Amination of benzylic C-H bonds by arylazides catalyzed by CoII-porphyrin complexes: A synthetic and mechanistic study

Article abstract of DOI:10.1002/chem.200390018

Co^II-porphyrin complexes catalyze the reaction of aromatic azides (ArN₃) with hydrocarbons that contain a benzylic group (ArR¹R²CH) to give the corresponding amines (ArR¹R²C-HAr). When at least one of the R substituents is hydrogen, the catalytic reaction proceeds further to give the imine ArRC=NAr in good yields. The reaction mechanism has been investigated. The reaction proceeds through a reversible coordination of the arylazide to the Con^II-porphyrin complex. This unstable adduct can either react with the hydrocarbon in the rate-determining step or decompose by a unimolecular mechanism to afford a putative nitrene complex, which reacts with more azide, but not with the hydrocarbon, to afford the byproduct diaryldiazene. The kinetics of the catalytic reaction have been investigated for a range of azides and substituted toluenes. Arvlazides with electron-withdrawing substituents react at a faster rate and a good correlation is found between the log(k) and the Taft parameters. On the other hand, an excellent correlation between the logarithm of the rate for substituted toluenes relative to that of toluene and a radical parameter (σ_JJ) alone was found, with no significant contribution by polar parameters. An explanation has been proposed for this anomalous effect and for the very high isotopic effect (k_H/k_D= 14) found.

Full text of DOI:10.1002/chem.200390018

FULL PAPER
À
Amination of Benzylic C H Bonds by Arylazides Catalyzed by
II
Co ± Porphyrin Complexes: A Synthetic and Mechanistic Study
[
a]
[a]
[a]
[b]
[a]
Fabio Ragaini, Andrea Penoni, Emma Gallo, Stefano Tollari, Claudia Li Gotti,
[
a]
[a]
[a]
Marta Lapadula, Enrica Mangioni, and Sergio Cenini*
II
Abstract: Co ± porphyrin complexes
hydrocarbon in the rate-determining
with electron-withdrawing substituents
react at a faster rate and a good corre-
lation is found between the log(k) and
the Taft parameters. On the other hand,
an excellent correlation between the
logarithm of the rate for substituted
toluenes relative to that of toluene and a
catalyze the reaction of aromatic azides
ArN ) with hydrocarbons that contain a
3
benzylic group (ArR R CH) to give the
step or decompose by a unimolecular
mechanism to afford a putative ™ni-
trene∫ complex, which reacts with more
azide, but not with the hydrocarbon, to
afford the byproduct diaryldiazene. The
kinetics of the catalytic reaction have
been investigated for a range of azides
and substituted toluenes. Arylazides
(
1
2
1
2
corresponding amines (ArR R CÀN-
HAr). When at least one of the R
substituents is hydrogen, the catalytic
reaction proceeds further to give the
imine ArRCˆNAr in good yields. The
Á
radical parameter (s † alone was found,
JJ
with no significant contribution by polar
parameters. An explanation has been
proposed for this anomalous effect and
for the very high isotopic effect (k /k ˆ
reaction mechanism has been investi-
gated. The reaction proceeds through a
reversible coordination of the arylazide
Keywords: amination
¥
azides
cobalt
¥
¥
H
D
CÀH activation
¥
II
to the Co ± porphyrin complex. This
unstable adduct can either react with the
14) found.
porphyrinoids
last reactions only the allylic C
ÀH bond of an olefin could be
functionalized.
Introduction
The development of intermolecular nitrogen-atom transfer
reactions catalyzed by metals, such as the amination of
Aromatic azides (ArN ) represent a wide class of easily
3
synthesized compounds. We have recently reported that
porphyrin complexes of cobalt(ii) are able to activate aromatic
azides for the amination under mild conditions of allylic CÀH
[1±1ꢀ0
saturated CÀH bonds, is a topic of high current interest.
Reasonably good results have only been obtained by using
[
1±40
[90
[
N(p-tolylsulfonyl)imino0phenyliodinane
(PhIˆNTs),
bonds, and we report here that the same catalytic system
[1, 50
chloramine-T,
and related derivatives as the nitrogen
also results in the more difficult activation of the CÀH bonds
of saturated organic compounds, to give secondary amines
and imines. While the use of porphyrin complexes of
manganese, iron, and ruthenium as catalysts in the amidation
source. However these compounds can only be obtained
from the corresponding sulfonamides,[ and, thus, they belong
to a limited family of reagents. Recently, we developed a
methodology for the intermolecular allylic amination of
unactivated olefins by nitroarenes in the presence of CO
and ruthenium complexes as catalysts.[ However, in these
70
[
2±40
reactions by PhIˆNTs has been reported,
to the best of our
knowledge, porphyrin complexes of cobalt(ii) have never been
80
employed before as catalysts for these reactions with the
exception of the allylic amination of cyclohexene mentioned
[
a0 Prof. Dr. S. Cenini, Prof. Dr. F. Ragaini, Dr. A. Penoni,
Dr. E. Gallo, Dipl. Chem. C. Li Gotti,
Dipl. Chem. M. Lapadula, Dipl. Chem. E. Mangioni
Dipartimento di Chimica Inorganica, Metallorganica e Analitica
and ISTM-CNR
[9, 1ꢀ0
before.
The use of organic azides as aminating agents
considerably extends the number of nitrogen-containing
organic compounds that can be potentially obtained in
catalytic reactions. Such a reaction has never been reported
before, apart from our preliminary communication on this
Via Venezian 21, 2ꢀ133 Milano (Italy)
Fax : (‡39) ꢀ2-5ꢀ3144ꢀ5
[110
topic.
It should be also pointed out that although the
E-mail: sergio.cenini@unimi.it
condensation of ketones with primary amines is a textbook
route to ketimines, the procedure is sometimes not practical
for the synthesis of aromatic ketimines.[ Thus other practical
routes to imines are desirable to supplement the classical
condensation pathway.
[
b0 Prof. Dr. S. Tollari
Dipartimento di Scienze Chimiche Fisiche e Matematiche
Via Valleggio 11, 221ꢀꢀ Como (Italy)
120
Supporting information for this article is available on the WWW under
http://www.chemeurj.org/ or from the author.
Chem. Eur. J. 2003, 9, No. 1¹ 2ꢀꢀ3 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
ꢀ947-6539/ꢀ3/ -ꢀ ꢀ9 2ꢀ 41 9 $ 2ꢀ.ꢀꢀ+.5ꢀ/ꢀ
249

S. Cenini et al.
FULL PAPER
Table 1. Amination of toluene by p-XC
6
H
4
N
3
catalyzed by Co ± porphyrin.[a0
Results
Catalyst
X
t [h0[b0 Imine 2
Amine 1
yield [%0 yield [%0
p-XC
6
H
4
NH
2
Diazene 3
yield [%0
yield [%0[
c,d0
[c0
[c0
[c0
Synthetic results: The reaction of hydrocarbons that
contain a benzylic group with para-nitrophenylazide
catalyzed by Co ± porphyrin derivatives leads to then
formation of the corresponding amines [Eq. (1)0. When at
least one of the R substituents is hydrogen, the catalytic
reaction proceeds further, at least in part, to give the
corresponding imine [Eq. (2)0:
[
Co(tpp)0
NO
OCH3
Cl
2
2
5
11
1
23 (2a)
±
28
22 (3a)
II
[Co(tpp)0
1
ꢀ
2
2 1b ) ꢀ ( ±
28 (2c)
25.5 (2 h)
2 1g ) 8 ( ±
14.9 (2e)
1
29
44.4
18
16.9
44.4
43.18.7 (
311
1
4
3
3
ꢀ
3b)25 (
[
[
[
[
Co(tpp)0
Co(tpp)0
Co(tpp)0
Co(tpp)0
±
±
18 (3c)
18 (3 h)
38.6 (3g)
49.2 (3e)
12.1 (3f)
3d)
H
CH
F
3
5
±
±
±
±5 (
±
±
±
±
±
[Co(tpp)0
Co(tpp)0
[Co(p-Cltpp)0
Co(p-Cltpp)0
[Co(p-Cltpp)0
Br
CN
NO2
OCH
Cl
8
4
2
1
2
37.2 (2f)
37.ꢀ (2d)
24 (2a)
[
CoIIÀporphyrin
1
2
1
2
3a)
ArR R CH ‡ ArN
3
À! ArR R CÀNHAr ‡ N
2
(1)
(2)
[
3
2
2b)6 (
3b)29 (
CoIIÀporphyrin
ArRHCÀNHAr ‡ ArN
À!ArRCˆNAr ‡ ArNH
‡ N
25 (2c)
28
16 (3c)
14.3 (3 h)
27.2 (3g)
3ꢀ.1( 3e)
3f)
± (3d)
19 (3a)
3b)2ꢀ (
22 (3c)
11.2 (3 h)
2ꢀ.1( 3g)
37.2 (3e)
13.4 (3f)
12.1 (3d)
29 (3a)
3b)24 (
22 (3c)
3 h)
3
2
2
[
[
[
[
Co(p-Cltpp)0
Co(p-Cltpp)0
Co(p-Cltpp)0
Co(p-Cltpp)0
H
CH
F
Br
CN
15
15
1ꢀ
7
7
3
23.8 (2 h)
18.7 (2g)
27.ꢀ (2e)
26.7 (2f)
38.ꢀ (2d)
21 2a)(
7 (2b)
33.4
27.9
3ꢀ.5
43.15.7 (
3
The reaction given in Equation (2) has been indepen-
dently observed by pre-forming the amines p-
YC H NHÀCH Ph and treating them with p-YC H N
±
[Co(p-Cltpp)0
5.8 (1d) 49.ꢀ
6
4
2
6
4
3
[
[
[
Co(p-MeOtpp)0 NO
Co(p-MeOtpp)0 OCH
Co(p-MeOtpp)0 Cl
2
±
±
±
±
±
±
±
28
1
36
39.ꢀ
35.9
25.4
39.4
(
Yˆ NO , Cl, OMe) in the presence of the catalyst. The
2
3
11.5
3
reaction quantitatively gives one equivalent of imine and
35 (2c)
one molecule of p-YC H NH per equivalent of amine,
6
4
2
[Co(p-MeOtpp)0 H
1ꢀ
21.ꢀ (2 h)
while no reaction is observed in the absence of the
catalyst.
Main byproducts of the reaction given in Equation (1)
are the azide-derived aniline and the diarylazo compound.
Other byproducts should be high-boiling materials, since
they could not be identified by gas-chromatography.
[Co(p-MeOtpp)0 CH3 12.5 18.1 (2g)
[Co(p-MeOtpp)0 F
6
4
2
91(1
1
4
24.8 (2e)
26.6 (2f)
3ꢀ.ꢀ (2d)
2a)
[
[
[
[
Co(p-MeOtpp)0 Br
Co(p-MeOtpp)0 CN
6.1( 1d) 48.5
Co(oep)0
Co(oep)0
NO
OCH
Cl
2
±
±
±
±
±
±
±
23
1
2ꢀ
3
5
2b)3 (
[Co(oep)0
[Co(oep)0
2ꢀ (2c)
II
H
11.5 18.2 (2 h)
5.8 ( 36.11
33.ꢀ
The following Co ± porphyrin complexes were tested
[
[
[
Co(oep)0
Co(oep)0
Co(oep)0
CH
F
Br
CN
3
13
5
21.7 (2g)
17.6 (2e)
27.2 (3g)
59 (3e)
9.7 (3f)
2.6 (3d)
as catalysts: [Co(tpp)0 (tpp ˆ dianion of meso-tetraphe-
17.6
32.9
nylporphyrin), [Co(p-MeOtpp)0 (p-MeOtpp ˆ dianion
3.5 25.8 (2f)
32.5 (2d)
of
meso-tetrakis-4-methoxyphenylporphyrin),
[Co-
[Co(oep)0
2
22.7 (1d) 38.5
(
p-Cltpp)0 (p-Cltpp ˆ dianion of meso-tetrakis-4-chloro-
[
1
a0 Experimental conditions: arylazide ˆ 1.4ꢀ mmol, molar ratio arylazide/Co ˆ
ꢀꢀ:1, in refluxing toluene (2ꢀ mL). [b0 Time required to reach the complete
phenylporphyrin), and [Co(oep)0 (oep ˆ dianion of beta-
conversion of the azide. [c0 Determined by gas-chromatographic analysis. [d0 Note
that for the imine the reaction requires two moles of azide per mole of the imine and
thus the maximum yield in imine is 5ꢀ%.
para position of the aromatic azide generally increased the
yields of the imine, except for the derivatives having the p-
NO group, which consistently gave lower selectivities than
2
expected based on the general trend. It should be noted that
imines with the p-nitro substituent hydrolyses very easily even
during the workup and gas-chromatographic analysis; this at
least partly explains the lower selectivity observed with
respect to that expected based on the high electron-with-
octaethylporphyrin). The results of a screening of the
reactions of toluene with a series of aryl azides bearing
different substituents in the para position and catalyzed by
several porphyrin complexes is reported in Table 1. The
results obtained with other benzylic compounds are reported
in Table 2.
Note that the stoichiometry in the reactions in Equa-
tions (1) and (2) implies that a maximum yield of 5ꢀ% can be
obtained with respect to the arylazide when the imine is the
product.
drawing power of the nitro group. If the data for the p-NO
2
2
group are neglected, a fair correlation (ꢀ.632 < R < ꢀ.831)
between the imine selectivity and the Hammet s constants is
observed for all catalysts and an even better correlation
2
‡
(ꢀ.721 < R < ꢀ.956) is observed with the s constants (Fig-
ure 1and Figures S 1± S4 in the Supporting Information).
Moreover, the slope of the lines derived from the linear
‡
regression of the selectivity versus s plots are quite close to
The reaction is sensitive to the nature of the azide
employed. With toluene as substrate, only in the case of p-
cyanophenylazide was the benzylamine observed among the
products (Table 1). In all other cases, the reaction in
Equation (2) was so fast relative to that in Equation (1) that
no benzylamine could be detected at any stage of the reaction.
With all catalysts, electron-withdrawing substituents in the
each other (2ꢀ.9, 21.5, 17.4, and 18.8 for [Co(tpp)0, [Co(p-
Cltpp)0, [Co(p-MeOtpp)0 and [Co(oep)0, respectively). It
should be stressed that, on theoretical grounds, no linear
correlation is in general to be expected between a selectivity
value and the s or s constants. However, the fact that such a
correlation is indeed observed and the similarity of the slopes
for different catalysts has some, at least qualitative, mecha-
‡
250
¹ 2ꢀꢀ3 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
ꢀ947-6539/ꢀ3/ꢀ9ꢀ1-ꢀ25ꢀ $ 2ꢀ.ꢀꢀ+.5ꢀ/ꢀ
Chem. Eur. J. 2003, 9, No. 1

Amination of Benzylic CÀH Bonds
249±259
2 6 4 3
Table 2. Catalytic amination of benzylic compounds by p-O NC H N .
Entry
Catalyst
Substrate
t [h0[a0
Imine
yield [%0[
Amine
yield [%0
p-Nitroaniline
yield [%0
Diazene 3a
yield [%0
b0
[b0
[b0
[b0
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
c0
c0
c0
c0
c0
c0
c0
c0
c0
c0
c0
c0
c0
c0
f0
[d0
[d0
[d0
1
2
3
4
5
6
7
8
9
ꢀ
1
2
3
4
5
6
7
8
9
[Co(tpp)0
[Co(oep)0
[Co(p-MeOtpp)0
[Co(p-Cltpp)0
[Co(tpp)0
isopropylbenzene
isopropylbenzene
isopropylbenzene
isopropylbenzene
sec-butylbenzene
sec-butylbenzene
sec-butylbenzene
sec-butylbenzene
cyclohexylbenzene
ethylbenzene
ethylbenzene
ethylbenzene
ethylbenzene
diphenylmethane
fluorene
1.5
6
2
5.5
4
5
±
±
±
±
±
±
±
±
±
15.8 (4)
23
28
24
2ꢀ
8
24
23
4ꢀ
8
14
5
15
21
1ꢀ
11
1ꢀ
[d0
[d0
[d0
27.2 ( 4)
[d0
[d0
[d0
42.3 (4)
[d0
[d0
[d0
52.7 (4)
[d0
[d0
[d0
15.ꢀ (5)
[d0
[d0
[d0
[Co(oep)0
13.6 (5)
[d0
[d0
[d0
[Co(p-MeOtpp)0
[Co(p-Cltpp)0
[Co(tpp)0
[Co(tpp)0
[Co(oep)0
[Co(p-MeOtpp)0
[Co(p-Cltpp)0
[Co(tpp)0
[Co(tpp)0
[Co(oep)0
5
3
1ꢀ.3 (5)
[d0
[d0
[d0
12.2 (5)
[d0
[d0
[d0
2.5
3
11
3.5
4
3
6
2.5
5
8
28.3 (6)
42.1
22.8
[e0
[e0
[e0
[e0
1
1
1
1
1
1
1
1
1
1
25 (7)
1
(8)
22 (8)
22 (8)
3
29
21
26
33
24
41
43
42
11
11
1ꢀ
[e0
[e0
[e0
[e0
8
(7)
[e0
[e0
[e0
[e0
24 (7)
[e0
[e0
[e0
2[e0
±
19 (7)
25 (8)
[e0
[e0
[e0
11.4 (9)
39.2 (11)
41.4 (11)
36.9 (11)
38.4 (11)
32[g0(12)
55.2 (10)
[d0
[d0
[d0
±
±
±
±
±
12
4
8
9
f0
[d0
[d0
[d0
fluorene
fluorene
fluorene
9,1ꢀ-Dihydroanthracene
f0
[d0
[d0
[d0
[Co(p-MeOtpp)0
[Co(p-Cltpp)0
[Co(tpp)0
f0
[d0
[d0
[d0
4ꢀ
46
f0
[e0
12
±
[
a0 Time required to reach the complete conversion of the aromatic azide. [b0 Calculated with respect to converted azide. Note that for the imine the reaction
requires two moles of azide per mole of the imine and thus the maximum selectivity in imine is 5ꢀ%. [c0 Reactions carried out in the substrate as solvent, Tˆ
1ꢀ8C with catalyst/azide ˆ 1:1ꢀꢀ. [d0 Determined by flash chromatography. [e0 Determined by gas-chromatographic analysis. [f0 Reactions carried out in
refluxing benzene, with catalyst/azide/hydrocarbon ˆ 1:1ꢀꢀ:1ꢀꢀ. [g0 Determined by gas-chromatographic analysis after hydrolysis of the diimine 12 to the
diketone 13. The diimine 12 can be reduced in the same pot to the diamine 14 by NaBH (28% total yield).
1
4
‡
Figure 1. Imine selectivity [%0 versus s constants of the R groups for the
reactions of toluene with p-RC , catalyzed by [Co(tpp)0. See Table 1
for experimental conditions. The line drawn corresponds to the best fit of
6
4 3
H N
‡
2
the equation: imine selectivity [%0 ˆ as ‡ b (a ˆ 2ꢀ.9, b ˆ 25.2, R ˆ
.759), for all substituents except NO
ꢀ
2
.
All catalysts gave comparable yields for the imine of
fluorene (Table 2), whereas [Co(p-Cltpp)0 gave the highest
yields of amine when isopropylbenzene was used as substrate,
although the time needed to reach a complete conversion of
the aromatic azide was 3.5 times longer than that required
when [Co(tpp)0 is used as catalyst. However, it should be
noted that for other substrates different orders of activity/
selectivity were observed for the same catalysts. For these
other substrates, data is insufficient for a statistical treatment,
but it is clear that any generalization is premature.
Steric hindrance lowers the selectivity, but does not prevent
the reaction and with sec-butylbenzene the amine was formed
in 1ꢀ ± 15% yield (Table 2). A better yield of 28.3% was
obtained for the cyclohexylbenzene derived amine.
If the amine is the desired product, in those cases in which
the imine is the only or predominant product, the latter can be
nistic implications that will be discussed later, apart from
being a useful predicting tool from a synthetic point of view.
The similarity of these plots also indicates that there is no
statistically significant difference among the results obtained
with the various porphyrins. Thus the identity of the porphyrin
appears to play only a minor role in the reaction.
reduced in situ by NaBH before isolation, as shown, for
4
example, by the case of the 9,1ꢀ-dihydroantracene-derived
imine.
Chem. Eur. J. 2003, 9, No. 1¹ 2ꢀꢀ3 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
ꢀ947-6539/ꢀ3/ -ꢀ ꢀ9$2ꢀ 521 1ꢀ .ꢀꢀ+.5ꢀ/ꢀ
251

S. Cenini et al.
FULL PAPER
The influence of the hydrocarbon concentration on the
selectivity has been investigated by employing toluene and p-
O NC H N as substrates, with benzene as a diluting solvent.
The results, reported in Table 3, clearly show that best results
are obtained in neat toluene, and increasing the benzene
groups of the ™ArN∫ moieties are equivalent, at least on the
H NMR timescale. An analogous result is observed when
[Co(p-MeOttp)0 is used in place of [Co(tpp)0. Examination of
the same reactions by UV-visible spectroscopy also indicates
the identity of the final products when the azide or the diazene
is employed as starting material. At least one complex has
1
2
6
4
3
been reported in the literature that has an unsubstituted
Table 3. Influence of the toluene concentration on the selectivity of the
catalyzed by [Co(tpp)0.[
6 4 3
H N
a0
diazene (HNˆNH) group bridging two ruthenium ± porphyrin
amination of toluene by p-O
2
NC
moieties.[
130
An analogous formulation as [{Co(porphy-
Toluene Toluene Imine 2a
PhCHO
yield [%0 yield [%0
p-O
2
NC
6
H
4
[b0
NH
2
Diazene 3a
yield [%0
rin)} (ArNˆNAr)0 (porphyrin ˆ tpp or p-MeOtpp, Arˆ p-
yield [%0[
b,c0
[b0
[b0
2
[
mL0
[m0
NO C H ) would also be consistent with our spectroscopic
2
6
4
3
5
7
2.82
4.69
6.57
9.39
12.8
17.1
27.13.9
29.5
1.8
2.4
2ꢀ.ꢀ
28.6
67.2
46.8
data, but all attempts to grow single crystals suitable for X-ray
diffraction always gave the Co ± porphyrin complex without
additional ligands. Thus it is possible that the observed shift in
the signals of the aryldiazene is only due to the proximity of
the paramagnetic cobalt complex, rather than to a true
coordination. If any coordination is present, it must be very
weak, as also indicated by the fact that addition of aryldiazene
to the catalytic reaction mixture does not slow down the
reaction (see later). An analogous outcome was observed
when [Co(tpp)0 was treated with p-MeOC H N . With the
33.6
39.2
1
ꢀ
4.5
34.5
36.ꢀ
[
a0 Experimental conditions: arylazide ˆ ꢀ.622 mmol, molar ratio aryla-
zide/Co ˆ 15:1, in toluene/benzene (total volume 1ꢀ mL) at 758C. Reac-
tions were run up to complete conversion of the azide. [b0 Determined by
gas-chromatographic analysis. [c0 Note that for the imine the reaction
requires two moles of azide per mole of the imine and thus the maximum
yield in imine is 5ꢀ%.
6
4
3
fraction at constant total volume decreases the imine (2a)
yield and increases that of the diazene (3a). For these
reactions, the amount of benzaldehyde formed by hydrolysis
of the imine during the reaction or the analysis has also been
measured. Some of these reactions were repeated to follow
the formation of the products as a function of time. In general,
the ratio between the relative formation rates for the different
products is approximately constant through all the reaction,
that is, no product is formed preferentially at the beginning of
the reaction (see Figure S5 in the Supporting Information for
one example).
sterically bulky azides Ph CN and adamantylazide no reac-
3
3
tion was observed. With p-CF C H N ; another reaction
3
6
4
3
prevailed, leading to fluorine loss from the CF group and
3
[140
formation of a complex that appears to be [Co(tpp)F0 based
on the spectroscopic data. This reaction was not investigated
further.
It should be noted that the signals due to the porphyrin in
1
the H NMR spectrum are immediately broadened when p-
O NC H N or p-MeOC H N (but not Ph CN or adaman-
2
6
4
3
6
4
3
3
3
tylazide) are added, regaining the original shape after the
azide has been completely consumed. This suggests a rever-
sible interaction of the azide with the cobalt complex.
The reaction of p-O NC H N with toluene catalyzed by
Since most metal ± porphyrin catalyzed amination reactions
reported up to date afford the sulfonylamide as product, we
2
6
4
3
also investigated the use of TsN and TsNˆIPh as aminating
3
Co ± porphyrin complexes is practically suppressed when
carried out in the presence of the spin-trap compound
TEMPO (2,2,6,6-tetramethyl-1-pyperidine-N-oxide). Since
there is no reaction between the Co ± porphyrin complexes
and TEMPO in the absence of the azide, this suggests that the
unpaired spin density of the Co ± porphyrin complex is partly
localized on the azide nitrogen atoms in the intermediate
agents. Tosylazide gave a very slow reaction. After 24 hours in
refluxing toluene ([Co(tpp)0 as catalyst, experimental con-
ditions as in Table 2), only a 68.5% conversion was observed.
The imine 16 was formed with a 28.2% selectivity, together
with the corresponding saturated amine 15 (21.5%) and
tosylamide (TsNH , 39.5%). Tosyliminoiodinane was even
2
less reactive and afforded only trace amounts (as detected by
GC-MS) of benzylic amine and imine; the only product
detected in large amounts was TsNH2.
[150
complex.
A kinetic study was also performed by employing p-
O NC H N and toluene as substrates and [Co(tpp)0 as
2
6
4
3
catalyst. The reaction was followed by IR spectroscopy,
À1
Mechanistic study: To clarify the mechanism of this reaction
we performed some stoichiometric reactions and a kinetic
study. No reaction is observed between [Co(tpp)0 and toluene
or between the azide and toluene alone, even at reflux
temperature. On the other hand, when [Co(tpp)0 was treated
with excess p-O NC H N at 758C in benzene, the diazoarene
monitoring the intensity of the 212ꢀ cm absorption of the
arylazide. All reactions showed an excellent first-order
dependence of the rate on azide concentration (R > ꢀ.98
2
for all reactions). The rate was unaffected by the addition of
the products of the reaction (imine, diazene, nitroaniline), an
indication that the observed kinetics is not an effect of
product inhibition; this is known in some cases to afford plots
easily mistaken for first-order behavior in one of the reagents.
The kinetics are also first-order with respect to [Co(tpp)0
(Figure S6, Supporting Information), but shows a saturation
behavior with respect to toluene concentration, the other
solvent being benzene (Figure 2). All kinetic experiments in
this paper were run at 758C to avoid boiling of benzene in
those reaction in which it was present.
2
6
4
3
p-O NC H NˆNC H NO -p was catalytically obtained. When
2
6
4
6
4
2
the same reaction was repeated at a lower temperature in
1
C D , new signals were observed in the H NMR spectrum
6
6
that cannot be attributed to any know compound, but are
identical to the ones of the product obtained by reaction of
[
Co(tpp)0 with p-O NC H NˆNC H NO -p. Observation of
2
6
4
6
4
2
only one set of resonances for the aryl groups of the
ArNˆNAr moiety of this complex indicates that the aromatic
252
¹ 2ꢀꢀ3 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
ꢀ947-6539/ꢀ3/ꢀ9ꢀ1-ꢀ252 $ 2ꢀ.ꢀꢀ+.5ꢀ/ꢀ
Chem. Eur. J. 2003, 9, No. 1

Amination of Benzylic CÀH Bonds
249±259
correlation quality by performing a dual parameter analysis
Á
2
introducing also a radical parameter (s , see later) left the R
JJ
value essentially unchanged; however, a better correlation
was observed if the Taft parameters,[ which allow a separate
160
description of field and resonance effects, were employed
170
(
Figure 4).[ The best fit for the equation: log(kArN3/kPhN3) ˆ
1Fs ‡ 1 s ‡ c was found for the values 1 ˆ 1.ꢀ86 and
F
R
R
F
2
1R ˆ ꢀ.2ꢀ9 (R ˆ ꢀ.882).
Figure 2. Influence of the toluene concentration on the value of kapp
À1 À1
[
s
m
0 in the equation: Àd[ArN
3
0/dt ˆ kapp[[Co(tpp)00[ArN 0 (Ar ˆ p-
3
O
2
6 4
NC H
À); Tˆ 758C. The line drawn corresponds to the best fit for
Equation (6).
When the selectivity was examined as a function of time for
different toluene/benzene mixtures the following observa-
tions were made.
Figure 4. Linear free energy correlation of log(kArN3/kPhN3) versus Taft s
F
1
) The amount of imine and p-nitroaniline grow at the same
rate, in accord with the stoichiometry reported in the
reactions in Equations (1) and (2).
and s
R
parameters. The best fit for the equation log(kArN3/kPhN3) ˆ 1
F
s
F
‡
2
1
R
s
R
‡ c was found for the values 1
F
ˆ 1.ꢀ86 and 1
R
ˆ ꢀ.2ꢀ9 (R ˆ ꢀ.882).
2
) The rate of formation of p-O NC H NˆNC H NO -p is
2
6
4
6
4
2
For the series of the substituted toluenes, we employed p-
proportional to the one of the imine for each reaction,
indicating the same dependency from the azide concen-
tration (for one example see Figure S5, supporting infor-
mation).
bromophenylazide as the aminating agent, because some of
the imines derived from p-nitrophenylazide hydrolyzed very
easily. However, even with p-bromophenylazide a small
amount of hydrolysis was observed. Since imine hydrolysis
produces the substituted toluene-derived aldehyde and this
aldehyde cannot form in any other way, the total amount of
imine initially formed was obtained by summing the moles of
imine detected by GC with those of the corresponding
aldehyde determined by the same technique. Benzylamine
was not detected in any of the reactions in this series, whereas
p-bromophenyldiazene was always the main byproduct; the
identity of which does not depend on the toluene derivative
employed. Given the nonlinear dependence of the rate on the
toluene concentration, competition reactions were run be-
tween the azide and an equimolar mixture of toluene and a
substituted toluene (Scheme 1).
3) The rate of formation (and the final selectivity, Table 3,
Figure 3) of imine increases with the toluene concentration
(
up to 1ꢀꢀ% toluene in the solvent mixture), while the one
of the diarylazo compound decreases in the same series,
indicating a common intermediate and a competition for
the formation of the two products.
Figure 3. Influence of the toluene concentration on reaction selectivity.
Selectivity ratio ˆ (imine ‡ aldehyde selectivity)/diazene selectivity. Rele-
vant data are reported in Table 3. The line drawn corresponds to the best fit
for the equation: selectivity ratio ˆ k
2
[toluene0/2k
3
with k
2
/2k
3
ˆ ꢀ.1ꢀ8
2
(
R ˆ ꢀ.965).
Two further kinetic series were measured. In the first,
toluene was kept constant as a substrate, and the reaction rate
for different azides was measured. In the second, differently
para-substituted toluenes were compared, keeping the azide
constant.
As far as the series of the azides is concerned, a fair
correlation was observed between the log of the ratio between
the rate of the amination reaction with the substituted azides
and the rate of the corresponding reaction with phenylazide
and the Hammet s constant (Figure S7, Supporting Informa-
2
tion), with 1 ˆ ꢀ.68 (R ˆ ꢀ.71). Attempts to improve the
Scheme 1.
Chem. Eur. J. 2003, 9, No. 1¹ 2ꢀꢀ3 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
ꢀ947-6539/ꢀ3/ -ꢀ ꢀ9 2ꢀ 51 3 $ 2ꢀ.ꢀꢀ+.5ꢀ/ꢀ
253

S. Cenini et al.
FULL PAPER
Since the amount of the two toluenes vary very little during
the reaction, the relative rate was determined by the ratio of
the products, as in Equation (3):
observed in the GC-MS spectrum of the solution after the
reaction. The primary isotopic effect determined, k /k ˆ 14
H
D
is again a value outside the normal range (usually 1± 7).
k
r
ˆ ([substituted imine0 ‡ [substituted benzaldehyde0)/
[unsubstituted imine0 ‡ [benzaldehyde0)
(
3)
(
Discussion
We note that this equation holds because no other toluene-
Synthetic aspects: In this paper we have reported on a new
reaction strategy for the amination of benzylic CÀH bonds,
derived product was obtained except for those mentioned. In
particular, no dibenzyl or other products derived from free-
radical reactions were observed in any case.
leading to amines and imines. Although much work has still to
be done before such a strategy will be fully developed, a few
points are worth being evidenced here.
No correlation at all is found between the log k values for
r
the substrates reported in Scheme 1and the Hammet s or the
While the synthesis of N-alkylanilines continues to attract
much attention,[ very few analogues of the N-tertiaryalkyl-
anilines here described have been reported in the literature,[
and the synthesis of several of them required several steps and
the use of reagents such as BuLi and PhLi, which are
incompatible with most functional groups.
‡
related s constants. In the literature, it has been reported that
2ꢀ0
the rates of analogous benzylation reactions by p-toluensul-
fonyliminoiodinane are sensitive to radical stabilization
effects, which are scarcely reproduced by the aforementioned
210
constants. A good correlation could be obtained[
4a,c,f0
by either
the use of a parameter called ™total effect∫ (TE),[ which
takes into accounts both polar and radical effects, or by a
multivariate analysis,[ in which a constant was employed
180
The synthesis of imines is generally performed by reaction
of amines or other compounds derived from them with
carbonyl compounds. The method gives excellent results in
some cases, but is unsatisfactory in others, most notably in
those in which the amine bears electron-withdrawing sub-
stituents. In contrast, imines of this kind are those for which
the best selectivities are observed with our method. More-
over, our reaction employs the hydrocarbon, rather then the
corresponding carbonyl compound, as a substrate. This point
is likely to make a difference when only the hydrocarbon is
available, since the selective oxidation of a benzylic CH2
group to carbonyl is not an easy reaction.
190
Á
that only describes the radical effects (s †, together with the
JJ
‡
standard Hammet s (or the modified s ) constant to describe
the polar contribution. In our system, use of the TE parameter
in place of s improved the correlation a bit, which remained
poor; however, an excellent agreement with the experimental
Á
data was obtained by use of the s ‡ s multivariate
JJ
correlation.
Á
Equation (4), with 1 ˆ 1.25, 1 ˆ ꢀ.ꢀꢀ961and c ˆ ꢀ.ꢀ322,
JJ
2
describes the experimental values with R ˆ ꢀ.9887 (Figure 5).
In situ reduction of imines to the corresponding amines by
NaBH4 and related reagents is widely reported in the
literature and also works well in the present system.
Á Á
logk
r
ˆ 1JJ
s
JJ ‡ 1s ‡ c
(4)
Concerning the substrates containing the tosyl group, there
is no evident reason for the low reactivity of tosylazide under
these conditions. A possible explanation is that one oxygen
atom of the sulfonyl group interacts with the cobalt center in
preference to the azide nitrogen, thus inhibiting the catalytic
reaction. The accumulation of an amount of benzylic amine
much larger than in any reaction with arylazides can be
attributed to the high electron-withdrawing power of the tosyl
group. The fast conversion of the benzylic amine to the imine
is clearly due to the activating power of an electron-rich
heteroatom directly attached to the benzylic carbon on the
reactivity of the remaining CÀH bonds. The presence of a
Figure 5. Linear free energy correlation of logk
r
[as defined in Eq. (4)0
Á
Á Á
versus (sJJ, s). The best fit for the equation logk
r
ˆ 1jj
s
JJ ‡ 1s ‡ c was
Á
2
found for the values 1JJ ˆ1.24797 and 1 ˆ ꢀ.ꢀꢀ961( R ˆ ꢀ.9887).
strongly electron-withdrawing tosyl group is expected to
decrease this activation. The very low reactivity of TsNˆIPh in
The overwhelming importance of the radical parameter is
completely unprecedented. Indeed, the contribution of the
polar parameter is so small (ꢀ.8% of that of the radical) that
the [Co(tpp)0-catalyzed reaction is surprising, since this
compound is the standard aminating agent in most metal-
catalyzed amination reactions of hydrocarbons. This inactivity
points to a different reaction mechanism for cobalt porphyrins
with respect to the one observed with other metals, in accord
with the following discussion.
Á
neglecting it completely (and fitting the data only to s † gives
JJ
2
an indistinguishably good fit (R ˆ ꢀ.9885 instead of ꢀ.9887,
Á
with 1 ˆ 1.26, Figure S8 in the Supporting Information). This
JJ
anomaly will be discussed later.
The isotopic effect on the benzylation reaction was
measured by running a reaction analogous to the preceding
ones, but with deuterated toluene in place of the substituted
toluene. Determination of the relative amounts of deuterated
products was made by integration of the masses of the peaks
Reaction mechanism: It is well known that arylazides
generate free nitrenes upon photochemical activation,[
and, when a metal is present, metal-nitrene (imido) inter-
mediates are generally supposed to be formed. However, in
the present case, the kinetic data discount this eventuality. In
220
254
¹ 2ꢀꢀ3 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
ꢀ947-6539/ꢀ3/ꢀ9ꢀ1-ꢀ254 $ 2ꢀ.ꢀꢀ+.5ꢀ/ꢀ
Chem. Eur. J. 2003, 9, No. 1

Amination of Benzylic CÀH Bonds
249±259
particular, a slow formation of an imido complex followed by
a fast reaction of this complex with toluene would give zero-
order kinetics with respect to toluene. On the other hand, a
fast formation of the imido complex followed by a slow
reaction with toluene would give zero-order kinetics with
respect to the azide. Moreover, in the latter case, the
intermediate complex should be at least spectroscopically
observable in solution during the reaction. None of these
conditions fit the experimental data.
amine can be formed both from the azide complex and the
imido complex. Again this would imply a higher selectivity in
aryldiazene at the beginning of the reaction.
The scenario in Scheme 2 implies that the rate of formation
of the imine is proportional to k [toluene0[[Co(tpp)(ArN )0
2
3
and the one of the diazene is proportional to 2k [[Co-
3
(tpp)(ArN )0. Since the concentration of the intermediate is
3
the same for both reactions, the ratio between the corre-
sponding selectivities will equal the one between the forma-
tion rates. Thus it follows that a linear relation is expected
between the ratio of the two selectivities and the toluene
concentration, as in Equation (6).
All of the kinetic data, on the other hand, can be explained
by the reaction pathway shown in Scheme 2. Application of
Imine selectivity/diazene selectivity ˆ k
2
[toluene0/2k
3
(6)
The results of the experiments run at different toluene
concentrations have been reported in Table 3. Since 2a partly
hydrolyzes even during the analysis, the selectivity of the
initially formed imine is considered to be the sum of the imine
and benzaldehyde selectivities. By plotting the ratio between
the (initial) selectivities in imine and diazene against the
molar concentration of toluene for the data in Table 3 (also
including the origin), a good linear correlation is indeed
Scheme 2.
2
observed with k /2k ˆ ꢀ.1ꢀ8 and R ˆ ꢀ.965 (Figure 3). This
2
3
further corroborates the kinetic model employed.
the steady-state approximation to the intermediate arylazide-
porphyrin complex leads to Equation (5):
It should be noted that, to the best of our knowledge, the
proposal of an arylazide complex as the aminating agent is
unprecedented; however, structurally related N-bound tosyl-
iminoiodinane complexes have been proposed on a kinetic
ground as the active intermediates in both the Mn/salen-
k
app ˆ 2k
1
(k
2
[toluene0 ‡ k
3
)/(k
2
[toluene0 ‡ k
3
‡ kÀ1
)
(5)
[230
catalyzed sulfimidation of sulfides and in the Mn/porphy-
rin-catalyzed aziridination of olefins.[ The latter intermedi-
ate was also proposed to be involved in the amidation of
saturated CÀH bonds by the same catalytic system, based on
240
which nicely fits the reaction rates observed at different
toluene concentrations, as shown in Figure 2. The fitting
À2 À1 À1
affords the value k ˆ 2.6  1ꢀ s m ; however, due to the
1
form of the equation and number of data points, different
the observation of a peak ascribable to the fragment
‡
groups of values for k , k , and k give equally good fittings
[Mn(porphyrin)(PhINTs)0 in the mass spectrum of the
2
3
À1
and the obtained values are not reliable.
In Scheme 2, a reversible interaction occurs between
reaction mixture of [Mn(porphyrin)(OH)(MeOH)0 with
PhINTs.[
30
[
Co(tpp)0 and the arylazide. The equilibrium is shifted
More information on the reactivity of the arylazide com-
plexes could be gained by the kinetic experiments run with
different azides and toluenes. Concerning the azides series,
the values obtained for the Hammet and Taft correlation
indicate that the nitrogen atom of the azide acts as an
electrophile in the transition state, which is not surprising. A
fair to good correlation is found even between the imine
towards the side of the starting materials, in accord with the
fact that the complex is not observable. This complex can
either react with toluene, to afford the benzylamine, or lose
dinitrogen in an unimolecular reaction, to afford another
intermediate, formulated as a true imido complex, that is
responsible for the formation of the aryldiazene in a fast
reaction. Formation of the aryldiazene by reaction of a second
molecule of azide directly with the same azide complex that
reacts with toluene is not consistent with the experimental
data. Indeed, such a scenario would imply a non-first-order
kinetics with respect to the azide for the formation of the
diazene; this would alter the global kinetics observed. More-
over, the concentration of toluene is approximately constant
during the reaction, whereas the one of the azide decreases.
Thus a direct competition of toluene and azide for the same
intermediate would result in a preferential formation of the
aryldiazene at the beginning of the reaction, whereas the
selectivities in aryldiazene and imine are constant during the
reaction (Figure S5, Supporting Information). The constant
selectivity during the reaction also excludes that the benzyl-
‡
selectivities and the s and s constants. As previously
mentioned, no such a linear correlation is to be expected on
theoretical grounds. However, the fact that it is observed and
the positive slope of both the rate and selectivity correlations
indicate that the arylazides reacting faster are also the ones
affording the highest selectivities. That both rate and selec-
tivity are favored by electron-withdrawing substituents sug-
gests that the rate-determining step is also the product-
determining one. This is in full agreement with the mechanism
proposed in Scheme 2 and the mathematical fitting derived
from it. Indeed, a faster reaction of the arylazide complex with
toluene not only will accelerate the global reaction rate by
increasing k , but will also increase the selectivity by increas-
2
ing the ratio k /k . Evidently, k is probably also affected by the
2
3
3
Chem. Eur. J. 2003, 9, No. 1¹ 2ꢀꢀ3 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
ꢀ947-6539/ꢀ3/ -ꢀ ꢀ9 2ꢀ 51 5 $ 2ꢀ.ꢀꢀ+.5ꢀ/ꢀ
255

S. Cenini et al.
FULL PAPER
identity of the azide, but it appears to be lesser so than k . The
2
‡
close similarity of the slopes of the selectivity versus s
correlations for different catalysts also strongly indicates
that the reaction mechanism for all the porphyrin complexes is
the same and the validity of Scheme 2 is not limited to
[Co(tpp)0.
In contrast to the arylazides series, the result of the fit for
the toluenes series was completely unexpected. As mentioned
before, an excellent correlation was found with a radical
parameter alone, with a statistically insignificant contribution
of a polar parameter. This situation is anomalous, since the
literature data[ indicate that even in the case of typical
Figure 6. Proposed transition state for the amination reaction.
190
.
radical reactions (e.g., the hydrogen abstraction by a Br
It must be noted that an increase in the steric hindrance
increases the importance of the quantum-mechanical
tunneling effect in those systems in which it is present.[
Using a terminology perhaps not completely rigorous, but
chemically intuitive, we can say that steric hindrance
prevents the reacting molecules from approaching each
other up to the correct distance for the hydrogen-atom
transfer to occur in a ™classical∫ way. Thus the hydrogen
atom has to do a ™long jump∫ that is characterized by a
high-energy barrier, since during the travel it loses most of
the stabilization due to the breaking bond, without gaining
significantly through the new bond formation. This may
well be the case in our system.
) Apart from the transferred hydrogen atom, the rest of the
toluene molecule remains relatively far from the polar
nitrogen atom. Since the magnitude of charge interactions
depends on the square of the distance, the porphyrin ±
azide moiety will have only a low polarizing effect on the
benzyl group, making charge stabilization on the latter of
negligible importance.
) In the intermediate arylazide complex, part of the
unpaired spin density of the paramagnetic cobalt(ii) center
is likely to be delocalized on the azide ligand. This may
explain the inhibition by the radical trap TEMPO, despite
that fact that an irreversible reaction between the Co ±
porphyrin complex and TEMPO in the absence of any
arylazide is not observed.
The only part of the mechanism for which data is
insufficient to give a complete picture of the intermediates
involved is the one concerning the reaction of the putative
radical or by N-bromosuccinimide) a polar contribution is
always present and is generally larger than the radical one.
Interestingly, in three cases that we are aware of in which the
radical parameter has been reported to be larger than the
polar one, the reactions involved are related to the one here
discussed, namely, the amidation of saturated CÀH bonds by
310
[30
[4f0
PhIˆNTs catalyzed by manganese and ruthenium com-
plexes and the copper-catalyzed aziridination reaction of
[250
substituted styrenes by PhIˆNTs. However the difference
among the absolute values of the two parameters was quite
small in all of these cases. It is worth remembering that the
first of these catalytic systems[30 is one of those for which a
metal ± PhINTs complex was proposed as the active aminating
species.
The anomalous importance of radical stabilization effects in
the toluene series is paralleled by the very high isotope effect
3
4
(
k /k ˆ 14). This value is clearly higher even of that
H
D
corresponding to the complete loss of the zero-point energy
[260
(
about 7 at 258C). However, even larger isotopic effects
have been previously observed and attributed to a relevant
role played by the quantum-mechanical tunneling effect in the
reactions under examination.[270 In general, k /k > 1 2 at 258C
H
D
are considered to be a reliable evidence for tunneling,[
27, 280
and the ratio over which tunneling is indicated decreases with
an increase in temperature (our value of 14 was measured at
758C). The quantum-mechanical tunneling effect can be
observed when the reaction involves the transfer of an
hydrogen atom that should pass through a very high energy
barrier. In these cases, the mass of the atom to be transferred
assumes a much greater importance than that the same mass
has on the zero-point energy of the bond to be broken. Thus
™
nitrene∫ intermediate. A speculative explanation of why it
1
only reacts with more arylazide and not with toluene is given
in the Supporting Information.[
an H atom can pass the barrier by tunnel effect much more
320
easily than a deuterium atom will, giving rise to a high
observed isotopic effect, which may even be larger than 5ꢀ.[
All of the reactivity and kinetic data point to the transition
state shown in Figure 6. For a discussion of alternative
coordination modes for the arylazide, see the Supporting
Information.[ The transition state proposed explains all of
the kinetic data observed, including the unusual ones. Indeed:
290
Conclusion
3ꢀ0
In this paper, we have reported on a new catalytic system for
the amination and imination of saturated benzylic CÀH bonds
1
2
2
) The nitrogen atom in the a-position with respect to the aryl
ring of the arylazide acts as an electrophile, in accord with
the reactivity order of several azides with respect to the
Hammet and Taft constants of the para substituents.
) The ™pocket∫ conformation of the arylazide makes steric
hindrance very severe for the incoming toluene molecule
and prohibits a too close approach of the whole molecule.
by arylazides. Both the synthetic and mechanistic aspects of
the reaction have been investigated and some unusual
findings have been explained. Yields are in most cases only
moderate; however, alternative synthetic strategies require
the availability of prefunctionalized starting materials or
several steps to reach the same compounds and the total yields
are likely to be even lower.
56
¹ 2ꢀꢀ3 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
ꢀ947-6539/ꢀ3/ꢀ9ꢀ1-ꢀ256 $ 2ꢀ.ꢀꢀ+.5ꢀ/ꢀ
Chem. Eur. J. 2003, 9, No. 1

Amination of Benzylic CÀH Bonds
249±259
1
Data for compound 8: H NMR (3ꢀꢀ MHz, CDCl
3
, TMS): d ˆ 8.ꢀ3 (d,
Experimental Section
3
3
J(H,H) ˆ 8.ꢀ Hz, 2H), 7.42 ± 7.25 (m, 5H), 6.48 (d, J(H,H) ˆ 8.ꢀ Hz, 2H),
3
3
4
.63 (s, 1H; NH), 4.59 (q, J ˆ 6.8 Hz, 1H; CH), 1.58 ppm (d, J(H,H) ˆ
.8 Hz, 3H; CH ); MS (7ꢀ eV, EI): m/z: 242, 227, 2ꢀ7, 181, 1ꢀ5, 91, 77.
General procedure: Unless otherwise specified, all reactions and manip-
ulations were performed under an N atmosphere by using standard
6
3
2
Data for compound 9: 1H NMR (3ꢀꢀ MHz, CDCl
, TMS): d ˆ 8.ꢀ3 (d,
Schlenk apparatus, cannula techniques, and magnetic stirring. Solvents
were dried and distilled by standard procedures and stored under
3
3
J(H,H) ˆ 8.4 Hz, 2H), 7.76 (m, 2H), 7.54- 7.28 (m, 6H), 7.15 (m, 2H),
3
[
330
[340
6.81ppm (d, J(H,H) ˆ 8.4 Hz, 2H).
dinitrogen. Arylazides,
tpp, p-Cltpp, p-MeOtpp
and their cobalt
[
350
1
complexes were synthesized by methods reported in the literature. All
other starting materials, including [Co(oep)0, were commercial products
and were used as received. Concerning the products, all anilines and
azobenzene (3h) are commercially available. The following products have
been previously reported in the literature and have been characterized by
comparison of their analytical data with the one reported in the literature
Data for compound 10: H NMR (3ꢀꢀ MHz, CDCl
3
, TMS): d ˆ 8.ꢀ6 (d,
3
3
J(H,H) ˆ 8.2 Hz, 2H), 7.39 ± 7.28 (m, 1ꢀH), 6.5ꢀ (d, J(H,H) ˆ 8.2 Hz, 2H),
3
3
5.63 (d, J(H,H) ˆ 3.6 Hz, 1H; CH), 5.ꢀꢀ ppm (d, J(H,H) ˆ 3.6 Hz, 1H;
NH).
Data for compound 11: 1H NMR (3ꢀꢀ MHz, CDCl
3
, TMS): d ˆ 8.32 (d,
3
J(H,H) ˆ 8.2 Hz, 2H), 7.88 (m, 1H), 7.63 (m, 2H), 7.51 (m, 1H), 7.39 (m,
[
360
[370
[380
or obtained from independently synthesized samples: 1a, 1b, 1c,
3
2
H), 7.11 (d, J(H,H) ˆ 8.2 Hz, 2H), 6.97 (m, 1H), 6.58 ppm (m, 1H);
[
390
[4ꢀ0
[410
[420
[430
[440
[420
[450
[460
[470
[460
1
3
d, 1e, 1 f, 1g, 1h, 2a, 2b,c, 2d, 2e, 2 f, 2g,[420 2h,
13
C NMR (3ꢀꢀ MHz, CDCl
3
): d ˆ 163.77, 158.4ꢀ, 144.81, 142.68, 137.37,
a, 3b ± d, 3e, 3 f, 3g, 4,[520 7,
[
480
[490
[5ꢀ0
[490
[510
[12a, 530
8,[540 9,[550 10, 11,
[560
[53b0
12,[570
1
[
35.37, 133.3ꢀ, 129.74, 126.12, 124.95, 119.38 ppm; MS (CI): m/z: 3ꢀ1
[
580
[590
[610
[620
[630
1
5, 16, 17a,[6ꢀ0 17b, 17c, 17d, 17e,[640 17 f.[650 In the following, the
‡
M ‡10; elemental analysis calcd (%) for C19
H
12
N
2
O
2
(3ꢀꢀ.3): C 75.99, H
1
H NMR spectra of some of them are reported when these data were not
available in the literature or when our spectra were of significantly better
quality than those reported. Gas chromatographic analyses were per-
formed on a Dani 861ꢀ capillary gas chromatograph equipped with a
4
.ꢀ3, N 9.33; found: C 75.86, H 4.3ꢀ, N 9.11.
1
Data for compound 14: H NMR (3ꢀꢀ MHz, CDCl
3
, TMS): d ˆ 8.ꢀ3 (m,
4
H), 7.47 ± 7.28 (m, 8H), 6.62 (m, 4H), 5.63 (s, 2H; NH), 5.28 ppm (s, 2H;
CH); C NMR (3ꢀꢀ MHz, CDCl
13
3
): d ˆ 153.45, 137.ꢀ3, 136.38, 128.86,
PS 255 column. R
i
values (R
i
ˆ response factor, relative to naphthalene as
‡
1
28.25, 127.2ꢀ, 126.ꢀ3, 112.6ꢀ ppm; MS (CI): m/z: 453 [M ‡10; elemental
an internal standard) were determined by the use of solutions of known
concentrations of the compounds. GC-MS analyses were performed on a
Shimadzu GCMS-5ꢀꢀꢀ instrument. NMR spectra were recorded on a
Bruker AC 3ꢀꢀ FT (3ꢀꢀ MHz). IR spectra were recorded on a FTS-7 Bio
Rad FT-IR spectrometer. Elemental analyses and mass spectra were
recorded in the analytical laboratories of Milan University.
analysis calcd (%) for C26
H
2ꢀ 4 4
N O (452.5): C 69.ꢀ2, H 4.46, N 12.38; found:
C 68.78, H 4.13, N 11.97.
1
Data for compound 17d: H NMR (3ꢀꢀ MHz, CDCl
3
, TMS): d ˆ 8.49 (s,
3
3
1
H), 8.ꢀ1(d, J(H,H) ˆ 8.2 Hz, 2H), 7.79 (d, J(H,H) ˆ 8.2 Hz, 2H), 7.54 (d,
3
3
J(H,H) ˆ 8.6 Hz, 2H), 7.13 ppm (d, J(H,H) ˆ 8.6 Hz, 2H).
1
Data for compound 17e: H NMR (3ꢀꢀ MHz, CDCl
3
, TMS): d ˆ 8.41(s,
Catalytic reactions: A two-necked flask equipped with a side port and a
reflux condenser with a nitrogen inlet on top and kept under N
3
1
(
H), 7.91 (m, 2H), 7.53 (d, J(H,H) ˆ 8.6 Hz, 2H), 7.18 (m, 2H), 7.ꢀ9 ppm
2
was
3
d, J(H,H) ˆ 8.6 Hz, 2H).
1
charged with the solid reagents in a N stream and then the liquid reagents
2
Data for compound 17 f: H NMR (3ꢀꢀ MHz, CDCl
3
, TMS): d ˆ 8.4ꢀ (s,
and, if required, the additional solvent was added at RT. The flask was then
heated to the desired temperature by a preheated oil bath. For reactions
run at reflux, the temperature of the oil bath was set ten degrees higher
than the boiling temperature of the solvent. The consumption of the
arylazide was monitored by TLC up to the point in which its spot was no
longer observable and then by IR spectroscopy. The reaction was
considered to be finished when the absorbance of the azide group in the
IR spectrum of the solution measured with a ꢀ.5 mm thick cell was at or
below ꢀ.ꢀ3. The so-obtained solution was either analyzed by GC or
evaporated to dryness in vacuo and the products separated by flash
chromatography on silica. Reagents amounts and reaction times are given
in the tables.
3
3
1
H), 7.78 (d, J(H,H) ˆ 8.4 Hz, 2H), 7.63 (d, J(H,H) ˆ 8.4 Hz, 2H), 7.52 (d,
3
3
J(H,H) ˆ 8.6 Hz, 2H), 7.1ꢀ ppm (d, J(H,H) ˆ 8.6 Hz, 2H).
Kinetic measurements: The catalyst, the azide, and toluene were added in
2
this order to a Schlenk flask under N . When required, benzene or a
substituted toluene were added at this point. The flask was capped with a
rubber septum and immediately placed in an oil bath preheated at 758C.
The solution was stirred for one minute to dissolve all reagents and ꢀ.2 mL
was withdrawn for IR analysis. The consumption of the azide was followed
À1
by IR spectroscopy (nN3 ˆ 212ꢀ cm ) by withdrawing samples of the
solution at regular times. Since all reactions were run in the presence of a
large excess of toluene, the apparent first-order constants were fitted to the
Data for compound 4: 1H NMR (3ꢀꢀ MHz, CDCl
equation Àd[ArN
3
0/dt ˆ kapp[[Co(tpp)00[ArN
3
0. The concentration of
3
, TMS): d ˆ 7.94 (d,
3
3
3
[Co(tpp)0 was calculated on the exact amount of catalyst weighed in each
run and was considered to remain constant during the reaction.
J(H,H) ˆ 8.2 Hz, 2H), 7.44 (d, J(H,H) ˆ 7.6 Hz, 2H), 7.37 (dd, J(H,H) ˆ
4
3
4
7
1
CH
1
2
7
.6 Hz, J(H,H) ˆ 1.9 Hz, 2H), 7.28 (dd, J(H,H) ˆ 7.6 Hz, J(H,H) ˆ 1.9 Hz,
3
H), 6.3ꢀ (d, J(H,H) ˆ 8.2 Hz, 2H), 5.ꢀ7 (s, 1H; NH), 1.73 ppm, (s, 6H;
2 6 4 3
Stoichiometric reaction of [Co(tpp)] with p-O NC H N
); 13C NMR (3ꢀꢀ MHz, CDCl
): d ˆ 168.22, 151.96, 145.78, 132.86,
1
À2
3
3
H NMR analysis: [Co(tpp)0 (1ꢀ.3 mg, 1.5 1ꢀ mmol) and p-nitrophenyl-
azide (2.7 mg, 1.5 1ꢀÀ2 mmol) were dissolved in C D (1mL) at 6ꢀ 8C. The
29.34, 129.22, 127.5ꢀ, 126.12, 125.68, 3ꢀ.77 ppm; MS (7ꢀ eV, EI): m/z: 256,
41, 195, 119, 91, 77; elemental analysis calcd (%) for C15 (256.3): C
ꢀ.29, H 6.29, N 1ꢀ.93; found: C 7ꢀ.ꢀ3, H 6.66, N 1ꢀ.58.
6
6
H
16
N
2
O
2
consumption of the arylazide was monitored by recording an
NMR spectrum every 5 min. New signals arose in the NMR
1
H
Data for compound 5: 1H NMR (3ꢀꢀ MHz, CDCl
(3ꢀꢀ MHz, C D , 6ꢀ8C) spectra at d ˆ 15.64 (brs, 8H), 12.92 (brs, 8H),
, TMS): d ˆ 7.9ꢀ (d,
6
6
3
1
3
3
9.56 (m, 12H), 7.82 (m, 4H), 7.41 (m, 4H). The H NMR spectrum was
identical to that of the product obtained by reaction of [Co(tpp)0 (17.3 mg,
J(H,H) ˆ 8.ꢀ Hz, 2H), 7.4ꢀ ± 7.29 (m, 5H), 6.28 (d, J(H,H) ˆ 8.ꢀ Hz, 2H),
3
4
ꢀ
.98 (s, 1H; NH), 1.97 (m, J(H,H) ˆ 8.2 Hz, 2H; CH
2
), 1.69 (s, 3H; CH
3
),
):
À2
À2
.85 ppm (t, 3J(H,H) ˆ 7.4 Hz, 3H; CH
); C NMR (3ꢀꢀ MHz, CDCl
13
6 6
2.6 1ꢀ mmol) with p-nitrophenyldiazene (3.7 mg, 1.3 1ꢀ mmol) in C D
3
3
(
1mL) at 6ꢀ 8C.
d ˆ 152.68, 149.45, 137.25, 13ꢀ.1ꢀ, 127.39, 126.27, 126.17, 113.84, 59.81, 36.91,
‡
2
5.4ꢀ, 8.55 ppm; MS (CI): m/z: 271[ M ‡10; elemental analysis calcd (%)
UV/Vis analysis: The reaction mixture of [Co(tpp)0 (1ꢀꢀ mg, ꢀ.15 mmol)
and p-nitrophenylazide (25.7 mg, ꢀ.16 mmol) in benzene (3ꢀ mL) was
heated at 8ꢀ8C until complete consumption of the arylazide (UV/vis signal
at 312 nm). A new signal at 344 nm was observed. Again the same signal is
observed when [Co(tpp)0 and p-nitrophenyldiazene are reacted with each
other.
for C16
18 2 2
H N O (27ꢀ.3): C 71.ꢀ9, H 6.71, N 1ꢀ.36; found: C 71.42, H 6.53, N
1
ꢀ.27.
Data for compound 6: 1H NMR (3ꢀꢀ MHz, CDCl
3
, TMS): d ˆ 7.92 (d,
3
3
J(H,H) ˆ 7.9 Hz, 2H), 7.46 (m, 2H), 7.31(m, 3H), 6.31(d,
J(H,H) ˆ
7
ꢀ
1
2
.9 Hz, 2H), 4.96 (s, 1H; NH), 2.28 (m, 2H), 1.94 ± 1.59 (m, 6H),
1
3
.9ꢀ ppm (m, 2H); C NMR (3ꢀꢀ MHz, CDCl
3
): d ˆ 152.31, 147.ꢀ5,
38.29, 127.41, 126.98, 126.17, 125.77, 113.93, 58.76, 36.69, 3ꢀ.11,
2.23 ppm; MS (7ꢀ eV, EI): m/z: 296, 253, 159, 91; elemental analysis calcd
(
2ꢀ 2 2
%) for C18H N O (296.4): C 72.95, H 6.8ꢀ, N 9.45; found: C 73.ꢀ8, H 6.45,
N 9.68.
Acknowledgement
Data for compound 7: 1H NMR (3ꢀꢀMHz, CDCl
, TMS): d ˆ 8.26 (d,
3
3
3
J(H,H) ˆ 8.ꢀ Hz, 2H), 7.35 ± 7.19 (m, 5H), 6.89 (d, J(H,H) ˆ 8.ꢀ Hz, 2H),
We thank MIUR (Programmi di Ricerca Scientifica di Rilevante Interesse
Nazionale) for financial support.
2
.26 ppm (s, 3H; CH ); MS (7ꢀ eV, EI): m/z: 24ꢀ, 225, 2ꢀ7, 179, 163, 117, 76.
3
Chem. Eur. J. 2003, 9, No. 1¹ 2ꢀꢀ3 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
ꢀ947-6539/ꢀ3/ -ꢀ ꢀ9 2ꢀ 51 7 $ 2ꢀ.ꢀꢀ+.5ꢀ/ꢀ
257

S. Cenini et al.
FULL PAPER
[
10 a) M. Johannsen, K. A. J˘rgensen, Chem. Rev. 1998, 98, 1689 ± 17ꢀ8
and references therein; b) Modern Amination Methods (Ed.: A.
Ricci), Wiley-VCH, Weinheim, 2000.
[210 a) N. C. Faibish, Y. S. Park, S. Lee, P. Beak, J. Am. Chem. Soc. 1997,
119, 11561 ± 1157ꢀ; b) H. J. Koh, H. W. Lee, I. Lee, J. Chem. Soc.
Perkin Trans. 2 1994, 125 ± 129; c) R. Chaabouni, A. Laurent, Bull.
Soc. Chim. Fr. 1973, 268ꢀ ± 2685; d) M. O. Abdel-Rahman, M. N.
Aboul-Enein, G. H. Tamman, Egypt. J. Chem. 1975, 18, 581± 586;
Chem. Abstr. 1978, 89, 197ꢀ17c.
[220 a) M. F. Budyka, M. M. Kantor, M. V. Alfimov Russ. Chem. Rev. 1992,
61, 25 ± 39; b) E. F. V. Scriven, K. Turnbull, Chem. Rev. 1988, 88, 297 ±
368.
[
20 a) R. Breslow, S. H. J. Gellman, J. Chem. Soc. Chem. Commun. 1982,
1
4ꢀꢀ ± 14ꢀ1; b) E. W. Svastits, J. H. Dawson, R. Breslow, S. H. Gell-
man, J. Am. Chem. Soc. 1985, 107, 6427 ± 6428; c) J.-P. Mahy, G. Bedi,
P. Battioni, D. Mansuy, Tetrahedron Lett. 1988, 1927 ± 193ꢀ; d) J.-P.
Mahy, G. Bedi, P. Battioni, D. Mansuy, New J. Chem. 1989, 13, 651±
6
57; e) I. Nageli, C. Baud, G. Bernardinelli, Y. Jaquier, M. Moran, P.
Muller, Helv. Chim. Acta, 1997, 80, 1ꢀ87 ± 11ꢀ5; f) Y. Kohmura, T.
Katsuki, Tetrahedron Lett. 2001, 42, 3339 ± 3342; g) J. Yang, R.
Weinberg, R. Breslow, Chem. Commun. 2000, 531± 532.
[230 C. Ohta, T. Katsuki, Tetrahedron Lett. 2001, 42, 3885 ± 3888.
[240 T.-S. Lai, H.-L. Kwong, C.-M. Che, S.-M. Peng, Chem. Commun. 1997,
2373 ± 2374.
[
[
30 a) X.-G. Zhou, X.-Q.Yu, J.-S. Huang, C.-M. Che, Chem. Commun.
[250 M. M. Diaz-Requejo, P. J. P e¬ rez, M. Brookhart, J. L. Templeton,
Organometallics 1997, 16, 4399 ± 44ꢀ2.
[260 F. H. Westheimer, Chem. Rev. 1961, 61, 265 ± 273.
[270 L. Melander, W. H. Saunders, Jr., Reaction Rates of Isotopic Mole-
cules, Wiley-Interscience, New York, 1980.
1
999, 2377 ± 2378; b) J.-L. Liang, J.-S. Huang, X.-Q. Yu, N. Zhu, C.-M.
Che, Chem. Eur. J. 2002, 8, 1563 ± 1572.
40 a) S.-M. Au, J.-S. Huang, W.-Y. Yu, W.-H. Fung, C.-M. Che, J. Am.
Chem. Soc. 1999, 121, 912ꢀ ± 9132; b) X.-Q. Yu, J.-S. Huang, X.-G.
Zhou, C.-M. Che, Org. Lett. 2000, 2, 2233 ± 2236; c) S.-M. Au, J.-S.
Huang, W.-Y. Yu, C.-M. Che, W.-Y. Yu, J. Org. Chem. 2000, 65, 7858 ±
[280 M. Amin, R. C. Price, W. H. Saunders, Jr., J. Am. Chem. Soc. 1990,
112, 4467 ± 4471.
7
864; d) C.-M. Che, W.-Y. Yu, Pure Appl. Chem. 1999, 71, 281± 288;
[290 a) In porphyrin chemistry, isotopic effects higher than seven have
been reported for several oxidation reactions by porphyrinoxoruthe-
e) S.-M. Au, S.-B. Zhang, W.-H. Fung, W.-Y. Yu, C.-M. Che, K.-K.
Cheung, Chem. Commun. 1998, 2677 ± 2678; f) J.-L. Liang, W.-Y. Yu,
C.-M. Che, Chem. Commun. 2002, 124 ± 125.
[
29b0
nium complexes.
In at least one case, an isotopic effect of eleven
was observed for an amidation reaction.[ In all of these cases, the
effect was said to be indicative of an hydrogen abstraction mechanism,
but the possible involvement of a quantum mechanical tunnel effect
was not discussed. On the other hand, the importance of tunneling
effect in related oxidation reactions by iron porphyrin has been
recognized;[ b) C.-M. Che, V. W.-W. Yam, Adv. Inorg. Chem. 1992,
39, 233 ± 325 and references therein; c) K. Yoshizawa, Coord. Chem.
Rev. 2002, 226, 251± 259 and references therein.
4a0
[
50 a) S. Cenini, A. Penoni, S. Tollari, J. Mol. Catal. A 1997, 124, 1ꢀ9 ± 113;
b) D. P. Albone, P. S. Aujla, P. C. Taylor, J. Org. Chem. 1998, 63, 9569 ±
9
3
571; c) B. M. Chanda, R. Vyas, A. V. Bedekar, J. Org. Chem. 2001, 66,
ꢀ ± 34.
29c0
[
[
[
60 Y. Kohmura, K. Kawasaki, T. Katsuki, Synlett 1997, 1456 ± 1458.
70 P. J. Stang, V. V. Zhdankin, Chem. Rev. 1996, 96, 1123 ± 1178.
80 a) S. Cenini, F. Ragaini, S. Tollari, D. Paone, J. Am. Chem. Soc. 1996,
1
18, 11964 ± 11965; b) F. Ragaini, S. Cenini, S. Tollari, G. Tummolillo,
[3ꢀ0 a) M. G. Fickes, W. M. Davis, C. C. Cummins, J. Am. Chem. Soc. 1995,
117, 6384 ± 6385; b) G. Proulx, R. G. Bergman, J. Am. Chem. Soc. 1995,
117, 6382 ± 6383; c) T. A. Hanna, A. M. Baranger, R. G. Bergman,
Angew. Chem. 1996, 108, 693 ± 696; Angew. Chem. Int. Ed. Engl. 1996,
35, 653 ± 655; d) G. Proulx, R. G. Bergman, Organometallics 1996, 15,
684 ± 692; e) M. Barz, E. Herdtweck, W. R. Thiel, Angew. Chem. 1998,
110, 238ꢀ ± 2383; Angew. Chem. Int. Ed. 1998, 37, 2262 ± 2265; f) G.
Guillemot, E. Solari, C. Floriani, C. Rizzoli, Organometallics 2001, 20,
6ꢀ7 ± 615; g) H. V. R. Dias, S. A. Polach, S.-K. Goh, E. F. Archibong,
D. S. Marynick, Inorg. Chem. 2000, 39, 3894 ± 39ꢀ1; h) I. Foch, L.
P a¬ rk a¬ nyi, G. Besenyei, L. I. Sim a¬ ndi, A. K a¬ lm a¬ n, J. Chem. Soc. Dalton
Trans. 1999, 293 ± 299; i) G. Besenyei, L. P a¬ rk a¬ nyi, I. Foch, L. I.
Sim a¬ ndi, Angew. Chem. 2000, 112, 986 ± 988; Angew. Chem. Int. Ed.
2000, 39, 956 ± 958.
R. Beltrami, Organometallics 1999, 18, 928 ± 942; c) F. Ragaini, S.
Cenini, E. Borsani, M. Domp e¬ , E. Gallo, M. Moret Organometallics
2
001, 20, 339ꢀ ± 3398.
90 S. Cenini, S. Tollari, A. Penoni, C. Cereda, J. Mol. Catal. A, 1999, 37,
35 ± 146.
[
1
[
[
[
1ꢀ0 E. Gallo, F. Ragaini, C. Perrone, A. Penoni, S. Tollari, S. Cenini,
unpublished results.
110 S. Cenini, E. Gallo, A. Penoni, F. Ragaini, S. Tollari Chem. Commun.
2
000, 2265 ± 2266.
120 a) M. C. Carreno, J. M. Cuerva, M. Ribagorda, A. M. Echavarren,
Angew. Chem. 1999, 111, 1552 ± 1555; Angew. Chem. Int. Ed. 1999, 38,
1
1
449 ± 1452; b) M. Tokunaga, M. Eckert, Y. Wakatsuki, Angew. Chem.
999, 111, 3416 ± 3419; Angew. Chem. Int. Ed. 1999, 38, 3222 ± 3225
and references therein.
130 J. P. Collman, J. E. Hutchison, M. S. Ennis, M. A. Lopez, R. Guilard, J.
Am. Chem. Soc. 1992, 114, 8ꢀ74 ± 8ꢀ8ꢀ.
140 V. Briois, C. Cartier, M. Momenteau, P. Maillard, J. Zarembowitch, E.
Dartyge, A. Fontaine, G. Tourillon, P. Thuery, M. Verdaguer J. Chim.
Phys. Phys.-Chim. Biol. 1989, 86, 1623 ± 1634.
[310 a) E. S. Lewis, H. L. Funderburk, J. Am. Chem. Soc. 1967, 89, 2322 ±
2327; b) B. T. Farrer, H. H. Thorp, Inorg. Chem. 1999, 38, 2497 ± 25ꢀ2.
[320 R. D. Webster, G. A. Hearth, A. M. Bond, J. Chem. Soc. Dalton Trans.
2001, 3189 ± 3195.
[330 M. Tanno, S. Sueyoshi, S. Kamiya, Chem. Pharm. Bull. 1982, 30, 3125 ±
3132.
[
[
[
150 a) Although no irreversible reaction occurs between [Co(tpp)0 and
TEMPO, it should be noted that a reversible interaction apparently
[340 R. A. W. Johnstone, M. L. P. G. Nunes, M. M. Pereira, A. M. d×A.
Rocha Gonsalves, A. Serra, Heterocycles 1996, 43, 1423 ± 1437.
[350 a) A. Shirazi, H. M. Goff, Inorg. Chem. 1982, 21, 342ꢀ ± 3425; b) A. D.
Adler, F. R. Longo, V. Varadi, Inorg. Synth. 1976, 16, 213 ± 22ꢀ.
[360 A. R. Katritzky, K. S. Laurenzo, J. Org. Chem. 1988, 53, 3978 ± 3982.
[370 S. Chandrasekhar, M. Venkat Reddy, L. Chandraiah, Synth. Commun.
1999, 29, 3981± 3987.
[380 T. Suwa, E. Sugiyama, I. Shibata, A. Baba, Synthesis 2000, 789 ± 8ꢀꢀ.
[390 R. Grigg, T. R. B. Mitchell, N. Tongpenyai, Synthesis 1981, 442 ± 444.
[4ꢀ0 K. Kanie, K. Mizuno, M. Kuroboshi, T. Hiyama, Bull. Chem. Soc. Jpn.
1998, 71, 1973 ± 1991.
1
occurs in solution, since the H NMR signals of the porphyrin are
immediately broadened by the addition of TEMPO to the solution.
Delocalization of the unpaired spin density over the azide moiety is in
line with what known for the complexes derived by interaction of
cobalt(ii) complexes with dioxygen;[15b0 b) For a review see: R. S.
Drago, B. B. Corden, Acc. Chem. Res. 1980, 13, 353 ± 36ꢀ.
[
[
160 C. Hansch, A. Leo, R. W. Taft, Chem. Rev. 1991, 91, 165 ± 195.
‡
[180
170 Worse correlations were observed with the s or with the TE (see
also later) parameters.
[
180 Y.-D. Wu, C.-L. Wong, K. W. K. Chan, G.-Z. Ji, X.-K. Jiang, J. Org.
Chem. 1996, 61, 746 ± 75ꢀ.
[410 B. Carboni, M. Vaultier, T. Courgeon, R. Carri e¡ , Bull. Soc. Chim. Fr.
1989, 844-849.
[
[
190 X.-K. Jiang, Acc. Chem. Res. 1997, 30, 283 ± 289 and references therein.
2ꢀ0 Leading references: a) J. F. Hartwig, Chem. Rev. 1998, 98, 2ꢀ46 ± 2ꢀ67;
b) B. H. Yang, S. L. Buchwald, J. Organomet. Chem. 1999, 576, 125 ±
[420 M. Tanaka, T. Kobayashi, Synthesis 1985, 967 ± 969.
[430 R. M. Lopez, G. C. Fu, Tetrahedron 1997, 53, 16349 ± 16354.
[440 L. Forlani, M. Sintoni, P. E. Todesco, Gazz. Chim. Ital. 1986, 116, 229 ±
232.
1
46; c) A. Klapars, J. C. Antilla, X. Huang, S. L. Buchwald, J. Am.
Chem. Soc. 2001, 123, 7727 ± 7729; d) S. Hayat, A. Rahman, M. I.
Choudhary, K. M. Khan, W. Shumann, E. Bayer, Tetrahedron 2001, 57,
[450 S. Bennour, J. Toullec, Can. J. Chem. 1999, 77, 76ꢀ ± 773.
[460 J. R. Hwu, W. N. Tseng, H. V. Patel, F. F. Wong, D.-N. Horng, B. R.
Liaw, L. C. Lin, J. Org. Chem. 1999, 64, 2211 ± 2218.
9
951± 9957; e) M. Beller, C. Breindl, T. H. Riermeier, M. Eichberger,
H. Trautwein, Angew. Chem. 1998, 110, 3571± 3573; Angew. Chem.
Int. Ed. 1998, 37, 3389 ± 3391.
[470 A. Echevarria, J. Miller, M. G. Nascimento, Magn. Reson. Chem. 1985,
23, 8ꢀ9 ± 813.
258
¹ 2ꢀꢀ3 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
ꢀ947-6539/ꢀ3/ꢀ9ꢀ1-ꢀ258 $ 2ꢀ.ꢀꢀ+.5ꢀ/ꢀ
Chem. Eur. J. 2003, 9, No. 1

Amination of Benzylic CÀH Bonds
249±259
[
[
480 J. Dale, L. Vikersveen, Acta Chem. Scand. Ser. B 1988, 42, 354 ± 361.
490 Y.-Z. Li, J. P. Kirby, M. W. George, M. Polyakoff, G. B. Schuster, J.
Am. Chem. Soc. 1988, 110, 8ꢀ92 ± 8ꢀ98.
5ꢀ0 X.-Y. Wang, Y.-L. Wang, Z.-Y. Zhang, C.-L. Wang, J.-P. Li, L. Shi, Z.-F.
Duan, Synth. Commun. 1999, 29, 481± 485.
[560 S. Chang, H. J. Koh, B.-S. Lee, I. Lee, J. Org. Chem. 1995, 60, 776ꢀ ±
7768.
[570 H. K. Hall, Jr., A. B. Padias, P. A. Williams, J.-M. Gosau, H. W.
Boone, D.-K. Park, Macromolecules 1995, 28, 1 ±8.
[580 E. Alonso, D. J. Ramon, M. Yus, Tetrahedron 1997, 53, 14355 ± 14368.
[590 F. E. Baz, M. Riv e¡ re-Baudet, M. Ahra, J. Organomet. Chem. 1997, 548,
123 ± 129.
[
[
[
510 R. H. Lowack, R. Weiss, J. Am. Chem. Soc. 1990, 112, 333 ± 338.
520 H. J. Koh, H. W. Lee, I. Lee, J. Chem. Soc. Perkin Trans. 2, 1994, 125 ±
1
29.
530 a) Y. S. Andreichikov, Y. V. Ionov, Zh. Org. Khim. 1982, 18, 243ꢀ ±
435; Russ. J. Org. Chem. 1982, 18, 2154 ± 2158 (English translation);
b) B. I. Buzykin, G. D. Lezhnina, Y. P. Kitaev, Zh. Org. Khim. 1975, 11,
48 ± 853; Russ. J. Org. Chem. 1975, 11, 837 ± 841(English translation).
[6ꢀ0 J. Zemlicka, J. P. Horwitz, J. Org. Chem. 1971, 36, 28ꢀ9 ± 2812.
[610 J. S. Sandhu, D. Mago, Tetrahedron Lett. 1975, 16, 1ꢀ91 ± 1ꢀ92.
[620 A. Bolognese, M. V. Diurno, O. Mazzoni, F. Giordano, Tetrahedron
1991, 47, 7417 ± 7428.
[630 I. I. Juchnovski, V. Ognyanova, Spectrosc. Lett. 1993, 26, 447 ± 452.
[640 E. Liepins, V. A. Pestunovich, A. V. Eremeev, D. A. Tikhomirov, N. P.
Gaidarova, Khim. Geterotsikl. Soedin. 1977, 9ꢀ6 ± 9ꢀ9; Chem. Abstr.
1977, 87, 183795b.
[
2
8
[
540 a) J. W. Bae, S. H. Lee, Y. J. Cho, C. H. Yoon, J. Chem. Soc. Perkin
Trans. 1 2000, 145 ± 146; b) A. Pelter, R. M. Rosser, S. Mills, J. Chem.
Soc. Perkin Trans. 1 1984, 717 ± 72ꢀ.
[
550 a) J. P. Wolfe, H. Tomori, J. P. Sadighi, J. Yin, S. L. Buchwald, J. Org.
Chem. 2000, 65, 1158 ± 1174; b) G. Reddelien, Chem. Ber. 1914, 47,
[650 a) O. H. Wheeler, P. H. Gore, J. Org. Chem. 1961, 26, 3298 ± 33ꢀ1; b) J.
Bernstein, I. Izak, J. Cryst. Mol. Struct. 1975, 5, 257 ± 266.
1355 ± 1364; c) W. A. Henderson, A. Zweig, Tetrahedron 1971, 27,
53ꢀ7 ± 5313; d) H. Abu-Eittah, Bull. Chem. Soc. Jpn. 1974, 47, 2576 ±
2579.
Received: June 25, 2ꢀꢀ2 [F42ꢀ60
Chem. Eur. J. 2003, 9, No. 1¹ 2ꢀꢀ3 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
ꢀ947-6539/ꢀ3/ -ꢀ ꢀ9 2ꢀ 51 9 $ 2ꢀ.ꢀꢀ+.5ꢀ/ꢀ
259