinylnitrene 5 reforms 1a
but can be intercepted with oxygen and bromine radicals. In
contrast, photolysis of 2-methyl-3-phenyl-2H-azirine (1b)
results only in singlet reactivity to yield ylide 8, and no
formation of triplet vinylnitrene 10 was observed.
of 1a in bromoform to determine whether vinylnitrene can be
trapped by bromine radicals. It should be noted that direct
irradiation of bromoform with light above 300 nm results in the
quantitative formation of bromine radicals.25 Because products
3 and 4 are formed only in bromoform, we theorize that the
triplet vinylnitrene 5 is trapped with bromine radicals, as shown
in Scheme 4, and that 3 is formed from the secondary
photolysis of 4. As expected, an independent photolysis of 4
through a Pyrex filter confirmed that 4 yields 3. We also
hypothesized that energy transfer from the excited state of 1a to
bromoform could result in the formation of bromine radicals in
addition to direct photolysis of bromoform. Furthermore, we
theorized that triplet vinylnitrene 5 can react directly with
bromoform to abstract bromine atom to form 4 in addition to
reacting with bromine radicals.
2. RESULTS
2.1. Product Studies. Photolysis of 1a through a Pyrex
filter in argon-saturated acetonitrile did not yield any new
photoproducts. In contrast, the photolysis of 1a in argon-
saturated bromoform resulted in the formation of ketones 4.
Photolysis of 1a in oxygen-saturated acetonitrile resulted in the
formation of benzaldehyde (2), and irradiation of 1a in oxygen-
saturated bromoform yielded 2, 3, and 4 (Scheme 3).
In contrast, the photolysis of 1b in argon- and oxygen-
saturated solutions through a Pyrex filter affords heterocycles 6
and 7 in a 1:3 ratio (Scheme 5). This is in agreement with the
Scheme 5
Scheme 3
We theorize that the irradiation of 1a through a Pyrex filter
yields triplet vinylnitrene 5, which undergoes intersystem
crossing to reform 1a, but in the presence of oxygen, triplet
vinylnitrene 5 is trapped to form 2 (Scheme 4). To better
probe the reactivity of vinylnitrene 5, we studied the photolysis
results obtained by Gakis et al.24 and Padwa et al.,26,27 who
previously reported that the irradiation of 1b results in 6 and 7.
These authors theorized that 6 and 7 are formed by trapping of
ylide 8 in a 1,3-dipolar addition to 1b. Furthermore, we also
reconfirmed that ylide 8 can be trapped by a 1,3-dipolar
Scheme 4
654
dx.doi.org/10.1021/jo402443w | J. Org. Chem. 2014, 79, 653−663

The Journal of Organic Chemistry
Article
addition to carbon dioxide to yield 9 by photolyzing 1b in CO2-
saturated acetonitrile.28−30
Thus, the product studies indicate that the irradiation of 1a
through a Pyrex filter yields triplet vinylnitrene 5, which can be
intercepted with oxygen or bromine radicals, whereas the
irradiation of 1b yields products from singlet ylide 8, which is
trapped by bimolecular 1,3-dipolar cycloaddition reactions.
2.2. Calculations. To better understand why the reactivity
of 1a and 1b differ and to better comprehend the reactivity of
vinylnitrene 5 and ylide 8, we calculated stationary points on
the singlet and triplet energy surfaces of 1a and 1b. The
calculations were performed using Gaussian09 at the B3LYP
level of theory using the 6-31+G(d) basis set.31−33
We optimized the ground states (S0) of 1a and 1b and used
time-dependent density functional theory (TD-DFT) to
estimate the vertical excitation energies of their first excited
singlet states (S1) and their first and second excited triplet
states (T1 and T2). The S1 are located at 106 kcal/mol (Figure
4) and 97 kcal/mol (Figure 5) above the corresponding S0 for
1a and 1b, respectively. The T1 of 1a and 1b are located at 83
and 70 kcal/mol and the T2 at 96 and 87 kcal/mol above the
corresponding S0 states. Thus, the conjugation of the CN
bond with the phenyl ring in 1b lowers the energies of S1, T1,
and T2 in 1b in comparison to the energies of S1, T1, and T2 in
1a.
Figure 1. Optimized structures of 1a and 1b, the T1 of 1a, and the T1
of 1b. The numbers in parentheses are the calculated spin densities for
the T1 states of 1a and 1b.
We optimized T1 of 1a and located it at 72 kcal/mol above
its S0, which is considerably lower than the energy for the
Franck−Condon state described above (TD-DFT calculations).
We have previously shown that the energy obtained from
optimization of triplet ketones with an (n,π*) configuration is
systematically underestimated by density functional theory,34
but the difference in the energies for T1 of 1a obtained from the
TD-DFT calculations and the optimization must be due to the
geometrical difference in the optimized structures of T1 and S0
of 1a. The most significant difference between the structures of
T1 and S0 of 1a is that the CN bond length is 1.47 Å in the
T1 of 1a, indicating that the bond is best described as a single
bond (Figure 1). In addition, spin density calculations showed
that the unpaired electrons are localized on the C and N atoms
in the azirine moiety, thus demonstrating that the T1 of 1a is
localized on the azirine moiety of the molecule (Figure 1).
The optimized structure of the T1 of 1b is 62 kcal/mol above
its S0, which is also considerably lower in energy than predicted
from the TD-DFT calculations, and as before, the discrepancy
must be due to geometrical differences between the T1 and S0
of 1b. The most significant difference between the optimized
structures of the T1 of 1b and its S0 is the CN bond, which
has a length of 1.44 Å (Figure 1). Furthermore, the C−C bond
between the azirine and the phenyl moiety is only 1.37 Å and
thus somewhat shorter than in the S0 of 1b. In addition, the C−
C bonds in the phenyl ring are not all equivalent. Thus, the T1
of 1b is delocalized over the phenyl and azirine moieties. Spin
density calculations further support this, as the unpaired
electron density is delocalized over both the phenyl and azirine
moieties (Figure 1).
Figure 2. Optimized structures of conformers A and B of triplet
vinylnitrene 5. The numbers in parentheses are the calculated spin
densities.
atoms in the vinylnitrene moiety and also on the ortho and para
C atoms in the phenyl ring.
We also optimized two minimal energy conformers, A and B,
of triplet vinylnitrene 10, the nitrene that corresponds to azirine
1b (Figure 3). Because the vinylnitrene moiety in 10 is in cross-
conjugation with the phenyl ring, the spin density is mainly
localized over the vinylnitrene moiety and not the phenyl
group.
We optimized two minimal-energy conformers, A and B, of
triplet vinylnitrene 5, with A being 3 kcal/mol more stable than
B. The Ph−C bond in vinylnitrene 5 is shorter than an average
single C−C bond because the vinylnitrene moiety is in
conjugation with the phenyl ring (Figure 2). This is further
highlighted by the spin density calculations, which show that
the unpaired electrons are mainly located on the N and β-C
The transition state for the formation of triplet vinylnitrene 5
from the T1 of 1a is located 10 kcal/mol above the T1 of 1a
(Figure 4). Thus, it is reasonable to expect that population of
the T1 of 1a will result in the formation of triplet vinylnitrene 5.
As a comparison, the transition state barrier for forming
vinylnitrene 10 from the T1 of 1b is 15 kcal/mol (Figure 5).
Thus, the delocalization of the T1 of 1b over the entire
655
dx.doi.org/10.1021/jo402443w | J. Org. Chem. 2014, 79, 653−663

The Journal of Organic Chemistry
Article
Scheme 6
our calculations suggest that the phenyl group in 5 does not
affect the triplet−singlet energy gap. In comparison, Nunes et
al.12 theorized that conical intersections between the triplet and
open-shell singlet are responsible for effective intersystem
crossing in vinylnitrene derivatives.
Because triplet vinylnitrenes typically undergo intersystem
crossing to form ketene imines in cryogenic matrixes,13,40−42 we
calculated the triplet transition state for the formation of triplet
ketene imine 11 from triplet vinylnitrene 5. This transition state
is located 66 kcal/mol above vinylnitrene 5; in comparison, the
transition state for the formation of triplet ketene imine 12
from triplet vinylnitrene 10 is 38 kcal/mol above 10. Thus, the
transition state barriers for triplet vinylnitrenes 5 and 10 to
rearrange to the corresponding triplet ketene imines indicate
that these rearrangements are not feasible. It is more likely that
triplet vinylnitrenes 5 and 10 undergo intersystem crossing to
form ketene imines 11 and 12, respectively.
Figure 3. Optimized structures of conformers A and B of triplet
vinylnitrene 10. The numbers in parentheses are the calculated spin
densities.
The calculated stationary points on the triplet surface for
vinylnitrene 5 reacting with oxygen to form benzaldehyde 2 are
displayed in Figure 6. Vinylnitrene 5 can react with oxygen to
form biradical 13, which is 8 kcal/mol more stable than 5 and
O2, and the calculated transition state barrier for this reaction is
6 kcal/mol above vinylnitrene 5 and O2. Thus, the calculations
show that vinylnitrene 5 is expected to react efficiently with
molecular oxygen at ambient temperature. We theorized that
biradical 13 can be cleaved to form biradical 14 and acetonitrile,
but these products are 2 kcal/mol less stable than 13, and the
calculated transition state barrier for cleavage of 13 to form 14
and acetonitrile is 18 kcal/mol above 13. Although 14 can
undergo intersystem crossing to form ylide 15, which is 27
kcal/mol more stable, it is more likely that biradical 13 forms
ylide 15 via intersystem crossing, followed by hydrolysis to
form benzaldehyde 2. Thus, the calculations support the theory
that triplet vinylnitrene 5 can react with oxygen at ambient
temperature to form benzaldehyde 2.
We calculated the feasibility of forming bromine radical from
bromoform by energy transfer from the S1 or T1 of 1a (Figure
7). The calculated bond dissociation energy for bromoform to
form bromine radical is only 65 kcal/mol, significantly less than
the energy of the S1 and T1 of 1a. Thus, the calculations
support the idea that 1a can be used as a sensitizer to cleave
bromoform to form bromine radicals. We could not calculate
the transition state for the reaction of bromine radical with
vinylnitrene 5 because of spin restrictions.
Figure 4. Calculated stationary points on the energy surface of 1a. The
numbers in parentheses are the calculated energies in kcal/mol. The
energies of S1 and T2 and the value of 83 kcal/mol for T1 were
obtained by TD-DFT calculations, whereas the other energies were
obtained by optimization calculations.
Figure 5. Calculated stationary points on the energy surface of 1b. The
numbers in parentheses are the calculated energies in kcal/mol. The
energies of S1 and T2 and the value of 70 kcal/mol for T1 were
obtained by TD-DFT calculations, whereas the other energies were
obtained by optimization calculations.
molecule stabilizes the T1 of 1b, which in turns makes it more
difficult to break the N−C bond in the azirine ring to form
vinylnitrene 10.
Addition of bromine radical to vinylnitrene 5 results in the
formation of radical 16, which is 56 kcal/mol more stable than
5 and bromine radical (Figure 7). Thus, the calculations show
that the formation of bromine radical by sensitization and the
addition of bromine radicals to vinylnitrene 5 are very feasible
at ambient temperature.
We optimized the open-shell singlet vinylnitrenes 5 and 10
to estimate their triplet−singlet energy gaps and found that the
energy gaps are 17 kcal/mol for 5 and 14 kcal/mol for 10
(Scheme 6). For comparison, we calculated the triplet−singlet
energy gap for vinylnitrene (CH2CH−N:) and obtained a
value of 16 kcal/mol, which is similar to the value of 15 kcal/
mol reported by Parasuk and Cramer for vinylnitrene.35
Although Wenthold and Hossain have demonstrated that
radical stabilization by electron delocalization effectively
reduces the triplet−singlet energy gap of vinylnitrenes,36,37
In addition, we calculated the transition state barrier for the
reaction of vinylnitrene 5 with bromoform to form 16 and
•CHBr2. The transition state barrier for 5 to abstract a bromine
•
atom from bromoform to yield 16 and CHBr2 is 11 kcal/mol
656
dx.doi.org/10.1021/jo402443w | J. Org. Chem. 2014, 79, 653−663

The Journal of Organic Chemistry
Article
Figure 6. Calculated stationary points on the energy surface of triplet vinylnitrene 5 reacting with O2. Energies are in kcal/mol.
Figure 9. Energy diagram for the formation of ylide 8 from 1b.
Figure 7. Energy calculation for the formation of bromine radicals
from bromoform and trapping of vinylnitrene 5 with bromine radical.
Energies are in kcal/mol.
(Figure 8). Thus, it is also possible for vinylnitrene 5 to abstract
a bromine atom from bromoform directly.
Finally, we calculated the singlet reactivity of azirine 1b to
form ylide 8 (Figure 9). The transition state barrier for the
formation of ylide 8 from azirine 1b was located 43 kcal/mol
above the S0 of 1b. The calculated transition state barrier for 1b
to form ylide 8 is similar to those we calculated for other azirine
derivatives.13
2.3. Laser Flash Photolysis. To identify the intermediates
formed upon the photolysis of 1a and 1b, we used laser flash
photolysis (excimer laser, 17 ns, 308 nm). Laser flash photolysis
of 1a in argon-saturated acetonitrile produced a broad transient
spectrum with λmax at ∼440 nm (Figure 10A). We assign this
transient spectrum to triplet vinylnitrene 5 on the basis of the
similarity between this transient spectrum and the calculated
absorption spectrum of 5 (Figure 10B). The TD-DFT
calculations in acetonitrile predicted the most significant
electronic transitions above 300 nm for vinylnitrene 5A at
301 nm (f = 0.070), 306 nm (f = 0.053), 405 nm (f = 0.0145),
and 420 nm (f = 0.0104) and for vinylnitrene 5B at 314 nm (f
= 0.0155), 384 nm (f = 0.0027), and 398 nm (f = 0.0088). It
should be emphasized that we could not measure the transient
absorption below 330 nm because of the ground-state
absorption of 1a, and therefore could not verify that
vinylnitrene 5 has a strong band at approximately 300 nm. In
Figure 10. (A) Transient UV−vis spectrum obtained by laser flash
photolysis of 1a. (B) Calculated absorption spectra of 5A, 5B and the
T1 of 1a in acetonitrile. (C) Kinetic trace obtained at 460 nm.
comparison, the calculated spectrum of the T1 of 1a has its
most significant electronic transitions at 303 nm (f = 0.0196),
304 nm (f = 0.0189), 342 nm (f = 0.0023), 391 nm (f =
0.0011), 480 nm (f = 0.0112), and 507 nm (f = 0.0008), which
•
Figure 8. Energy diagram for the reaction of vinylnitrene 5 with bromoform to form 16 and CHBr2.
657
dx.doi.org/10.1021/jo402443w | J. Org. Chem. 2014, 79, 653−663

The Journal of Organic Chemistry
Article
is a considerably broader absorption range than in the observed
spectrum, thus further supporting the assignment of the
transient to vinylnitrene 5. The transient absorption was
formed faster than the time resolution of the laser flash
photolysis apparatus. The decay of the transient absorption at
460 nm (Figure 10C) was best fit as an exponential decay with
a rate constant of 1.5 × 105 s−1, corresponding to a lifetime of 7
μs. In oxygen-saturated solutions, the transient absorption was
fully quenched, whereas in air-saturated methanol, the rate
constant for the decay was 1.7 × 106 s−1. The rate constant for
quenching of triplet vinylnitrene with oxygen is 8.9 × 108 M−1
s−1 because the concentration of oxygen in air-saturated
acetonitrile is 0.0019 M.38 This rate constant for oxygen
quenching is similar to those we have measured for different
triplet vinylnitrene derivatives (7 × 108 and 2 × 109 M−1
s−1).11,13
experiments clearly demonstrate that 1a exhibits triplet
reactivity to form triplet vinylnitrene 5 whereas 1b reacts on
the singlet surface to form ylide 8.
2.4. Stern−Volmer Analysis. To further characterize
triplet vinylnitrene 5 and its reactivity, we analyzed how its
formation and decay are affected by the addition of isoprene.
Figure 12A shows the kinetic traces for triplet vinylnitrene 5 at
Laser flash photolysis of azirine 1b in acetonitrile produced
transient spectra with λmax at ∼320 nm (Figure 11A). We assign
Figure 12. (A) Kinetic traces obtained at 460 nm from the laser flash
photolysis of 1a with various isoprene concentrations. (B) Stern−
Volmer plot obtained from the kinetic traces shown in (A). I0 is the
absorption intensity in the absence of isoprene, and I is the absorption
intensity at various isoprene concentrations.
460 nm with various amounts of isoprene. The intensity of the
absorption was reduced upon the addition of isoprene, whereas
the lifetime of 5 was not significantly affected, indicating that
the isoprene quenches the precursor to vinylnitrene 5. The
Stern−Volmer plot yielded a straight line with a slope of 3.4
M−1 (Figure 12B). Thus, the lifetime of the precursor to triplet
vinylnitrene 5 can be deduced as being between 3.4 and 0.34 ns
by assuming that the quenching is diffusion-controlled (i.e., that
kq is between 109 and 1010 M−1 s−1).
Figure 11. (A) Transient spectrum obtained by laser flash photolysis
of 1b in argon-saturated acetonitrile. (B) TD-DFT calculated
spectrum for 8. (C) Kinetic trace obtained at 320 nm.
We also studied the effect of bromoform on the formation
and decay of triplet vinylnitrene 5 in acetonitrile. The yield for
the formation of vinylnitrene 5 decreased as the concentration
of bromoform increased (Figure 13A), demonstrating that the
precursor to vinylnitrene 5 is quenched with bromoform. A
Stern−Volmer plot of I0/I versus [bromoform] gave a slope of
2.1 M−1 (Figure 13B), and with the assumption that the
quenching is diffusion-controlled, the precursor has a lifetime
between 0.21 and 2.1 ns, which is the same lifetime, within the
experimental error, as obtained for the T1 of 1a by isoprene
quenching. Hence, we conclude that T1 of 1a is quenched by
bromoform to form bromine radicals. Addition of bromoform
reduced not only the yield of vinylnitrene 5 but also its lifetime.
Plotting the rate constant for the decay of vinylnitrene 5 as a
function of the bromoform concentration yielded a straight line
with a slope of 1.7 × 105 M−1s−1 (Figure 13C). The reaction of
vinylnitrene 5 with bromoform is significantly slower than that
of vinylnitrene 5 with oxygen, which fits with the result that the
transition state barrier for reaction of vinylnitrene 5 with
bromoform (11 kcal/mol) is significantly larger than that for
the reaction of vinylnitrene 5 with oxygen (6 kcal/mol).
the absorbance at 320 nm to ylide 8 on the following basis.
Stenkeen and co-workers have shown previously that laser flash
photolysis by 248 nm irradiation of 3-phenyl-2H-azirine
derivatives yields ylides with λmax at ∼290 nm and a shoulder
at ∼400 nm.23 Laser flash photolysis of 1b with 308 nm
irradiation does not allow detection of the transient absorption
below 320 nm because of ground-state absorption of 1, and
thus, we observed only the tail absorption for ylide 8. In
addition, TD-DFT calculations showed that the major
absorptions for 8 are placed at 270 nm (f = 0.0636), 306 nm
(f = 0.0002) and 456 nm (f = 0.0017) (Figure 11B), which fits
with the conclusion that we observed only the tail of the
absorption of ylide 8. The lifetime of ylide 8 at 320 nm is on
the order of several milliseconds (Figure 11C). As expected, the
transient spectrum of ylide 8 and its lifetime were not affected
by oxygen.
Thus, the results of laser flash photolysis of 1a and 1b with a
308 nm laser are in agreement with the product studies
performed by irradiation through Pyrex filter, as both sets of
658
dx.doi.org/10.1021/jo402443w | J. Org. Chem. 2014, 79, 653−663

The Journal of Organic Chemistry
Article
imine 11 (Scheme 7) on the basis of the following observations.
The calculated IR spectrum of 11 has an intense band at 2122
Scheme 7
cm−1 (f = 776) due to CCN stretching, and scaling by
0.961339 places this band at 2040 cm−1, which matches very
closely with the observed bands at 2054 and 2043 cm−1. The
multiplet of this band is likely due to the presence of different
conformers of 11, which can also be trapped in different matrix
sites. We assign the band at 1171 cm−1 to the stretch that is
coupled between the phenyl ring and the CCN moiety,
calculated to be at 1206 cm−1 (f = 42) or 1159 cm−1 with
scaling. The band at 829 cm−1 is assigned to the coupled C
CN and −CH3 stretch at 903 cm−1 (f = 14), 868 cm−1 with
scaling. Finally, we assign the band at 689 cm−1 to a coupled
CCN and phenyl bending, which is calculated to be at 649
cm−1 (f = 22).
The photolysis of 1b in argon matrixes at 14 K resulted in
the reduction or depletion of the bands at 1742, 1635, 1586,
1386, 1337, 1330, 1327, 1151, and 751 cm−1 (Figure 15).
Concurrently, bands at 1895, 1887, 1859, 1374, 1292, 1263,
1137, 1069, 785, 783, 772, 770, 700, and 602 cm−1 were
formed, and we assign them to ylide 8 (Scheme 8) on the basis
of the following observations. The calculated IR spectrum of 8
has an intense band at 1968 cm−1 (f = 133) due to its CN
C stretching. Scaling this band by 0.9613 places it at 1892 cm−1,
which fits nicely with the bands observed at 1895, 1887, and
1859 cm−1. As before, we theorize that the multiplicity of this
band is due to the presence of different conformers of 8, which
can also be trapped at different matrix sites. The band at 1374
cm−1 is assigned to wagging of the methyl group, which is
calculated to be at 1426 cm−1 (f = 16); scaling places this band
at 1371 cm−1. The bands at 1292, 1263, and 1137 cm−1 are
assigned to C−H wagging coupled with twisting of the
aromatic ring and are calculated to be at 1336, 1303, and
1157 cm−1. Furthermore, the band at 1069 cm−1 is assigned to
twisting of the aromatic ring, which is calculated to be at 1103
cm−1 (f = 23). Finally, we assign the bands at 786 and 783
cm−1 to a coupled phenyl and CN bending, calculated to be
at 800 cm−1 (f = 39). The bands at 772, 770, and 700 cm−1 are
assigned to NC−H bending coupled to aromatic C−H
bending, calculated to be at 788 cm−1 (f = 31) and 721 cm−1 (f
= 71). Finally, we also observed CNC bending at 602
cm−1, calculated to be at 620 cm−1 (f = 88).
Figure 13. (A) Kinetic traces obtained at 460 nm from the laser flash
photolysis of 1a at various bromoform concentrations. (B) Stern−
Volmer plot obtained from the kinetic traces shown in (A). I0 is the
absorption intensity in the absence of bromoform, and I is the
absorption intensity at various bromoform concentrations. (C) The
decay rate constant at 460 nm as a function of the bromoform
concentration.
However, because bromine radicals are formed both by direct
photolysis of bromoform and by sensitization of bromoform
with 1a, it is just as likely that vinylnitrene 5 reacts with
bromine radicals rather than bromoform and that the lower rate
is a reflection of the lower concentration of bromine radicals
compared with bromoform.
2.5. Matrix Isolation. We studied the photolysis of 1a and
1b through a Pyrex filter in argon matrixes at 14 K to confirm
that 1a and 1b exhibit triplet and singlet reactivity, respectively.
We deposited 1a and 1b into argon matrixes in several different
experiments, entraining the vapor over a sample of pure azirine
in flowing argon and depositing the resulting sample on a 14 K
cryogenic surface.
Photolysis of 1a through a Pyrex filter for 2 h in argon
matrixes resulted in reductions of the IR bands at 1787, 1294,
1267, 1078, 1031, and 968 cm−1 (Figure 14). Concurrently,
new bands were observed at 2054, 2043, 1071, 829, and 689
cm−1. We assign these new bands to the formation of ketene
Thus, the photolysis of 1a in argon matrixes through a Pyrex
filter yields 11; ketene imine formation is a characteristic of the
reactivity of triplet vinylnitrenes in matrices.13,40−42 In contrast,
the photolysis of 1b in argon matrixes results in ylide formation,
as generally observed for the singlet reactivity of azirines.
3. DISCUSSION
Product studies showed that azirine 1b does not yield any
products when irradiated with light above 300 nm. However,
laser flash photolysis and matrix isolation of 1a with light above
300 nm confirmed the formation of triplet vinylnitrene 5 with a
lifetime of 7 μs. The lifetime of vinylnitrene 5 is somewhat
Figure 14. Matrix isolation spectra of 11. (A) IR spectrum obtained
after the photolysis of 1a through a Pyrex filter. (B) Difference of the
IR spectra of 1a before and after irradiation through a Pyrex filter. The
positive bands belong to 11 and the negative bands to 1a.
659
dx.doi.org/10.1021/jo402443w | J. Org. Chem. 2014, 79, 653−663

The Journal of Organic Chemistry
Article
Figure 15. Difference of the IR spectra obtained before and after the irradiation of 1b through a Pyrex filter. The positive bands belong to 8 and the
negative bands to 1b.
must be slower than in 1a, as it does not compete with the ylide
formation. The intersystem crossing in 1b is presumably slow
because the azirine and phenyl moieties are fully conjugated
and thus in more rigid alignment than in 1a, where the azirine
and phenyl π system are orthogonal to each other. Finally,
because of the conjugation of the phenyl ring with the azirine
moiety in 1b, the transition state barrier for breaking the C−N
bond to form vinylnitrene 10 is considerably larger than that for
formaton of vinylnitrene 5 from 1a.
Scheme 8
longer than we previously reported for vinylnitrenes 17 and 18
(depicted in Scheme 9).11,13
Scheme 9
4. CONCLUSION
We have shown that the photolysis of azirine 1a through a
Pyrex filter results in triplet vinylnitrene formation, whereas
irradiation of 1b yields only ylide formation. The triplet
vinylnitrene 5 has a lifetime of a few microseconds and can be
trapped both with oxygen and bromine radicals.
We theorize that the greater spin delocalization of 5 stabilizes
it more than vinylnitrenes 17 and 18. The rate constant for
interception of vinylnitrene 5 with oxygen is 8.9 × 108 M−1 s−1,
whereas that for trapping of vinylnitrene 5 with bromoform is
significantly lower (1.7 × 105 M−1 s−1). The lower rate for the
reaction of vinylnitrene 5 with bromoform can be a reflection of
the larger transition state barrier for the reaction of 5 with
bromoform compared with oxygen. However, since direct
photolysis of bromoform yields bromine radicals, it is just as
likely that the lower rate for the reaction of vinylnitrene 5 with
bromoform is a reflection of the fact that the Stern−Volmer
plot was graphed using the concentration of bromoform rather
than bromine radicals and that a larger rate for the trapping of
vinylnitrene 5 would be obtained as a function of bromine
radicals.
5. EXPERIMENTAL SECTION
5.1. Molecular Modeling. All of the geometries were optimized at
the B3LYP level of theory with the 6-31G+(d) basis set as
implemented in Gaussian09. Transition states were confirmed to
have one imaginary vibrational frequency by analytical determination
of the second derivatives of the energy with respect to internal
coordinates. Intrinsic reaction coordinate (IRC) calculations were used
to verify that each located transition state corresponded to the
attributed reactant and product.43,44 The absorption spectra were
calculated using time-dependent density functional theory (TD-
DFT).45−47 The effect of solvation was calculated using the self-
consistent reaction field (SCRF) method with the integral equation
formalism polarization continuum model (IEFPCM) with acetonitrile,
methanol, and dichloromethane as solvents.48−51
5.2. Laser Flash Photolysis. Laser flash photolysis was performed
with an excimer laser (308 nm, 17 ns). The system has been described
in detail elsewhere.52 In a typical experiment, a stock solution of 1a or
1b was prepared with spectroscopic-grade acetonitrile such that the
solutions had an absorbance between 0.3 and 0.6 at 308 nm. Typically,
∼1 mL of the stock solution was placed in a 10 mm × 10 mm × 48
mm quartz cuvette and purged with argon or oxygen for 5 min. The
rate was obtained by fitting the average of two to three kinetic traces.
5.3. Matrix Isolation. Matrix isolation studies were performed
using conventional equipment.53
In comparison, the irradiation of azirine 1b solely yields ylide
8, as the formation of vinylnitrene 10 in solution or ketene
imine 12 in matrixes was not observed. Thus, when 1b is
irradiated, it forms the S1 of 1b, which presumably undergoes
internal conversion to a vibrationally hot ground state to
overcome the energy barrier of the transition state to form ylide
8 (43 kcal/mol). Because the azirine moiety is in conjugation
with the benzene ring, selective excitation of only the phenyl
moiety is not possible. In addition, intersystem crossing in 1b
660
dx.doi.org/10.1021/jo402443w | J. Org. Chem. 2014, 79, 653−663

The Journal of Organic Chemistry
Article
5.4. Preparation of Azirines 1a and 1b. Synthesis of (2-Azido-
1-iodopropyl)benzene. (2-Azido-1-iodopropyl)benzene was prepared
following the procedure reported by Nair et al.54 To a mixture of trans-
propenylbenzene (2.0 g, 16.9 mmol), sodium azide (1.10 g, 16.9
mmol), and sodium iodide (2.50 g, 16.9 mmol) in methanol (25 mL)
at 0 °C was added a solution of ceric ammonium nitrate (19.5 g, 35.6
mmol) in methanol (100 mL) dropwise. When TLC showed that the
starting material was exhausted, saturated aqueous NaHSO3 (50 mL)
was added, and the resulting mixture was extracted with dichloro-
methane (3 × 125 mL). The combined organic extracts were washed
with distilled water (50 mL) and saturated brine (50 mL) and dried
over anhydrous MgSO4, and the solvent was removed under vacuum.
The residue was purified on a silica column with hexane and ethyl
acetate (9:1) as an eluent to yield (2-azido-1-iodopropyl)benzene (2.3
g, 7.9 mmol, 47% yield) as a slightly yellowish oil.
residue was subjected to silica gel chromatography and eluted with
hexane and ethyl acetate (92:8) to yield (1-azidoprop-1-en-1-
yl)benzene as a pale-yellow oil (1.79 g, 11.3 mmol, 72% yield).
1
IR (CDCl3): 2104, 1652, 1492, 1443, 1265, 763, 700 cm−1. H
NMR (CDCl3, 400 MHz): δ 1.71−1.72 (d, J = 4 Hz, 3H), 5.45−5.51
(q, J = 8 Hz, 1H), 7.31−7.43 (m, 5H). 13C NMR (CDCl3, 100 MHz):
δ 137.3, 133.2, 128.8, 128.6, 128.4, 112.1, 13.8. GC/MS (EI): m/z 131
(M+ − N2), 130, 104, 103 (100), 76, 64.
Photolysis of (2-Azidoprop-1-en-1-yl)benzene To Give 1a. A
solution of (Z)-(2-azidoprop-1-en-1-yl)benzene (450 mg, 2.2 mmol)
in chloroform (100 mL) was purged with argon for 15 min and
irradiated using a high-pressure mercury lamp through a Pyrex filter
(>310 nm) for 8 h. The solvent was removed under vacuum to obtain
1a as a dark-brown oil (359 mg, 2.1 mmol, 97% yield). The
spectroscopic characterization of 1a matched the previously published
data.55,57
1
IR (CDCl3): 2984, 2099, 1451, 1259 cm−1. H NMR (CDCl3, 400
MHz): δ 1.25−1.27 (d, J = 8 Hz, 3H), 1.54−1.55 (d, J = 4 Hz, 3H),
3.81−3.88 (m, 1H), 3.93−4.00 (m, 1H), 4.95−4.97 (d, J = 8 Hz, 1H),
5.00−5.02 (d, J = 8 Hz, 1H), 7.29−7.50 (m, 10H). 13C NMR (CDCl3,
100 MHz): δ 140.7, 140.4, 128.9, 128.7, 128.5, 128.4, 128.1, 63.7, 63.2,
37.4, 36.3, 19.8, 18.4. GC/MS (EI): m/z 217, 127, 117, 105, 91 (100),
77, 65.
IR (CDCl3): 3029, 1764, 1668, 1598, 1495, 1454, 1261, 913, 766
1
cm−1. H NMR (CDCl3, 400 MHz): δ 2.52 (s, 3H), 2.89 (s, 1H),
7.05−7.07 (d, J = 8 Hz, 2H), 7.21−7.31 (m, 3H). 13C NMR (CDCl3,
100 MHz): δ 164.5, 141.2, 128.3, 126.8, 125.6, 33.4, 12.9. GC/MS
(EI): m/z 131, 130 (100), 90, 89, 77, 65.
Photolysis of (1-Azidoprop-1-en-1-yl)benzene To Give 1b. A
solution of (1-azidoprop-1-en-1-yl)benzene (300 mg, 1.88 mmol) in
chloroform (60 mL) was purged with argon for 15 min and irradiated
using a high-pressure mercury lamp through a Pyrex filter (>310 nm)
for 10 h. GC/MS analysis of the reaction mixture showed some
remaining (1-azidoprop-1-enyl)benzene (5%) and the formation of 1b.
The solvent was removed under vacuum, and the resulting oil was
purified on a silica column with 10% ethyl acetate in hexane to obtain
1b as a pale-yellow oil (222 mg, 1.70 mmol, 90% yield). The spectra of
1b are in agreement with those already published.55
Synthesis of (2-Azidoprop-1-en-1-yl)benzene. To 2.3 g (7.9
mmol) of (2-azido-1-iodopropyl)benzene in dry diethyl ether (100
mL) in an ice bath was added potassium tert-butoxide (1.2 g, 10.3
mmol). The reaction mixture was stirred for 4 h at 0 °C and was then
washed twice with water (100 mL). The organic extract was dried over
magnesium sulfate, and the diethyl ether was removed under vacuum
at ambient temperature. The residue was subjected to chromatography
on silica and eluted with hexane and ethyl acetate (9:1) to yield (Z)-
(2-azidoprop-1-en-1-yl)benzene (0.91 g, 5.7 mmol, 70% yield) as a
colorless oil and (E)-(2-azidoprop-1-enyl)benzene (0.13 g, 0.8 mmol,
10% yield) as a colorless oil. The spectra of (2-azidoprop-1-en-
yl)benzene are in agreement with those already published.55
IR (CDCl3): 2121, 2099, 2055, 1646, 1597, 1572, 1492, 1265 cm−1.
1H NMR (CDCl3, 400 MHz): (Z isomer) δ 2.20 (s, 3H), 5.62 (br s,
1H), 7.56−7.15 (m, 5H); (E isomer) δ 2.06 (s, 3H), 6.18 (br s, 1H),
7.35−7.17 (m, 5H). 13C NMR (CDCl3, 100 MHz): (Z isomer) δ
135.3, 131.1, 128.5, 128.2, 126.6, 115.7, 19.8; (E isomer) δ 136.1,
135.4, 128.6, 128.4, 126.6, 115.9, 15.6.
IR (CDCl3): 1737, 1597, 1578, 1487, 1451, 1320, 936, 760, 689
1
cm−1. H NMR (CDCl3, 400 MHz): δ 1.36−1.37 (d, J = 4 Hz, 3H),
2.29−2.33 (q, J = 4 Hz, 1H), 7.54−7.57 (m, 3H), 7.85−7.87 (m, 2H).
13C NMR (CDCl3, 100 MHz): δ 172.5, 132.7, 129.3, 129.1, 125.7,
27.5, 18.9. GC/MS (EI): m/z 131, 130, 104, 103 (100), 77, 76.
5.5. Photolysis of Azirines 1a and 1b. Photolysis of 1a in
Argon-Saturated Acetonitrile. Azirine 1a (100 mg, 0.76 mmol) was
dissolved in acetonitrile, and argon was bubbled through the solution
for 15 min. The solution was irradiated using a high-pressure mercury
lamp through a Pyrex filter (>310 nm) for 20 h. GC/MS analysis of
the irradiated solution showed no new products.
Synthesis of (1-Azido-2-iodopropyl)benzene. Sodium azide (3.58
g, 22 mmol) in acetonitrile (50 mL) was placed in a 100 mL three-
neck round-bottom flask fitted with two 50 mL pressure-equalizing
dropping funnels. The flask was cooled in an ice bath, and iodine
monochloride (2.75 g, 42.4 mmol) was added dropwise from one of
the dropping funnels over a period of 15 min while the reaction
mixture was stirred with a magnetic stirrer. The solution was stirred for
an additional 10 min, and then trans-propenylbenzene (2.0 g, 16.9
mmol) was added from the other dropping funnel over a 15 min
period. The resulting reaction mixture was stirred for 12 h at ambient
temperature, poured into water (50 mL), and extracted with three
portions of diethyl ether (50 mL). The combined organic extracts
were washed with 5% aqueous sodium thiosulfate (100 mL) and water
(100 mL). The ether extract was dried over magnesium sulfate, and
the solvent was removed under vacuum at ambient temperature to
yield the product as a pale-yellow oil (4.5 g) of sufficient purity to be
used for the next step.
Photolysis of 1a in Oxygen-Saturated Acetonitrile. Azirine 1a
(100 mg, 0.76 mmol) was dissolved in acetonitrile, and oxygen was
bubbled through the solution for 15 min. The solution was irradiated
using a high-pressure mercury lamp through a Pyrex filter (>310 nm)
1
for 20 h. The H NMR spectra of the reaction mixture showed the
formation of benzaldehyde (2) (11% yield) as the major product along
with remaining starting material.
Photolysis of 1a in Argon-Saturated Bromoform. Azirine 1a (100
mg, 0.76 mmol) was dissolved in bromoform, and argon was bubbled
through the solution for 15 min. The solution was irradiated using a
high-pressure mercury lamp through a Pyrex filter (>310 nm) for 6 h.
GC/MS analysis of the reaction mixture showed trace amounts of 1-
phenylpropan-2-one (3) and 1-bromo-1-phenylpropan-2-one (4)
(14%). The formation of 3 and 4 was verified by injecting authentic
samples into the GC/MS instrument.
Photolysis of 1a in Oxygen-Saturated Bromoform. Azirine 1a
(100 mg, 0.76 mmol) was dissolved in bromoform, and oxygen was
bubbled through the solution for 15 min. The solution was irradiated
using a high-pressure mercury lamp through a Pyrex filter (>310 nm)
for 15 h. GC/MS analysis of the reaction mixture showed the
formation of 2 (10%), 3 (9%), and 4 (81%).
1
IR (CDCl3): 2983, 2099, 1492, 1451, 1259, 752, 700 cm−1. H
NMR (CDCl3, 400 MHz): δ 1.87−1.88 (d, J = 4 Hz, 3H), 4.31−4.38
(m, 1H), 4.75−4.77 (d, J = 8 Hz, 1H), 7.32−7.43 (m, 5H). 13C NMR
(CDCl3, 100 MHz): δ 137.4, 128.82, 128.75, 127.4, 72.7, 28.8, 23.5.
GC/MS (EI): m/z 287, 245, 155 (100), 127, 117, 105, 91, 77, 65.
Synthesis of (1-Azidoprop-1-en-1-yl)benzene. (1-Azidoprop-1-en-
1-yl)benzene was synthesized following the procedure of Hassner and
Fowler.56 A solution of (1-azido-2-iodopropyl)benzene (4.5 g, 15.7
mmol) in dry diethyl ether (100 mL) was cooled in an ice bath, and to
it was added potassium tert-butoxide (2.3 g, 20.4 mmol). The reaction
was stirred for 4 h at 0 °C and washed with two portions of water (100
mL). The organic layer was dried over magnesium sulfate, and the
solvent was removed under vacuum at ambient temperature. The
Photolysis of 1b in Argon-Saturated Acetonitrile. Azirine 1b (100
mg, 0.76 mmol) was dissolved in acetonitrile, and argon was bubbled
through the solution for 15 min. The solution was irradiated using a
high-pressure mercury lamp through a Pyrex filter (>310 nm) for 6 h.
1
The solvent was then removed under vacuum to yield a crude oil. H
NMR spectroscopy of the reaction mixture showed two significant
components, which were assigned as exo-2,6-dimethyl-4,5-diphenyl-
661
dx.doi.org/10.1021/jo402443w | J. Org. Chem. 2014, 79, 653−663

The Journal of Organic Chemistry
Article
1,3-diazabicyclo[3.1.0]hex-3-ene (6) and endo-2,6-dimethyl-4,5-di-
(10) Murthy, R. S.; Muthukrishnan, S.; Rajam, S.; Mandel, S. M.;
Ault, B. S.; Gudmundsdottir, A. D. J. Photochem. Photobiol., A 2009,
201, 157.
(11) Gamage, D. W.; Li, Q.; Ranaweera, R. A. A. U.; Sarkar, S. K.;
Weragoda, G. K.; Carr, P. L.; Gudmundsdottir, A. D. J. Org. Chem.
2013, 78, 11349.
1
phenyl-l,3-diazabicyclo-[3.1.0]hex-3-ene (7) on the basis of their H
NMR spectra, which matched those reported previously.27
1
6: H NMR (CDCl3, 400 MHz): δ 1.25 (d, J = 4 Hz, 3H), 1.48−
1.50 (d, J = 8 Hz, 3H), 2.04−2.08 (q, J = 4 Hz, 1H), 5.59−5.64 (q, J =
1
8 Hz, 1H), 7.30−7.50 (m, 10H). 7: H NMR (CDCl3, 400 MHz): δ
1.24 (d, J = 4 Hz, 3H), 1.52 (d, J = 8 Hz, 3H), 2.07 (q, J = 4 Hz, 1H),
5.0 (q, J = 8 Hz, 1H), 7.30−7.50 (m, 10H).
(12) 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.
Photolysis of 1b in Oxygen-Saturated Acetonitrile. Azirine 1b
(100 mg, 0.76 mmol) was dissolved in acetonitrile, and oxygen was
bubbled through the solution for 15 min. The solution was irradiated
using a high-pressure mercury lamp through a Pyrex filter (>310 nm)
for 6 h. The solvent was then removed under vacuum to yield a crude
oil. 1H NMR spectroscopy showed two products, which were
identified as 6 and 7.
(13) Rajam, S.; Murthy, R. S.; Jadhav, A. V.; Li, Q.; Keller, C.; Carra,
C.; Pace, T. C. S.; Bohne, C.; Ault, B. S.; Gudmundsdottir, A. D. J. Org.
Chem. 2011, 76, 9934.
(14) Singh, B.; Zweig, A.; Gallivan, J. B. J. Am. Chem. Soc. 1972, 94,
1199.
(15) Inui, H.; Murata, S. J. Am. Chem. Soc. 2005, 127, 2628.
(16) Murata, S.; Tomioka, H. Chem. Lett. 1992, 57.
(17) Nunes, C. M.; Reva, I.; Fausto, R. J. Org. Chem. 2013, 78, 10657.
(18) Gritsan, N. P.; Pritchina, E. S. J. Inf. Rec. Mater. 1989, 17, 391.
(19) Pritchina, E. A.; Gritsan, N. P. J. Photochem. Photobiol., A 1988,
43, 165.
(20) Barcus, R. L.; Hadel, L. M.; Johnston, L. J.; Platz, M. S.; Savino,
T. G.; Scaiano, J. C. J. Am. Chem. Soc. 1986, 108, 3928.
(21) Mueller, F.; Mattay, J. Chem. Ber. 1993, 126, 543.
(22) Orton, E.; Collins, S. T.; Pimentel, G. C. J. Phys. Chem. 1986, 90,
6139.
(23) Albrecht, E.; Mattay, J.; Steenken, S. J. Am. Chem. Soc. 1997,
119, 11605.
(24) Gakis, N.; Maerky, M.; Hansen, H. J.; Schmid, H. Helv. Chim.
Acta 1972, 55, 748.
(25) Bayes, K. D.; Toohey, D. W.; Friedl, R. R.; Sander, S. P. J.
Geophys. Res.: Atmos. 2003, 108, ACH4/1.
(26) Padwa, A.; Smolanoff, J.; Wetmore, S. I., Jr. J. Chem. Soc., Chem.
Commun. 1972, 409.
(27) Padwa, A.; Smolanoff, J.; Wetmore, S. I., Jr. J. Org. Chem. 1973,
38, 1333.
Photolysis of 1b in CO2-Saturated Acetonitrile. Azirine 1b (100
mg, 0.76 mmol) was dissolved in acetonitrile. CO2 was bubbled
through the solution while it was irradiated using a high-pressure
mercury lamp through a Pyrex filter (>310 nm) for 4 h. GC/MS was
used to monitor the reaction progress. When GC/MS showed that the
starting material was depleted, the solvent was removed under vacuum
to produce 2-methyl-4-phenyloxazol-5(2H)-one (9) in 95% yield. The
spectroscopic characterization matched the published data.28
IR (CDCl3): 2985, 1179, 1618, 1302, 1149, 1066, 928, 808, 747,
1
686 cm−1. H NMR (CDCl3, 400 MHz): δ 1.66−1.68 (d, J = 8 Hz,
3H), 6.10−6.15 (q, J = 8 Hz, 1H), 7.48−7.57 (m, 3H), 8.37−8.39 (d, J
= 8 Hz, 2H).
ASSOCIATED CONTENT
■
S
* Supporting Information
Cartesian coordinates and energies of 1a, 1b, 5, 8, and 10−16
1
and H and 13C NMR spectra of 1a, 1b, 2, 3, 4, 6, 7, and 9.
This material is available free of charge via the Internet at
http://pubs.acs.org.
(28) Padwa, A.; Akiba, M.; Cohen, L. A.; MacDonald, J. G. J. Org.
Chem. 1983, 48, 695.
AUTHOR INFORMATION
(29) Jackson, B.; Gakis, N.; Marky, M.; Hansen, H. J.; von
̈
■
Philipsborn, W.; Schmid, H. Helv. Chim. Acta 1972, 55, 916.
Corresponding Author
(30) Padwa, A.; Wetmore, S. I. J. Am. Chem. Soc. 1974, 96, 2414.
(31) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.;
Robb, M. A.; Cheeseman, J. R.; Scalmani, G.; Barone, V.; Mennucci,
B.; Petersson, G. A.; Nakatsuji, H.; Caricato, M.; Li, X.; Hratchian, H.
P.; Izmaylov, A. F.; Bloino, J.; Zheng, G.; Sonnenberg, J. L.; Hada, M.;
Ehara, M.; Toyota, K.; Fukuda, R.; Hasegawa, J.; Ishida, M.; Nakajima,
T.; Honda, Y.; Kitao, O.; Nakai, H.; Vreven, T.; Montgomery, J. A., Jr.;
Peralta, J. E.; Ogliaro, F.; Bearpark, M.; Heyd, J. J.; Brothers, E.; Kudin,
K. N.; Staroverov, V. N.; Kobayashi, R.; Normand, J.; Raghavachari, K.;
Rendell, A.; Burant, J. C.; Iyengar, S. S.; Tomasi, J.; Cossi, M.; Rega,
N.; Millam, J. M.; Klene, M.; Knox, J. E.; Cross, J. B.; Bakken, V.;
Adamo, C.; Jaramillo, J.; Gomperts, R.; Stratmann, R. E.; Yazyev, O.;
Austin, A. J.; Cammi, R.; Pomelli, C.; Ochterski, J. W.; Martin, R. L.;
Morokuma, K.; Zakrzewski, V. G.; Voth, G. A.; Salvador, P.;
*E-mail: Anna.Gudmundsdottir@uc.edu.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
We thank the National Science Foundation (CHE-1057481)
and the Ohio Supercomputer Center for supporting this work.
REFERENCES
■
(1) Platz, M. S. Nitrenes. In Reactive Intermediate Chemistry; Moss, R.
A., Platz, M. S., Jones, M., Jr., Eds.; John Wiley & Sons: Hoboken, NJ,
2004; Chapter 11.
̈
Dannenberg, J. J.; Dapprich, S.; Daniels, A. D.; Farkas, O.;
(2) Reiser, A.; Leyshon, L. J. J. Am. Chem. Soc. 1971, 93, 4051.
(3) Schrock, A. K.; Schuster, G. B. J. Am. Chem. Soc. 1984, 106, 5228.
(4) Muthukrishnan, S.; Ranaweera, R. A. A. U.; Gudmundsdottir, A.
D. In Nitrenes and Nitrenium Ions; Falvey, D. E., Gudmundsdottir, A.
D., Eds.; John Wiley & Sons: Hoboken, NJ, 2013; Vol. 6.
(5) Singh, P. N. D.; Mandel, S. M.; Sankaranarayanan, J.;
Muthukrishnan, S.; Chang, M.; Robinson, R. M.; Lahti, P. M.; Ault,
B. S.; Gudmundsdottir, A. D. J. Am. Chem. Soc. 2007, 129, 16263.
(6) Sankaranarayanan, J.; Bort, L. N.; Mandel, S. M.; Chen, P.;
Krause, J. A.; Brooks, E. E.; Tsang, P.; Gudmundsdottir, A. D. Org.
Lett. 2008, 10, 937.
(7) Sankaranarayanan, J.; Rajam, S.; Hadad, C. M.; Gudmundsdottir,
A. D. J. Phys. Org. Chem. 2010, 23, 370.
(8) Lwowski, W.; Woerner, F. P. J. Am. Chem. Soc. 1965, 87, 5491.
(9) Lwowski, W.; Mattingly, T. W., Jr. J. Am. Chem. Soc. 1965, 87,
1947.
Foresman, J. B.; Ortiz, J. V.; Cioslowski, J.; Fox, D. J. Gaussian 09,
revision A.1; Gaussian, Inc.: Wallingford, CT, 2009.
(32) Becke, A. D. J. Chem. Phys. 1993, 98, 5648.
(33) Lee, C.; Yang, W.; Parr, R. G. Phys. Rev. B 1988, 37, 785.
(34) Muthukrishnan, S.; Mandel, S. M.; Hackett, J. C.; Singh, P. N.
D.; Hadad, C. M.; Krause, J. A.; Gudmundsdottir, A. D. J. Org. Chem.
2007, 72, 2757.
(35) Parasuk, V.; Cramer, C. J. Chem. Phys. Lett. 1996, 260, 7.
(36) Wenthold, P. G. J. Org. Chem. 2012, 77, 208.
(37) Hossain, E.; Wenthold, P. G. Comput. Theor. Chem. 2013, 1020,
180.
(38) Clark, W. D. K.; Steel, C. J. Am. Chem. Soc. 1971, 93, 6347.
(39) Foresman, J. B.; Frisch, Æ. Exploring Chemistry with Electronic
Structure Methods, 2nd ed.; Gaussian, Inc.: Pittsburgh, PA, 1996.
(40) Inui, H.; Murata, S. Chem. Lett. 2001, 30, 832.
662
dx.doi.org/10.1021/jo402443w | J. Org. Chem. 2014, 79, 653−663

The Journal of Organic Chemistry
Article
́
(41) Kaczor, A.; Gomez-Zavaglia, A.; Cardoso, A. L.; Pinho e Melo,
T. M. V. D.; Fausto, R. J. Phys. Chem. A 2006, 110, 10742.
(42) Inui, H.; Murata, S. Chem. Commun. 2001, 1036.
(43) Gonzalez, C.; Schlegel, H. B. J. Chem. Phys. 1989, 90, 2154.
(44) Gonzalez, C.; Schlegel, H. B. J. Phys. Chem. 1990, 94, 5523.
(45) Bauernschmitt, R.; Ahlrichs, R. Chem. Phys. Lett. 1996, 256, 454.
(46) Stratmann, R. E.; Scuseria, G. E.; Frisch, M. J. J. Chem. Phys.
1998, 109, 8218.
(47) Foresman, J. B.; Head-Gordon, M.; Pople, J. A.; Frisch, M. J. J.
Phys. Chem. 1992, 96, 135.
(48) Cances, E.; Mennucci, B. J. Chem. Phys. 2001, 114, 4744.
(49) Cramer, C. J.; Truhlar, D. G. Chem. Rev. 1999, 99, 2161.
(50) Mennucci, B.; Cances, E.; Tomasi, J. J. Phys. Chem. B 1997, 101,
10506.
(51) Tomasi, J.; Mennucci, B.; Cammi, R. Chem. Rev. 2005, 105,
2999.
(52) Muthukrishnan, S.; Sankaranarayanan, J.; Klima, R. F.; Pace, T.
C. S.; Bohne, C.; Gudmundsdottir, A. D. Org. Lett. 2009, 11, 2345.
(53) Ault, B. S. J. Am. Chem. Soc. 1978, 100, 2426.
(54) Nair, V.; George, T. G.; Sheeba, V.; Augustine, A.; Balagopal, L.;
Nair, L. G. Synlett 2000, 1597.
(55) Bader, H.; Hansen, H. J. Helv. Chim. Acta 1978, 61, 286.
(56) Hassner, A.; Fowler, F. W. Tetrahedron Lett. 1967, 8, 1545.
(57) Singh, P. N. D.; Carter, C. L.; Gudmundsdottir, A. D.
Tetrahedron Lett. 2003, 44, 6763.
663
dx.doi.org/10.1021/jo402443w | J. Org. Chem. 2014, 79, 653−663