[a]
À
Table 1. Optimisation of the intramolecular C H amination of 2b.
[Ru(TPP)CO]
T
t[b]
Yield [%][c]
[%]
[8C]
[h]
1b
3b
4b
5b
1
2
3
5
none
2
0.4
1
0.4
1
none
180
180
180
120
120
120
120
13
2
3
10
1.5
7
7
92
8
18
10
24
4
–
23
27
10
14
46
–
–
54
41
40
46
32
–
Scheme 2. Suggested route for the formation of phenanthridine.
12
31
12
11
2
6[d]
7[d]
8[d]
sumption of the azide due to the presence of air in the reac-
tion mixture. Leaving the mixture exposed to air at 1208C
in the presence of the catalyst for another two hours con-
cludes the oxidation process. To support the role of path B
in the formation of 5b, we repeated the reaction but left the
reaction mixture under a nitrogen atmosphere for 5 h after
the complete conversion of the azide starting material. Com-
plete conversion of 4b into 5b was not observed due to the
lack of air.
1
1.5
97
[a] General procedure: 2b (70.0 mg, 3.13ꢃ10À4 mol) was dissolved in 1,2-
dichlorobenzene (20 mL) under dinitrogen. [b] Time required to reach
complete conversion of 2b. [c] Determined by 1H NMR spectroscopy
with 2,4-dinitrotoluene as the internal standard. [d] The reaction mixture
was irradiated with a halogen lamp (400 W).
well as the catalyst loading. The positive effect of a halogen
lamp assumes practical importance in terms of low cost and
availability.
Considering the product distributions reported in Table 1,
entry 7, the O2 concentration (ꢀ1.3 mm) involved in path B
can either originate from impurities in the nitrogen gas em-
ployed or be introduced during the preparation of the
sample for NMR spectroscopy. To reduce the oxygen con-
tamination, the annulation reactions of 2b were carried out
under argon and at a higher biaryl azide concentration
(7.61ꢃ10À2 m instead of the 1.56ꢃ10À2 m used in the reac-
Data collected to date indicate that 1 mol% catalyst at
1208C, coupled with a halogen lamp, are the optimal condi-
tions to perform intramolecular amination reactions. In fact,
the use of a lower catalyst loading (Table 1, entry 6) was as-
sociated with the formation of a significant amount of carba-
zole by the competitive uncatalyzed reaction. As shown in
Table 1, dihydrophenanthridine (4b) and phenanthridine
(5b) are obtained in variable amounts depending on the ex-
perimental conditions and for every set of reaction condi-
tions the formation of the 2-amine biaryl compound (1b)
was observed.
1
tions reported in Table 1, entry 7). The H NMR spectrum,
run under argon after the complete consumption of 2b, re-
vealed the presence of 1b (16%), 3b (4%), 4b (60%) and
5b (17%). The employment of more strict experimental
conditions allowed the formation of a 5b/1b ratio of almost
1:1 and increased the yield of 4b due to nearly complete
suppression of path B.
On the other hand, to favour the oxidation of 4b into 5b,
we performed the intramolecular amination of 2b by using
the experimental conditions described for Table 1, entry 7
but the reaction was exposed to air from the beginning. The
1H NMR spectrum of the crude mixture registered after the
complete consumption of the azide showed the presence of
5b (17%), 4b (66%), 1b (10%) and 3b (traces). Only if the
reaction mixture was left to stir at 1208C for a further 2 h
was 4b completely transformed into 5b, indicating that the
Our previous study on the intermolecular benzylic amina-
tion of hydrocarbons by aryl azides[32] indicated that the first
À
step of the C H amination was the formation of a benzylic
amine that yields the corresponding imine through a reaction
with another azide molecule. The stoichiometric by-products
of this second step are the anilines. If we envisage a similar
mechanism for the intramolecular amination, we should
obtain equal amounts of phenanthridine (5b) and 2-amine
biaryl (1b), and considering complete azide conversion, the
remainder of the material must form dihydrophenathridine
(4b). On the other hand, the formation of a 1b/5b ratio
smaller than one indicated the presence of a competitive ox-
idation mechanism (see Table 1, entry 2). Taking into ac-
count that the oxidation of 4b to 5b could be due to
À
oxidation is a slower process than the C H amination. The
decrease in the yield of 5b (the yield decreased from 30 to
17% by running the reaction in air rather than under a N2
atmosphere) could be due to a different reaction rate of
path A with respect to path B (Scheme 2). Moreover,
path A may even be inhibited by dioxygen and therefore
the formation of 5b should be mainly due to path B.[32]
The role of ruthenium in the conversion of 4b into 5b
was confirmed by stirring a ruthenium-free mixture of the
oxygen[26] and promoted by [Ru
ACTHNUTRGNE(UNG TPP)CO], a catalytic mix-
ture containing both heterocyclic compounds was left to stir
at 1208C and directly exposed to air for 2 h by simply re-
moving the flask stopper. Complete conversion of 4b into
5b was observed. This experimental result indicates that the
phenanthridine formation is probably due to the two mecha-
nisms indicated in Scheme 2, paths A and B. We suggest
that, with the dehydrogenation of 4b to form 5b being a fa-
vourable process, it can commence before complete con-
1
two compounds in air at 1208C for 7 h. The H NMR spec-
tra registered at the beginning and end of the reaction
showed an unvaried product distribution, indicating that, for
the oxidation process of 4b to 5b to occur, a catalytic
&
2
&
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 0000, 00, 0 – 0
ÝÝ
These are not the final page numbers!