Photochemistry of fluorinated 4-iodophenylnitrenes: Matrix isolation and spectroscopic characterization of phenylnitrene-4-yls

Article abstract of DOI:10.1021/jo901145h

(Chemical Equation Presented) The photochemistry of a series of fluorinated p-iodophenyl azides 2 has been investigated using matrix isolation IR and EPR spectroscopy. In all cases, the corresponding phenylnitrenes 1 were formed as primary photoproducts. Further irradiation of the nitrenes 1 resulted in the formation of azirines 3, ketenimines 4, and nitreno radicals 5. The yield of 5 depends on the number of ortho fluorine substituents: with two ortho fluorine atoms the highest yield is observed, whereas without fluorine atoms the yield is too low for IR spectroscopic detection. The interconversion between the isomers 1, 3, and 4 proved to be rather complex. If the fluorine atoms are distributed unsymmetrically, two isomers of azirines 3 and ketenimines 4 can be formed. The yields of these isomers depend critically on the irradiation conditions. 2009 American Chemical Society.

Full text of DOI:10.1021/jo901145h

pubs.acs.org/joc
Photochemistry of Fluorinated 4-Iodophenylnitrenes: Matrix Isolation
and Spectroscopic Characterization of Phenylnitrene-4-yls
Dirk Grote and Wolfram Sander*
€
Lehrstuhl fu€r Organische Chemie II der Ruhr-Universitat, D-44780 Bochum, Germany
wolfram.sander@rub.de
Received May 29, 2009
The photochemistry of a series of fluorinated p-iodophenyl azides 2 has been investigated using
matrix isolation IR and EPR spectroscopy. In all cases, the corresponding phenylnitrenes 1 were
formed as primary photoproducts. Further irradiation of the nitrenes 1 resulted in the formation of
azirines 3, ketenimines 4, and nitreno radicals 5. The yield of 5 depends on the number of ortho
fluorine substituents: with two ortho fluorine atoms the highest yield is observed, whereas without
fluorine atoms the yield is too low for IR spectroscopic detection. The interconversion between the
isomers 1, 3, and 4 proved to be rather complex. If the fluorine atoms are distributed unsymme-
trically, two isomers of azirines 3 and ketenimines 4 can be formed. The yields of these isomers depend
critically on the irradiation conditions.
Introduction
years, various arylnitrenes were investigated using matrix
isolation spectroscopy, time-resolved UV-vis spectroscopy,
and quantum chemical calculations.5-11
The mechanism of these nitrene rearrangements has been
subject to controversial discussions for years, but due to
extensive experimental and theoretical calculations the me-
chanism shown in Scheme 1 for the parent phenylnitrene 1a
Phenylnitrenes 1 are reactive intermediates with triplet
ground states that have been extensively studied in recent
years using matrix isolation spectroscopy and laser flash
photolysis.1 Phenylnitrenes 1 are conveniently generated by
photolysis of the corresponding phenyl azides 2 which under
loss of molecular nitrogen yield the singlet nitrenes S-1 as the
primary photoproducts. The fate of S-1 with lifetimes in the
order of several ns depends on the reaction conditions: they
are either trapped by suitable reagents, deactivate to the
ground state triplet nitrenes T-1 via intersystem crossing, or
rearrange to azirines 3 or ketenimines 4.2-4 During the last
(5) Maltsev, A.; Bally, T.; Tsao, M.-L.; Platz, M. S.; Kuhn, A.; Vosswinkel,
M.; Wentrup, C. J. Am. Chem. Soc. 2004, 126, 237–249.
(6) Kvaskoff, D.; Bednarek, P.; George, L.; Pankajakshan, S.; Wentrup,
C. J. Org. Chem. 2005, 70, 7947–7955.
(7) Vosswinkel, M.; Luerssen, H.; Kvaskoff, D.; Wentrup, C. J. Org.
Chem. 2009, 74, 1171–1178.
(8) Kvaskoff, D.; Mitschke, U.; Addicott, C.; Finnerty, J.; Bednarek, P.;
Wentrup, C. Aust. J. Chem. 2009, 62, 275–286.
(1) Gritsan, N. P.; Platz, M. S. Chem. Rev. 2006, 106, 3844–3867.
(2) Hayes, J. C.; Sheridan, R. S. J. Am. Chem. Soc. 1990, 112, 5879–5881.
(3) Borden, W. T.; Gritsan, N. P.; Hadad, C. M.; Karney, W. L.;
Kemnitz, C. R.; Platz, M. S. Acc. Chem. Res. 2000, 33, 765–771.
(4) Platz, M. S. In Reactive Intermediate Chemistry; Moss, R. A., Platz,
M. S., Jones, M., Jr., Eds.; Wiley: New York, 2004.
(9) Kvaskoff, D.; Bednarek, P.; George, L.; Waich, K.; Wentrup, C.
J. Org. Chem. 2006, 71, 4049–4058.
(10) Addicott, C.; Wong, M. W.; Wentrup, C. J. Org. Chem. 2002, 67,
8538–8546.
(11) Kuhn, A.; Vosswinkel, M.; Wentrup, C. J. Org. Chem. 2002, 67,
9023–9030.
7370 J. Org. Chem. 2009, 74, 7370–7382
Published on Web 09/04/2009
DOI: 10.1021/jo901145h
r
2009 American Chemical Society

Grote and Sander
JOCArticle
SCHEME 1. Rearrangement of Phenylnitrene 1a
CHART 1. Phenyl Azides 2 Investigated in This Study
is now generally accepted. Irradiation of phenyl azide 2a,
matrix-isolated in argon at cryogenic temperatures, results in
the formation of 1a, 3a, and 4a in photostationary equilibria.
According to ab initio calculations, singlet nitrene S-1a first
rearranges to azirine 3a via addition of the nitrogen atom to
one of the neighboring aromatic C-C bonds. This rearran-
gement is slightly endothermic by 4 kcal/mol. Azirine 3a
subsequently opens up to the cyclic ketenimine 4a which is
1.6 kcal/mol more stable than S-1a.12 However, with a single-
t-triplet splitting ΔEST=18 kcal/mol the triplet phenylnitrene
T-1a is the most stable of these isomers.
results in a number of close lying “low spin” (doublet) states
and a “high spin” (quartet) state.
Substituents in the ortho position, especially by fluor-
ine atoms as in 1b and 1c, lead to a substantial increase in
the activation barrier for the rearrangement to azirines 3
and thus to higher yields of singlet nitrene trapping
products in solution.1 The photochemistry of matrix-
isolated 2,6-difluorophenylnitrene T-1b was investigated
by several authors13-15 but only recently could it be
shown that indeed both 3b and 4b are formed in the
photochemical rearrangement. The question why ortho
fluorine substitution raises the barrier toward ring
expansion of phenylnitrene has also been subject to a
theoretical study.16 Thus, phenyl azides bearing fluorine
substituents in the ortho positions are the most suitable
precursors for the generation of kinetically stabilized
phenylnitrenes.
According to high-level ab initio calculations, in phenyl-
nitrene-4-yl 5a the quartet state is energetically preferred by
5.3 kcal/mol.17 These calculations also predict a quartet
ground state for phenylnitrene-2-yl 7, whereas phenylni-
trene-3-yl 6 is predicted to exhibit a low-spin doublet state
as ground state.18
The only derivative of 5 that could be isolated and spectro-
scopically characterized, so far, is the perfluorinated triradi-
cal 5f (Chart 1).19 The IR spectrum of matrix-isolated 5f is in
good agreement with the spectrum calculated for the quartet
state Q-5f but not with the lowest lying doublet state D-5f.
The assignment of a quartet state for 5f was later confirmed
by EPR spectroscopy.20 Here, we describe the matrix isola-
tion and spectroscopic characterization (IR and EPR) of a
series of new nitreno radicals 5.
Results
Introduction of an additional radical center in the phenyl
ring leads to the interesting phenylnitrene-yls 5-7. These
nitreno radicals are characterized best as σ, σ, π triradicals
with two unpaired σ electrons located at the nitrogen atom
and the radical center of the phenyl ring, respectively, and
the third unpaired electron being delocalized in the π
system. The interaction between these three unpaired elec-
trons
The photochemistry of the eight p-iodophenyl azides 2d-l
has been investigated using matrix isolation spectroscopy
with IR and EPR detection (Chart 1). In all cases, the
primary photoproducts isolated in high yields were the
corresponding triplet nitrenes T-1. The IR spectra of the
nitrenes are in good agreement with DFT calculations at the
UB3LYP/6-311G(d,p) level of theory. The EPR spectra
show strong triplet signals with zfs parameters D and E as
expected for phenylnitrenes (Table 1).
(12) Karney, W. L.; Borden, W. T. J. Am. Chem. Soc. 1997, 119, 1378–
1387.
(13) Morawietz, J.; Sander, W. J. Org. Chem. 1996, 61, 4351–4354.
(14) Carra, C.; Nussbaum, R.; Bally, T. ChemPhysChem 2006, 7, 1268–
1275.
(17) Bettinger, H. F.; Sander, W. J. Am. Chem. Soc. 2003, 125, 9726–9733.
(18) Sander, W.; Winkler, M.; Cakir, B.; Grote, D.; Bettinger, H. F.
J. Org. Chem. 2007, 72, 715–724.
(15) Mandel, S.; Liu, J.; Hadad Christopher, M.; Platz Matthew, S.
J. Phys. Chem. A 2005, 109, 2816–21.
(19) Wenk, H. H.; Sander, W. Angew. Chem., Int. Ed. 2002, 41, 2742–
2745.
(16) Karney, W. L.; Borden, W. T. J. Am. Chem. Soc. 1997, 119, 3347–
3350.
(20) Sander, W.; Grote, D.; Kossmann, S.; Neese, F. J. Am. Chem. Soc.
2008, 130, 4396–4403.
J. Org. Chem. Vol. 74, No. 19, 2009 7371

Grote and Sander
JOCArticle
TABLE 1. zfs Parameters of the Triplet and Quartet Species (giso = 2.003)
T-1
Q-5
fluorine distribution
|D/hc| (cm-1
)
|E/hc| (cm-1
)
nitrogen spin densitya
|D/hc| (cm-1
)
|E/hc| (cm-1
)
2,3,6- (e)
2,6- (d)
2,3,5,6- (f)
2,3,5- (g)
2,5- (h)
2,3- (i)
0.981
0.925
1.103
1.073
0.954
0.953
0.900
1.053
0.0071
0.0086
0.0120
0.0091
0.0120
0.0032
0.0036
0.0103
1.5725
1.5694
1.5762
1.5792
1.5689
1.5782
1.5712
1.5852
0.289
0.291
0.285
0.283
0.278
0.287
0.285
0.041
0.040
0.043
0.042
0.040
0.040
0.040
2- (k)
3,5- (l)
aUB3LYP/6-311G(d,p).
SCHEME 2. Photochemistry of 2d
SCHEME 3. Photochemistry of 2e
FIGURE 1. IR spectra showing the 320 nm photochemistry of
azide 2d in argon at 4 K. (a) Difference IR spectrum: bands pointing
downward disappear during irradiation and are assigned to 2d;
bands pointing upward appear and are assigned to the photopro-
ducts T-1d (major product), 3d, and 4d. (b) Calculated IR spectrum
of T-1d (c) Calculated IR spectrum of 3d. (d) Calculated IR
spectrum of 4d. All calculations at the (U)B3LYP/6-311G(d,p) level
of theory.
be presented in detail, while for the other precursors only
deviations from the general scheme will be discussed.
Photochemistry of 2,6-Difluoro-4-iodophenyl Azide 2d. The
photochemistry of 2d, 2e, and 2f (which was described
previously)19,20 with two fluorine atoms in the ortho posi-
tions is very similar, and here we focus on 2d as a represen-
tative sample (Scheme 2). The photochemistry of 2e is
complicated by the asymmetric distribution of the fluorine
substituents which increases the number of the isomers
formed during photolysis (Scheme 3).
UV irradiation (λ=254 or >320 nm) of phenyl azide 2d in
solid argon or neon at 4 K results in a rapid decrease of all IR
absorptions of the azide and formation of a new main
product with very strong IR bands at 1562.9, 1476.2,
1434.2, and 1013.4 cm-1 assigned to nitrene T-1d. DFT
calculations (UB3LYP/6-311G(d,p)) predict the strongest
IR absorptions of T-1d at 1589.6, 1502.0, 1458.1, and 1027.4
cm-1 (unscaled), in very good agreement with the experiment
(Figure 1). The EPR spectrum of T-1d is characteristic of a
triplet phenylnitrene with zero field splitting (zfs) parameters
Irradiation of the nitrenes T-1 produced complex product
mixtures with a composition much depending on the light
source (UV or visible light, narrow or broadband
irradiation), the matrix temperature (between 3 and 15 K),
and the matrix host (neon or argon). Three basic types of
substitution patterns in the precursor molecules will be
discussed separately: (i) phenyl azides bearing two fluorine
atoms in the ortho positions 2d-f, (ii) one fluorine and one
hydrogen atom 2g-k, and (iii) two hydrogen atoms 2l. For
each of these classes of phenyl azides one representative will
|D/hc|=0.925 cm-1 and |E/hc|=0.0086 cm-1 20
.
A second product that is formed during the UV irradiation
is didehydroazepine (ketenimine) 4d with a characteristic
strong IR absorption at 1840.7 cm-1, assigned to the
NdCdC str. vibration (Table 2). Absorptions in this range
7372 J. Org. Chem. Vol. 74, No. 19, 2009

Grote and Sander
JOCArticle
TABLE 2. IR Spectroscopic Data of 4 and 40, Matrix-Isolated in Argon
at 4 K
TABLE 3. IR Spectroscopic Data of Azirines 3 and 30, Matrix-Isolated
in Argon at 4 K
aArgon, 4 K. bB3LYP/6-311G(d,p), unscaled. Relative intensities
d
based on the strongest absorption. cOverlap with other bands. Neon,
4 K.
aArgon, 4 K; bB3LYP/6-311G(d,p), unscaled. Relative intensity
based on the strongest absorption.
calculation of the NdCdC str vibration is also observed in
other ketenimines.
Azirine 3d is also found in the product mixture after
irradiation of 2d upon λ > 320 nm. The azirine was identified
by comparison of its IR spectrum with that of 3b and other
related azirines and with results from DFT calculations. A
characteristic absorption of 3dis the CdN stretching vibration
at 1680.8 cm-1 (Table 3). Irradiation of a product mixture
formed on irradiation of precursor 2d with visible light (λ >
420 nm) results in a decrease of 4d and T-1d and an increase of
3d (Figure 2). A careful analysis of the relative amounts of 1d,
3d, and 4d reveals that on broadband UV irradiation (λ > 320
nm) these products are formed in photostationary equilibria
with 1d and 4d being the major products. However, the
photochemistry is much more selective with 254 nm; here only
1d and 4d but no 3d are formed (Scheme 2).
Prolonged 254 nm irradiation (argon, 4 K) of the product
mixture produces a further compound with IR bands at
1405.5, 1326.2, 1185.3, 997.9, and 814.3 cm-1 (Table 4)
which is in better agreement with calculations for quartet
triradical Q-5d than for doublet D-5d. However, the differ-
ences between the calculated spectra of Q-5d and D-5d are
not very pronounced (Figure 3), and the assignment is
therefore based on the EPR spectra. This is different to the
tetrafluorinated system 5f where the differences in the
IR spectra calculated for the quartet state and the doublet
state are large enough for an assignment based on the IR
spectra.19
FIGURE 2. IR spectra showing the photochemical interconversion
between T1-d, 3d, and 4d during 420 nm irradiation in argon at 4 K.
(a) Calculated IR spectrum of 4d. (b) Calculated IR spectrum of
T-1d. (c) Difference IR spectrum: bands pointing upward are
appearing and assigned to 3d; bands pointing downward disappear
and are assigned to T1-d and 4d. (d) Calculated IR spectrum of 3d.
All calculations at the (U)B3LYP/6-311G(d,p) level of theory.
have been reported for a number of didehydroazepines; e.g.,
4a shows a strong band at 1895 cm-1 and 4b at 1830 cm-1 14
.
DFT calculations predict the NdCdC str vibration of 4d at
1929.1 cm-1, which is considerably blue-shifted compared to
the experimental value, while the other vibrations of 4d are in
good agreement with the experiment. The large error for the
J. Org. Chem. Vol. 74, No. 19, 2009 7373

Grote and Sander
JOCArticle
TABLE 4. IR Spectroscopic Data of 5d, Matrix-Isolated in Argon at
4 K, Compared to Calculations for the Quartet and Doublet State
Ar, 4 K
Q-5d
D-5d
ν~exp/cm-1
ν~calcd/cm-1 a
ν~calcd/cm-1a
b
a,b
a,b
mode (Irel,exp
)
symmetry (Irel,calcd
)
symmetry (Irel,calcd
)
13
14
15
16
17
814.3 (0.14)
b1
a1
a2
b2
a1
b2
a1
b2
a1
a1
b2
b2
a1
814.8 (0.30)
820.3 (0.02)
835.1 (0.00)
1010.1 (0.51)
1048.7 (0.03)
1186.9 (0.18)
1253.7 (0.03)
1287.6 (0.04)
1327.3 (0.17)
1360.0 (0.00)
1426.9 (1.00)
1534.9 (0.20)
1541.1 (0.66)
b1
a1
a2
b2
a1
b2
a1
b2
a1
a1
b2
a1
b2
811.0 (0.26)
820.9 (0.02)
847.7 (0.00)
1014.0 (0.43)
1059.3 (0.02)
1185.2 (0.18)
1246.8 (0.00)
1267.9 (0.00)
1305.0 (0.19)
1371.4 (0.00)
1415.8 (1.00)
1553.6 (0.56)
1554.7 (0.22)
997.9 (0.84)
18 1185.3 (0.35)
19
20
21 1326.2 (0.32)
22
23 1405.5 (1.00)
24
25
FIGURE 4. (a) EPR spectrum of Q-5d (argon, 4 K) formed after
308 nm irradiation of 2d. (b) Simulated spectrum using S=3/2, |D/
hc| = 0.291 cm-1, and |E/hc| = 0.040 cm-1. The intense signal
between 6500 and 7000 G is assigned to T-1d.
aUB3LYP/6-311G(d,p), unscaled. bRelative intensity based on the
strongest absorption.
FIGURE 5. IR spectra showing the 254 nm photochemistry of
nitrene T-1e in argon at 4 K after 47 min of irradiation.
(a) Calculated IR spectrum of nitrene T-1e. (b) Difference IR
spectrum: bands of T-1e pointing downward disappear; bands
pointing upward assigned to Q-5e and 40e appear during irradiation.
(c) Calculated IR spectrum of Q-5e. (d) Calculated IR spectrum of
40e. The second isomer 4e is not observed. All calculations at the
(U)B3LYP/6-311G(d,p) level of theory.
FIGURE 3. IR spectra showing the 254 nm photochemistry of
nitrene T-1d in argon at 4 K. (a) Calculated IR spectrum of T-1d.
(b) Difference spectrum: bands pointing downward disappear;
bands pointing upward appear during irradiation. Bands marked
with b assigned to 4d. (c) Calculated IR spectrum of quartet Q-5d.
(d) Calculated IR spectrum of doublet D-5d. All calculations at the
(U)B3LYP/6-311G(d,p) level of theory.
The EPR spectrum of matrix-isolated 5d obtained after
UV irradiation of 2d allows a definitive assignment of the
quartet spin state Q-5d. The spectrum shows a pattern
characteristic of a quartet spin state with zfs parameters
|D/hc|=0.291 cm-1 and |E/hc|=0.040 cm-1 (Figure 4). The
EPR spectrum of Q-5d is very similar to that obtained for Q-
5f which was discussed in detail elsewhere.20
not yet understood. Thus, photolysis of 1d in argon at 10 K
or in neon at 4 K produces only very low yields of 5d. The
highest yields are observed in solid argon at 4 K.
Due to the asymmetric fluorine substitution, the re-
arrangement of nitrene 1e can principally produce the
two isomeric azirines 3e and 30e and the two isomeric
ketenimines 4e and 40e. Irradiation of T-1e with UV light
(254 nm) produces 40e as the major product, but almost no
4e (Figure 5). In contrast, visible light irradiation (λ >
420 nm) of T-1e yields 3e as the major product and only
traces of 30e. However, if a matrix containing ketenimine
40e is irradiated with visible light, azirine 30e is produced as
a major product. On the other hand, UV irradiation of a
matrix containing azirine 3e produces ketenimine 4e as a
major product (Figure 6). Thus, depending on the sequence
of UV and visible light irradiation, the isomeric azirine 3e
or the isomeric ketenimine 40e is obtained as major product.
This puzzling photochemical behavior will be discussed
later.
Irradiation with visible light reduces the amount of 5d in
the matrix. Presumably, this irradiation leads to local heat-
ing of the matrix and thermal recombination of 5d with
iodine atoms in proximity. The matrix and temperature
effects controlling the formation of 5d are very subtle and
7374 J. Org. Chem. Vol. 74, No. 19, 2009

Grote and Sander
JOCArticle
FIGURE 7. IR spectra showing the 254 nm photochemistry of
nitrene T-1h in argon at 4 K after 80 min of irradiation.
(a) Calculated IR spectrum of T-1h. (b) Difference IR spectrum:
bands of T-1h pointing downward disappear; bands pointing
upward assigned to Q-5h and 4h appear during irradiation. The
second ketenimine 40h is not observed. (c) Calculated IR spectrum of
4h. (d) Calculated IR spectrum of Q-5h. All calculations at the
(U)B3LYP/6-311G(d,p) level of theory.
FIGURE 6. IR spectra showing the 254 nm photochemistry of
nitrene T-1e in argon at 4 K after 10 min of irradiation.
(a) Calculated IR spectrum of 30e. (b) Calculated IR spectrum of
3e. (c) Difference IR spectrum: bands of 3e and 30e pointing down-
ward disappear, bands pointing upward assigned to 4e and 40e
appear during irradiation. Bands marked with b assigned to T-1e.
(d) Calculated IR spectrum of 40e. (e) Calculated IR spectrum of 4e.
All calculations at the (U)B3LYP/6-311G(d,p) level of theory.
SCHEME 4. Photochemistry of 2h
TABLE 5. IR Spectroscopic Data of 5e, Matrix-Isolated in Argon at
4 K, Compared to Calculations for the Quartet and the Doublet State
Ar, 4 K
Q-5e
D-5e
ν~calcd/cm-1
ν~exp/cm-1a
ν~calcd/cm-1 a
b
a,b
a,b
mode (Irel,exp
)
symmetry (Irel,calcd
)
symmetryc (Irel,calcd
)
14
15
16
17
a0
a0
a0
a0
a0
a0
a0
a0
a0
a0
a0
a0
a0
695.6 (0.01)
729.5 (0.06)
818.5 (0.08)
966.8 (0.27)
1071.6 (0.14)
1197.1 (0.14)
1263.6 (0.00)
1294.9 (0.00)
1325.9 (0.12)
1351.7 (0.03)
1449.1 (1.00)
1541.1 (0.66)
1578.5 (0.09)
696.1 (0.01)
728.3 (0.05)
823.8 (0.07)
968.5 (0.23)
1075.0 (0.12)
1192.9 (0.10)
1262.0 (0.03)
1270.2 (0.04)
1308.3 (0.09)
1359.2 (0.02)
1442.1 (1.00)
1550.9 (0.15)
1595.1 (0.05)
727.3 (0.22)
953.0 (0.83)
18 1057.3 (0.25)
19 1188.7 (0.11)
20
21
22
23
24 1431.0 (1.00)
25
26
aUB3LYP/6-311G(d,p), unscaled. bRelative intensity based on the
strongest absorption. cCalculation performed with C1 symmetry.
The quartet nitreno radical Q-5e with a characteristic IR
absorption at 1431 cm-1 is formed as a minor product during
prolonged 254 nm irradiation (Table 5). Q-5e is also formed
in a neon matrix, but the yield is much lower than in solid
argon. Annealing of an argon matrix containing Q-5e at 30
K or irradiation with λ > 420 nm results in recombination of
the iodine atom with 5e producing nitrene 1e or azirine 3e.
Photochemistry of 2,5-Difluoro-4-iodophenyl Azide 2h. The
photochemistry of the azides 2g, 2h, 2i, and 2k bearing only
one ortho fluorine substituent is very similar. The asym-
metric fluorine substitution leads to two isomeric azirines 3
and 30 and ketenimines 4 and 40 which are formed in variable
amounts depending on the experimental conditions. The
overall trend in the selectivity of the formation of these
products is quite similar in all four systems; we therefore
focus on the photochemistry of 2h as a prototype (Scheme 4).
UV irradiation (λ = 254 nm) of phenyl azide 2h in solid
argon or neon at 4 K results in a rapid decrease of all IR
absorptions of the azide and formation of a new product with
three characteristic bands at 1555.0, 1492.2, and 1428.6 cm-1
(in argon). The new product was assigned to triplet nitrene T-
1hby comparison of the IR spectrum with DFT calculations at
the UB3LYP/6-311G(d,p) level of theory. The EPR spectrum
of T-1h with |D/hc|=0.954 cm-1 and |E/hc|=0.012 cm-1 is
characteristic of triplet phenylnitrenes and in line with the
other nitrenes investigated in this study (Table 1).
Short irradiation times of the precursor leads to T-1h
exclusively. However, during prolonged UV irradiation
two new sets of IR bands are formed while all bands assigned
to T-1h decrease. A characteristic band at 1888.3 cm-1
(argon, 4 K) is assigned to the NdCdC str vibration of a
ketenimine 4. A closer investigation reveals that only isomer
J. Org. Chem. Vol. 74, No. 19, 2009 7375

Grote and Sander
JOCArticle
TABLE 6. IR Spectroscopic Data of 5h, Matrix-Isolated in Argon at
4 K, Compared to Calculations for the Quartet and the Doublet State
Ar, 4 K
Q-5h
D-5h
ν~exp/cm-1
ν~calcd/cm-1a
ν~calcd/cm-1a
b
a,b
a,b
mode (Irel,exp
)
symmetry (Irel,calcd
)
symmetry (Irel, calcd
)
12
13
14
15
16
a0
a0
a00
a0
a0
a0
a0
a0
a0
a’
a0
a0
a0
a0
729.4 (0.12)
768.7 (0.06)
818.0 (0.00)
857.9 (0.25)
944.8 (0.11)
1157.8 (0.34)
1194.2 (0.08)
1216.3 (0.04)
1282.5 (0.01)
1300.3 (0.08)
1366.3 (0.10)
1443.0 (1.00)
1518.2 (0.22)
1563.6 (0.08)
a0
a0
a00
a00
a0
a0
a0
a0
a0
a0
a0
a0
a0
a0
728.6 (0.11)
774.6 (0.05)
821.7 (0.00)
857.7 (0.20)
949.4 (0.11)
1160.2 (0.33)
1196.6 (0.04)
1212.6 (0.01)
1262.5 (0.05)
1280.2 (0.02)
1357.9 (0.13)
1436.5 (1.00)
1532.8 (0.13)
1581.5 (0.06)
17 1139.3 (0.39)
18
19
20
21
22
23 1411.4 (1.00)
24 1478.5 (0.61)
25
aUB3LYP/6-311G(d,p), unscaled. bRelative intensity based on the
strongest absorption.
FIGURE 9. IR spectra showing the 320-400 nm photochemistry
of nitrene T-1h and ketenimine 4h in argon at 4 K after 10 min of
irradiation. (a) Calculated IR spectrum of 4h. (b) Calculated IR
spectrum of T-1h. (c) Difference IR spectrum: bands of T-1h
pointing downward disappear; bands pointing upward assigned to
3h and 30h appear during irradiation. (d) Calculated IR spectrum of
30h. (e) Calculated IR spectrum of 3h. All calculations at the
(U)B3LYP/6-311G(d,p) level of theory.
UV photolysis of azides 2g, 2i, and 2k results in the
expected formation of the corresponding triplet nitrenes
T-1 as primary products in high yields. However, while the
further photochemistry of 1g and 1k shows a quite similar
selectivity than that of 1h, the photochemistry of 1i upon UV
irradiation is much less selective. UV irradiation (λ=254 nm)
of the nitrenes 1g and 1k results in the selective rearrange-
ment to ketenimines 4g and 4k, respectively. Thus, as with
1h, the nitrogen atoms insert selectively into the adjacent
C-C bonds bearing the fluorine substituents. Only traces of
the isomer 40g, but not of 40k, are found. In contrast, UV
photolysis of 1i shows no selectivity, and both 4i and 40i are
formed in similar yields. The photochemistry of 2g is sum-
marized in Scheme 5.
FIGURE 8. (a) EPR spectrum of Q-5h (argon, 15 K) formed after
308 nm irradiation of T-1h. (b) Spectrum simulated with S=3/2, |D/
hc|=0.278 cm-1, |E/hc|=0.040 cm-1, and g=2.003. Due to slightly
smaller zfs parameters, z4 appears as shoulder on the low field side
of y1.
4h is formed, while 40h is not observed (Figure 7). Thus, the
nitrogen atom of nitrene T-1h is inserting into the C-C bond
bearing the fluorine substituent.
The most intense band of the remaining IR absorptions at
1411.4 cm-1 is assigned to nitreno radical 5h. For the quartet
and the doublet state of 5h, DFT calculations (UB3LYP/6-
311G(d,p)) predict bands at 1443.0 and 1436.5 cm-1, respec-
tively, which are too close to allow an assignment (Table 6).
However, the EPR spectrum clearly shows a quartet system
Q-5h with zfs parameters |D/hc|=0.278 cm-1 and |E/hc|=
0.040 cm-1, similar to that of Q-5f (Figure 8).
Upon irradiation with visible light (λ > 420 nm), the IR
absorptions of 1h and 5h decrease while new bands
appear in the spectrum. These new absorptions were
identified as 30h by comparison with DFT calculations
(Supporting Information, Figure S4). The most charac-
teristic absorption is the CdN stretching vibration at
1718.1 cm-1 (argon, 4 K). The decrease of nitreno radical
5h upon visible light irradiation is most likely caused by
local heating of the matrix resulting in a thermal recom-
bination of 5h with an iodine atom. Azirine 3h is obtained
in small amounts during 320-400 nm irradiation
(Figure 9).
Subsequent irradiation of the product mixtures obtained
by UV irradiation of 1g, 1k, and 1i with visible light (λ >
420 nm) results in the formation of 30g, 30k, and 30i, respec-
tively, as the major products. These products are formed by
a formal addition of the nitrene nitrogen atom to the
adjacent C-C bond without fluorine substitution. The
selectivity of this rearrangement is high; only small amounts
of 3i but no 3k are observed under these conditions. For 3g,
this selectivity is still apparent, but less striking (Supporting
Information, Figure S1). While the irradiation of nitrenes
with visible light yields higher amounts of azirine 30i
(Supporting Information, Figure S2), 30h, 30k, and 30g, the
7376 J. Org. Chem. Vol. 74, No. 19, 2009

Grote and Sander
JOCArticle
SCHEME 5. Photochemistry of 2g
TABLE 8. IR Spectroscopic Data of 5i, Matrix-Isolated in Argon at 4 K,
Compared to Calculations for the Quartet State
Ar, 4 K
Q-5i
b
a,b
mode
ν~exp/cm-1 (Irel,exp
)
symmetry
ν~calcd/cm-1a (Irel,calc
)
12
13
14
15
16
17
18
19
20
21
22
23
24
25
a0
a00
a0
a00
a0
a0
a0
a0
a0
a0
a0
a0
a0
a0
702.0 (0.20)
766.9 (0.41)
804.2 (0.05)
934.0 (0.00)
1024.5 (0.62)
1111.7 (0.23)
1175.3 (0.28)
1225.6 (0.31)
1299.6 (0.03)
1316.2 (0.47)
1405.5 (0.49)
1425.8 (1.00)
1525.8 (0.35)
1563.0 (0.99)
1007.9 (0.50)
1159.7 (0.23)
1400.6 (0.31)
1536.0 (1.00)
aUB3LYP/6-311G(d,p), unscaled. bRelative intensity based on stron-
gest absorption.
TABLE 9. IR Spectroscopic Dataof 5k, Matrix-Isolated in Argon at 4 K,
Compared to Calculations for the Quartet and the Doublet State
Ar, 4 K
Q-5k
D-5k
ν~exp/cm-1
ν~calc/cm-1a
ν~calcd/cm-1a
b
a,b
a,b
mode (Irel,exp
)
symmetry (Irel,calc
)
symmetry (Irel,calcd
)
TABLE 7. IR Spectroscopic Data of 5g, Matrix-Isolated in Argon at 4 K,
Compared to Calculations for the Quartet and the Doublet State
6
7
8
513.5 (0.46)
a0
a000
a
a00
a0
a000
a00
a
a00
a0
a0
a0
a0
a0
a0
a0
a0
a0
a0
520.2 (0.20)
552.1 (0.05)
567.1 (0.03)
688.7 (0.00)
763.5 (0.21)
782.0 (0.74)
838.1 (0.11)
839.6 (0.47)
945.8 (0.01)
1029.5 (0.10)
1110.4 (0.28)
1173.6 (0.59)
1227.3 (0.28)
1285.2 (0.14)
1316.1 (0.32)
1396.1 (0.54)
1419.9 (1.00)
1522.4 (0.70)
1529.4 (0.83)
a00
a00
a
a00
a0
a000
a00
a
a00
a0
a0
a0
a0
a0
a0
a0
a0
a0
a0
459.9 (0.01)
519.3 (0.14)
567.2 (0.02)
679.0 (0.02)
769.1 (0.18)
779.0 (0.73)
838.3 (0.09)
846.5 (0.26)
940.4 (0.00)
1039.6 (0.10)
1109.4 (0.16)
1182.8 (0.61)
1228.3 (0.22)
1266.4 (0.02)
1278.7 (0.30)
1388.7 (1.00)
1421.3 (0.65)
1536.4 (0.83)
1548.4 (0.42)
Ar, 4 K
Q-5g
D-5g
9
ν~exp/cm-1
ν~calcd/cm-1a
ν~calcd/cm-1a
10
11
12
13
14
15
750.4 (0.15)
766.6 (0.23)
b
a,b
a,b
mode (Irel,exp
)
symmetry (Irel,calc
)
symmetry (Irel, calc
)
15
16
a0
a00
a0
a0
a0
a0
a0
a0
a0
a0
a0
a0
772.2(0.11)
822.2 (0.14)
1014.6 (0.33)
1072.9 (0.48)
1177.0 (0.08)
1217.0 (0.03)
1306.3 (0.00)
1323.7 (0.32)
1401.4 (0.11)
1428.1 (1.00)
1555.2 (0.23)
1564.7 (0.60)
a0
a00
a0
a0
a0
a0
a0
a0
a0
a0
a0
a0
774.9(0.14)
810.9 (0.13)
1015.7 (0.42)
1074.8 (0.54)
1170.5 (0.14)
1219.6 (0.02)
1278.8 (0.09)
1302.7 (0.40)
1409.5 (0.65)
1428.5 (1.00)
1563.1 (0.28)
1584.9 (0.62)
17 1002.3 (0.88)
18 1061.9 (0.38)
19
20
21
22
23
24 1408.5 (1.00)
25
26
16 1087.1 (0.15)
17 1152.4 (0.38)
18
19
20
21 1372.4 (0.92)
22 1394.4 (0.31)
23 1487.1 (0.23)
24 1492.4 (1.00)
aUB3LYP/6-311G(d,p), unscaled. bRelative intensity based on stron-
gest absorption.
aUB3LYP/6-311G(d,p), unscaled. bRelative intensity based on stron-
gest absorption.
azirines 3g and 3i are obtained by visible light irradiation
from the corresponding ketenimines 4g and 4i. This photo-
chemical rearrangement is most selective for 4g (Support-
ing Information, Figure S3). Compound 3k is not formed
at all.
In analogy to nitrene 1h, prolonged UV irradiation of the
nitrenes 1g, 1i, and 1k produces the corresponding nitreno
radicals 5 which were characterized by their IR and EPR
spectra. In all three cases, the EPR spectra correspond to
quartet systems Q-5 (Tables 7-9). The EPR spectra could be
simulated with the zfs parameters given in Table 1.
An additional intense and broad absorption at 1892.5 cm-1
(argon, 4 K), characteristic of the NdCdC str vibration of a
ketenimine, indicates the formation of 4l. In contrast to the
photochemistry of the other systems described here, no traces
of nitreno radical 5l could be detected by IR or EPR spectros-
copy. Irradiation with visible light (λ > 420 nm) yields, as
expected, the azirine 3l with the characteristic CdN str.
vibration at 1739.1 cm-1 (Figure 11).
Discussion
Nitrene EPR Investigation. The EPR spectrum of nitrene
T-1f (|D/hc| = 1.108 cm-1, |E/hc| = 0.012 cm-1) has been
discussed in detail previously.20 The D values of the triplet
nitrenes T-1d-T-1l are close to that of T-1f in the character-
istic range of aryl nitrenes (Table 1).21 They strongly depend
on the spin density at the nitrene center which is influenced
by substituent effects. A strong linear correlation between
Photochemistry of 3,5-Difluoro-4-iodophenylazide 2l. UV
irradiation (λ=254 nm) of 2l in solid argon or neon at 4 K
results in the rapid decrease of the IR absorptions of the
azide (Scheme 6). Simultaneously, new intense absorptions
are formed at 1507.9 and 1300.6 cm-1 (argon, 4 K) which are
assigned to T-1l by comparison with UB3LYP/6-311G(d,p)
calculations (Figure 10). The EPR spectrum shows the
formation of a triplet system with zfs parameters |D/hc|=
1.053 and |E/hc|=0.0103 cm-1
.
(21) Wasserman, E. Prog. Phys. Org. Chem. 1971, 8, 319–36.
J. Org. Chem. Vol. 74, No. 19, 2009 7377

Grote and Sander
JOCArticle
FIGURE 11. IR spectra showing the 420 nm photochemistry of 4l
in argon at 4 K after 30 min irradiation. (a) Calculated IR spectrum
of 4l. b) Difference IR spectrum: bands of 4l pointing downward
disappear; bands pointing upward assigned to 3l appear during
irradiation. (c) Calculated IR spectrum of 3l. All calculations at the
(U)B3LYP/6-311G(d,p) level of theory.
FIGURE 10. IR spectra showing the 254 nm photochemistry of 2l
in argon at 4 K. (a) IR spectrum of 2l after deposition at 4 K.
(b) Difference IR spectrum: bands of 2l pointing downward dis-
appear; bands pointing upward assigned to T-1l and 4l appear
during irradiation. Nitrene radical 5l was not observed. (c) Calcu-
lated IR spectrum of T-1l. (d) Calculated IR spectrum of 4l. All
calculations at the (U)B3LYP/6-311G(d,p) level of theory.
CHART 2. Characteristic CdCdN and CdN Stretching Vi-
brations of the Ketenimines 6a,2 6b-d,18 and Azirines 7a,b13 and
7c,d,18 respectively, Described in the Literature
SCHEME 6. Photochemistry of 2l
the D values of aryl nitrenes and the spin densities at the
nitrogen atoms was recently found by Wentrup et al.9 The
D values of T-1d-T-1l only roughly fit into this scheme, since
in these nitrenes the influence of the heavy iodine atom on the
D value has to be taken into account. The zfs parameters of
molecules containing heavy atoms are not only influenced by
spin-spin dipolar interactions but also by strong spin-orbit
coupling (SOC) contributions. The E values of 1 are un-
commonly large compared to most other arylnitrenes. This
also indicates the influence of the iodine substituent on the
zfs parameters due to a significant contribution of SOC, in
accordance with previous calculations for T-1f.20
Nitrene Rearrangements. The IR spectra of triplet nitrenes
T-1, azirines 3, and ketenimines 4 are in accordance with
DFT calculations (Tables 2 and 3 and Table T1, Supporting
Information) and agree well with similar molecules described
in the literature.2,13,14,18 The characteristic symmetrical CdC
str vibration of the nitrenes 1d-k is found between 1545 and
1597 cm-1. For T-1l, the calculated intensity of this vibration
is very low, and it could not be observed in the experimental
spectrum. For the other nitrenes it was found that the
frequency and intensity of this vibration strongly depends
on the pattern of the fluorine substitution, in agreement with
theoretical predictions. The three nitrenes with two ortho
fluorine substituents show the highest frequency for the
symmetrical CdC str vibration. In general, the frequency
increases while the intensity decreases with the number of
fluorine substituents. The C-N str vibration is also char-
acteristic and found in the spectral range between 1284 and
1363 cm-1. The C3C4C5 deformation vibrations are located
in the fingerprint region between 803 and 1056 cm-1 (see
Supporting Information Table T1).
The CdCdN str vibrations of ketenimines 4 and 40 are
located between 1840.7 and 1893.5 cm-1, in good agreement
with similar systems published in the literature (Chart 2). The
CdN str vibrations of the azirines 3 and 30 are found between
1680.8 and 1739.1 cm-1 (Table 3).
Phenyl nitrenes with fluorine substituents initially were
regarded as photochemically stable.22 Later it was shown
that selective photolysis of fluorinated phenylnitrenes pro-
duces azirines13 and ketenimines, although fluorination
markedly increases the photostability of the nitrenes.18,14
(22) Dunkin, I. R.; Thomson, P. C. P. J. Chem. Soc., Chem. Commun.
1982, 1192–1193.
7378 J. Org. Chem. Vol. 74, No. 19, 2009

Grote and Sander
JOCArticle
Hayes and Sheridan2 observed that phenylnitrene and
ketenimine are formed with a constant initial ratio, regard-
less which matrix was used, which indicates that these
products interconvert and are formed in photostationary
equilibria. The increased stability of fluorinated phenylni-
trenes results in higher yields of the nitrenes in the product
mixtures. The thermal barrier for the addition of the nitrogen
atom into adjacent C-C bonds is predicted to be 3.5-
4.5 kcal/mol higher for the ortho-fluorinated systems than
for the parent phenylnitrene.12 Platz et al. found that the
rearrangement of phenylnitrene takes place on the excited π2
state of the singlet nitrene. This state is destabilized by
fluorine substituents, and hence these rearrangements be-
come less favored.23 This interpretation was later supported
by DFT calculations of Smith and Cramer.24 The photo-
chemistry of phenylnitrene with one fluorine atom in ortho
position was studied by Gritsan et al. with laser flash
photolysis and theoretical methods. Evidence was presented
that the azirine is an intermediate and that the corresponding
singlet nitrene and azirine interconvert in solution.25
Our results are in accordance with these studies. In all
cases, selective irradiations result in the interconversion
between nitrene 1, azirine 3, and ketenimine 4. However,
unsymmetrical substitution leads to the preferential forma-
tion of isomers that cannot be easily understood. A compli-
cation for the interpretation of the experimental results is
that there is no simple way to discriminate between true
photochemical reactions and hot ground state reactions. In
the latter case, the thermal deactivation of electronically
excited states results in thermally excited molecules which
undergo subsequent thermal rearrangements. To under-
stand the thermal pathways we calculated the relative en-
ergies of the isomers and the activation barriers at the
((U)B3LYP/6-311G(d,p)) level of theory. Interestingly,
there is a correlation between the activation barriers for the
ring closure of the nitrenes 1 and the relative yields of the
isomeric azirines 3, 30 in the matrix. In contrast, the yields of
the ketenimines 4, 40 correlate with their relative stabilities
(Table 10, Figure 12, Supporting Information Figures
S5-S8).
The nitrenes 1 are produced by UV irradiation (254 nm) of
the azides 2 in yields depending on the constitution of the
nitrene and the irradiation conditions. Since phenylnitrenes
are stabilized by ortho fluorine substituents, the yield of 1 is
highest with two ortho fluorine atoms and smallest without
fluorine atoms in ortho positions. Visible light irradiation of
the nitrenes 1 produces azirines, whereas prolonged UV
irradiation yields the ketenimines.
The influence of one ortho fluorine substituent on the
relative yields of the isomeric azirines 3 and 30 and the
isomeric ketenimines 4 and 40 is most pronounced for 1h
and 1k. Visible light irradiation (420 nm) produces exclu-
sively azirine 30h by formal insertion of the nitrogen atom
away from the ortho fluorine substituent. The calculated
barrier for the formation of 30h is 22.8 kcal/mol, while that
for 3h is 27.5 kcal/mol. At the same level of theory 3h is
calculated to be more stable than 30h by 4.3 kcal/mol. This is
TABLE 10. Energies (kcal mol-1) of the Rearrangement Products and
Their Transition States Relative to S-1 (B3LYP/6-311G(d,p))
TS1 TS10
3
30
TS2 TS20
4
40
2,6- (d) 26.0
2,3,6- (e) 24.6 26.7
2,3,5,6- (f) 24.7
10.4
16.5
3.2
4.5 -1.9
-1.5
6.8 10.0 16.0 15.3
5.1 12.7
2,3,5- (g)
2,5- (h)
2,3- (i)
2- (k)
26.0 23.3
4.1 12.8 10.1 20.9 -1.8
8.8 13.1 12.5 23.1 -3.0
4.8 15.1 11.3 21.1
8.9 14.7 12.4 22.1 -0.3
10.3 16.1
5.7
11.5
3.8
9.5
4.0
27.5 22.8
25.7 23.3
27.3 22.6
23.6
1.4
3,5- (l)
in accordance with earlier theoretical studies on fluorinated
phenylnitrenes.25 A careful inspection of the experimental
results reveals a general trend: the isomer of the azirine that is
preferentially formed is the kinetically favored but thermo-
dynamically less stable isomer; thus, the photochemical
rearrangements follow the thermal activation barriers.
UV irradiation (254 nm) of nitrene 1h produces keteni-
mine 4h but not 40h. This is striking since the preferred azirine
30h is expected to rearrange to 40h via ring-opening. DFT
calculations reveal that ketenimine 4h with no fluorine atom
attached to the cumulene double bond is considerably more
stable than isomer 40h and that the barrier for its formation
via ring-opening from azirine 3h is also lower (Figure 12a,
Table 10). The same is true for 4k and 40k. On the other hand,
the difference between both barriers for the ring-opening
from azirine 3i and 30i to 4i and 40i, respectively, and
particularly between the energies of 4i and 40i are much less
pronounced (Supporting Information, Figure S6 and
Table 10). The experimental results show that 4i and 40i are
formed in comparable yields in the matrix upon 254 nm
irradiation. The barriers of the ring-opening to form 4k and
40k, respectively, differ in a similar way than those for the
formation of 4e and 40e, but the difference in their thermo-
dynamic stability is more pronounced for 4k and 40k.
Since 40k could not be isolated in the matrix at all, the selecti-
vity of the ring-opening reaction seems to be thermodynami-
cally controlled and the thermal barriers for the ring-opening
of 3 are less related to the distribution of the isomers of 4.
Irradiation of 4g with visible light (420 nm) produces 3g as
a major product, the isomer that is obtained only as a minor
product of the visible light irradiation of nitrene 1g. Thus, by
selecting the sequence of irradiations, either 3g or 30g can be
obtained as the major product. This suggests that there is no
photostationary equilibrium between the isomers of 1g, in
particular not between 3g and 30g.
There are two possible mechanisms for the formation of 4g
via UV irradiation of 1g: (i) 3g is formed as an intermediate
with a low stationary concentration that subsequently re-
arranges to 4g or (ii) 4g is directly formed from 1g without
passing 3g as intermediate. Although there is no direct
evidence, mechanism (ii) seems to be more plausible since
(i) requires a switch in the selectivities by changing the
irradiation wavelength which is difficult to understand.
In some cases, the less stable ketenimine can be obtained
by irradiating the corresponding azirine with UV light. This
is especially striking for 1g where the more stable ketenimine
4g is almost the only product formed after UV irradiation.
However, 40g can be synthesized as minor products by
irradiating 30g with UV light, although most of the 30g reacts
back to the nitrene under these conditions.
(23) Platz, M. S. Acc. Chem. Res. 1995, 28, 487–492.
(24) Smith, B. A.; Cramer, C. J. J. Am. Chem. Soc. 1996, 118, 5490–5491.
(25) Gritsan, N. P.; Gudmundsdottir, A. D.; Tigelaar, D.; Zhu, Z.;
Karney, W. L.; Hadad, C. M.; Platz, M. S. J. Am. Chem. Soc. 2001, 123,
1951–1962.
J. Org. Chem. Vol. 74, No. 19, 2009 7379

Grote and Sander
JOCArticle
FIGURE 12. Relative energies of the rearrangement products and transition states of (a) 1h and )b) 1e.
An interesting case of selectivity is nitrene 1e with two fluorine
atoms in the ortho position but only one in the meta position. The
calculations predict that azirine 3e is 3.2 kcal/mol more stable
and the barrier for its formation is 2.1 kcal/mol lower than that of
its isomer 30e (Figure 12b). Of the two ketenimines, 40e is more
stable by 6.4 kcal/mol. Again, a selective photochemistry is
observed: UV irradiation of 1e preferentially produces the most
stable ketenimin isomer 40e, whereas visible light irradiation gives
mainly azirine 3e, which is formed by ring closure of the nitrene
with a lower barrier than 30e (in this case it is also thermodyna-
mically more stable than its isomer 30e).
Nitreno Radicals Q-5. Nitreno radicals 5 are an interesting
class of high-spin molecules with a quartet ground state. The
electronic structure of a nitreno radical 5 is described best as
a σ,σ,π-triradical with one unpaired electron located at a σ
orbital at the nitrogen atom, one at a σ orbital at the para
carbon atom, and one delocalized in the π system.17 This
delocalized unpaired π electron accounts for a partial nitrene
character at the nitrogen atom and partial carbene character
at the para carbon atom. To retain the local high-spin
characters of the nitrene and carbene centers, a quartet state
is energetically favored for the overall spin system.20
Of all systems investigated in this study, 1k shows the
highest selectivity. While in most systems the less stable
ketenimine and the azirine with the higher barrier for ring
closure from the nitrene are also formed as minor products,
30k and 4k are formed exclusively without any trace of 3k and
40k detectable in the matrix.
Although details of the mechanism of the formation of
azirines 3 and ketenimines 4 are not known, it is obvious that
DFT calculations of the relative stabilities of the isomers and
of the barriers for their formation can be used to predict the
distribution of isomers even in a semiquantitative way. A
larger difference in the thermal stability or the activation
barriers, respectively, results in a larger preference of one of
the isomers.
EPR signals of the nitreno radicals 5 are observed during
308 nm irradiation (XeCl excimer laser) of the precursors at
4-15 K. Annealing at 30 K results in the disappearance of
these signals and subsequent 308 nm irradiation at 4 K in
their reappearance. This behavior is characteristic of radical
pairs produced by photolysis of matrix isolated precursors.
The radical pair is formed in a matrix cage, and annealing
leads to rapid thermal radical recombination. Nitrenes 1 and
7380 J. Org. Chem. Vol. 74, No. 19, 2009

Grote and Sander
JOCArticle
FIGURE 14. Simulation of the field dependence of the most intense
axial transitions of a quartet system with D values between 0.05 and
0.285 cm-1 and E/D between 0 and 0.35. The values |D/hc| = 0.285
and |E/hc| = 0.043 cm-1 of 5f are marked.
a function of the zfs parameters (Figure 14). The most noticeable
difference between the quartet spectra are observed in the low-
field region for the transitions z4 and y1. While for Q-5 h
transition z4 appears at lower field than y1 (Figure 8), for Q-
5d with a slightly larger Dvalue it is located at higher field than y1
(Figure 4). The zfs parameters in nitreno radicals are largely
dominated by one-center spin-spin interactions,20 and since the
spin densities at the nitrene and the carbene centers in Q-5 show
only small variations, similar zfs parameters can be expected for
all nitreno radicals (Table 11).
Calculations show that the magnetic z-axis of the quartet
nitreno radical is located parallel to the N-C bond, the
x-axis is perpendicular to the molecular plane, and the y-axis
lies in the molecular plane. Thus, the nitrene character
prevails over the carbene character. However, the carbene
character is reflected by an enhanced dipolar field in the
y-direction which is sufficiently strong to give significant
magnetically inequivalent y- and x-directions. Thus, the
nonzero E-value in Q-5f and the other nitreno radicals
mainly reflects the carbene character of the system. This
carbene character, in turn, is controlled by the properties of
the π-system since the out-of-plane spin-density arises from
the conjugation of the nitrene out-of-plane SOMO and the
π-system of the ring.20
To synthesize the nitreno radicals in argon matrices the
corresponding azides 2 have to be irradiated with UV light
(254 nm). Under these conditions, the formation of 5 com-
petes with the rearrangement to the ketenimines 4. Only
nitrenes 1 with one or two fluorine atoms in the ortho
positions are stable enough toward rearrangement to pro-
duce nitreno radicals 5 in yields high enough to allow their
spectroscopic detection. Thus, nitreno radical 5l could be
detected neither by IR nor by the more sensitive EPR
spectroscopy.
FIGURE 13. Energy of the four quartet sublevels of the external field
H parallel to z, x, and y-axes, respectively, calculated for a quartet
system using the zfs parameters |D/hc|=0.285 cm-1, |E/hc|=0.043 cm-1
(giso = 2.003) of Q-5f as a representative example for the assignment
of the transitions of the nitrino radicals in this study.
the other intermediates do not show this behavior (no
reaction with N2 under the conditions of matrix isolation)
which allows to assign signals of the nitreno radicals 5 even if
the yield is low.
The EPR spectra of Q-5 could be reproduced by simula-
tion of quartet high-spin systems with suitable zfs para-
meters, which allows the assignment of all experimentally
observed EPR transitions. For Q-5f the zfs parameters |D/hc|=
0.285 and |E/hc|= 0.043 cm-1 were reported previously.20
The field dependence of the magnetic energy levels of 5f was
calculated for the principal axes using the solution of the
Hamilton matrix of a quartet state26,27 with the parameters
derived from the simulation (Figure 13). Two signals cannot
be assigned to axial transitions. These off-axis transitions are
typical for EPR spectra of high-spin systems with S > 1. The zfs
parameters of the quartet nitreno radicals Q-5d-Q-5k are quite
similar. The D and E values are found in narrow ranges between
0.278 and 0.29 cm-1 and 0.040 and 0.043 cm-1, respectively
(Table 1). The small differences in the appearance of the EPR
spectra are explained by the field dependence of the transitions as
The formation of nitreno radicals 5 is most efficient in
argon at 4 K, whereas at 10 K or in neon at 4 K the yields are
much lower. In several instances, the nitreno radicals could
be detected in argon at 10 K only by EPR, but not by IR
spectroscopy, whereas at 4 K clear IR signals were obtained.
This is probably related to the immobilization of radical
pairs (5 and the iodine atom) in the matrix cages. A slight
(26) Rai, U. S.; Symons, M. C. R.; Wyatt, J. L.; Bowman, W. R. J. Chem.
Soc., Faraday Trans. 1993, 89, 1199–201.
(27) Atherton, N. M. Principles of Electron Spin Resonance; Prentice Hall:
New York, 1993.
J. Org. Chem. Vol. 74, No. 19, 2009 7381

Grote and Sander
JOCArticle
TABLE 11. Spin Densities at the Nitrene and the Carbene Centers
of Q-5
spin densitya
Q5
at N
at C4
2,3,5,6- (f)
2,3,6- (e)
2,6- (d)
2,3,5- (g)
2,5- (h)
2,3- (i)
1.5771
1.5749
1.5739
1.5803
1.5709
1.5811
1.5742
1.2480
1.2420
1.2398
1.2565
1.2586
1.2452
1.2493
2- (k)
aUB3LYP/6-311G(d,p).
FIGURE 15. Geometries of the 4A2 state of Q-5d and the 2A2 state
of D-5d on the UB3LYP/6-311G(d,p) level of theory. Both states
show C2v symmetry and the bond lengths and angles, as well as the
calculated IR spectra, are quite similar.
increase of the temperature results in an efficient recombina-
tion of the radical pair under formation of 1. The effect of
argon vs neon matrices is not understood, we can only
speculate that the different shape of matrix cages in these rare
gases influences the stability of the radical pairs.
using a N2(l)-cooled MCT detector in the range 400-4000 cm-1
.
X-band EPR spectra were recorded with a Bruker Elexsys E500
EPR spectrometer with an ER077R magnet (75 mm pole cap
distance), an ER047 XG-T microwave bridge, and an oxygen-free
high-conductivity copper rod (75 mm length, 3 mm diameter) cooled
by a closed-cycle cryostat.
Theoretical studies with 4-dehydrophenylnitrene have
4
shown that the ground state of 5a is the A2 state, only
2-5 kcal/mol lower in energy than the lowest lying doublet
state, which is a 2A2 state.17 Thus, the lowest-energy doublet
state of 5a has the same electronic configuration as the
quartet ground state. The geometries of both states are quite
similar and show only minor bond length alternation, mostly
Broadband irradiation was carried out with mercury high-
pressure arc lamps in housings equipped with quartz optics and
dichroic mirrors in combination with cutoff filters (50% trans-
mission at the wavelength specified). IR irradiation from the
lamps was absorbed by a 10 cm path of water. For 254 nm
irradiation a low pressure mercury arc lamp was used. A XeCl
excimer laser was used for 308 nm irradiation.
Computational Methods. Optimized geometries and vibra-
tional frequencies of all species were calculated with the Gauss-
ian 03 suite of programs.30 The computer simulation of the EPR
spectrum was performed by using the XSophe computer simula-
tion software suite (version 1.0.4),31 developed by the Centre for
Magnetic Resonance and Department of Mathematics, Uni-
versity of Queensland, Brisbane, Australia, and Bruker Analy-
tik GmbH, Rheinstetten. Germany.
˚
reflected in a 0.028 A longer C-N bond in the doublet
geometry. Similar results are found for the geometries and
infrared spectra of the nitreno radicals 5d-l. In Figure 15 the
geometries of the 4A2 state of Q-5d and the 2A2 state of D-5d
are shown as typical examples for nitreno radicals. Both C2v
symmetrical states show almost similar bond lengths and
bond angles. As a consequence, the calculated doublet and
quartet infrared spectra are very similar with only small
frequency shifts (Table 4). This makes it difficult to deter-
mine the spin state of these nitreno radicals by infrared
spectroscopy, in particular since only the strongest bands
are visible in the experiment. Hence, only EPR spectroscopy,
which is very sensitive to paramagnetic species, proves the
quartet state of the nitreno radicals. An interesting case is 5f,
where the doublet state has a nonplanar CS symmetrical
geometry, while Q-5f is planar with C2v symmetry. As a result,
the calculated IR spectra of the quartet and doublet states of 5f
are different enough and allowed a definitive assignment of the
electronic state of 5f based only on IR spectroscopy.19
Acknowledgment. This work was financially supported by
the Deutsche Forschungsgemeinschaft and the Fonds der
Chemischen Industrie.
Supporting Information Available: Additional figures and
tables; synthetic, spectroscopic, and analytical data. Geome-
tries, total energies, and IR spectroscopic data of the photo-
1
products. Relative energies of all rearrangement products; H
and 13C NMR spectra. This material is available free of charge
via the Internet at http://pubs.acs.org.
Experimental Section
(30) Frisch, M. J. T.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb,
M. A.; Cheeseman, J. R.; Montgomery, Jr., J. A.; Vreven, T.; Kudin, K. N.;
Burant, J. C.; Millam, J. M.; Iyengar, S. S.; Tomasi, J.; Barone, V.; Mennucci,
B.; Cossi, M.; Scalmani, G.; Rega, N.; Petersson, G. A.; Nakatsuji, H.; Hada,
M.; Ehara, M.; Toyota, K.; Fukuda, R.; Hasegawa, J.; Ishida, M.; Nakajima,
T.; Honda, Y.; Kitao, O.; Nakai, H.; Klene, M.; Li, X.; Knox, J. E.; Hratchian,
H. P.; 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.; Ayala, P. Y.; Morokuma, K.; Voth, G. A.; Salvador, P.; Dannenberg,
J. J.; Zakrzewski, V. G.; Dapprich, S.; Daniels, A. D.; Strain, M. C.;
Farkas, O.; Malick, D. K.; Rabuck, A. D.; Raghavachari, K.; Foresman,
J. B.; Ortiz, J. V.; Cui, Q.; Baboul, A. G.; Clifford, S.; Cioslowski, J.;
Stefanov, B. B.; Liu, G.; Liashenko, A.; Piskorz, P.; Komaromi, I.; Martin,
R. L.; Fox, D. J.; Keith, T.; Al-Laham, M. A.; Peng, C. Y.; Nanayakkara, A.;
Challacombe, M.; Gill, P. M. W.; Johnson, B.; Chen, W.; Wong, M. W.;
Gonzalez, C.; Pople, J. A. Gaussian 03, Revision C.02; Gaussian, Inc.:
Wallingford CT, 2004.
Synthesis. All phenyl azides were synthesized in analogy to
the synthesis of 2,3,4,5-tetrafluoroaniline and 2,3,4,5-tetrafluor-
ophenylazide described in the literature.19 The 2,6-difluoro-4-
iodoaniline was synthesized starting from 2,6-difluoroaniline
according to a procedure described by Kutepov et al.28 The
synthetic, spectroscopic, and analytical data are given in the
Supporting Information.
Matrix Isolation. Matrix isolation experiments were per-
formed by standard techniques29 with closed cycle helium
cryostats allowing cooling of a CsI spectroscopic window to 4 K.
FTIR spectra were recorded with a standard resolution of 0.5 cm-1
,
(28) Kutepov, D. F.; Khokhlov, D. N.; Tuzhilkina, V. L. Zh. Obsh. Khim.
1960, 30, 2484–9.
(31) Griffin, M.; Muys, A.; Noble, C.; Wang, D.; Eldershaw, C.; Gates,
K. E.; Burrage, K.; Hanson, G. R. Mol. Phys. Rep. 1999, 26, 60–84.
(29) Dunkin, I. R. Matrix Isolation Techniques: A Practical Approach;
Oxford University Press: Oxford, 1998.
7382 J. Org. Chem. Vol. 74, No. 19, 2009