Angewandte
Chemie
Table 1: Amination of cyclohexanol with ammonia in different solvents.[a]
catalysts for the amination of alcohols.[17] Therefore, we tested
this catalyst precursor with 23 different ligands[18] in the
benchmark reaction. At 1308C the majority of catalytic
systems decomposed and gave unsatisfying results (< 5%
Entry Solvent
mNH [g] Conv.[b] [%] Yprim.[b] [%] Ysec.[b] [%]
3
1
2
3
4
5
heptane
toluene
THF
0.2
0.2
0.2
0.2
68
72
62
62
66
50
79
80
78
79
80
93
95
36
36
30
49
46
35
44
48
52
55
66
65
87
26
20
24
11
13
7
25
23
18
14
4
diglyme
t-amyl alcohol 0.2
t-amyl alcohol 0.2
t-amyl alcohol 0.2
t-amyl alcohol 0.2
t-amyl alcohol 0.3
t-amyl alcohol 0.6
t-amyl alcohol 1.0
6[c]
7[d]
8[e]
9[e]
10[e]
11[e]
12[d,e] t-amyl alcohol 0.3
13[d,e] t-amyl alcohol 0.6
20
5
[a] Reaction conditions: 1 mmol cyclohexanol, 0.2 g (6 bar) ammonia at
RT, 0.02 mmol [Ru3(CO)12], 0.06 mmol CataCXiumPCy, 1408C, 20 h.
[b] Conversion and yield (based on cyclohexanol) were determined by GC
analysis with hexadecane as an internal standard. [c] Addition of 20 mL
water. [d] Molecular sieves were suspended above the reaction mixture.
[e] 1508C.
Indeed, the conversion increased to 79%; however, a
significant amount of dicyclohexylamine was still obtained.
The conversion was also increased at higher temperature
(1508C), but again the amount of dicyclohexylamine
increased as well (Table 1, entries 5 and 8). To reduce the
formation of this secondary amine, we increased the amount
of ammonia. Without molecular sieves present slightly higher
yields of the primary amine were observed (Table 1,
entries 8–11). However, in the presence of molecular sieves
a conversion of 95% was achieved and the yield of cyclo-
hexylamine increased to 87% (Table 1, entry 13)! Interest-
ingly, the amount of ammonia does not influence the
conversion of the alcohol.
yield of cyclohexylamine). Standard mono- and bidentate
arylphosphines such as triphenylphosphine (1), Xantphos[19]
(4,5-bis(diphenylphosphino)-9,9-dimethylxanthene (2)), 1,3-
bis(diphenylphosphino)propane (3), and tris(4-methoxyphe-
nyl)phosphine (4) showed no activity (< 5% yield of cyclo-
hexylamine), while more electron-rich phosphines like tricy-
clohexylphosphine (5), benzyldi-1-adamantylphosphine (6),
and n-butyldi-1-adamantyl-phosphine (7) were slightly active
in this reaction (10–20% yield of cyclohexylamine). Of all the
ligands tested, CataCXiumPCy (2-(dicyclohexylphosphino)-
1-phenyl-1H-pyrrole (8)) showed the highest reactivity and
gave cyclohexylamine in 30% yield.
Applying [Ru3(CO)12]/CataCXiumPCy as the most prom-
ising catalyst system, we investigated the influence of the
critical reaction parameters in more detail. The concentration
of ammonia should have a significant influence on the
reactivity and chemoselectivity of the reaction.
To demonstrate the general applicability of the
[Ru3(CO)12]/CataCXiumPCy system for this reaction and
the scope of the process, we tested various secondary alcohols.
In general, catalytic experiments were conducted in the
presence of 2 mol% [Ru3(CO)12] and 6 mol% Cata-
CXiumPCy with 1 mmol alcohol in 1 mL t-amyl alcohol
along with molecular sieves at 1508C. As shown in Table 2
various secondary alcohols reacted with ammonia to give the
desired products in good to excellent yields.
In most cases the use of 0.6 g ammonia resulted in the
formation of significant amounts of the ketone. Hence, the
amount of ammonia was increased to 1 g (Table 2, entries 3
and 6–11) to enhance the nucleophilic attack of ammonia to
the ketone (Scheme 2). We assume that in those cases this is
the bottleneck of the reaction. Excellent yields > 90% were
observed with 1-dimethylamino-2-propanol (Table 2,
entry 11), 2-adamantanol, and 1,4-dioxaspiro[4.5]decan-8-ol
(Table 2, entries 5 and 13). In the case of sterically hindered 2-
adamantanol the reaction had to be conducted at higher
temperature (Table 2, entries 4 and 5). It should be noted that
even at high temperature the selectivity towards the forma-
tion of the primary amine was very high and the catalytic
Of the different solvents tested, diglyme and tert-amyl
alcohol gave the best results (Table 1, entries 4 and 5). The
differences between the yields and the conversions are caused
by the formation of the ketone along with the primary and
secondary imines. When the the reaction was run in heptane,
toluene, and tetrahydrofuran (THF), the selectivity for the
formation of the primary amine dropped (Table 1, entries 1–
3). For further experiments tert-amyl alcohol was chosen
because it can be removed easily from the products. Notably,
water, which is formed during the reaction, has a negative
influence on the conversion (Table 1, entries 5 and 6), which is
in agreement with the proposed mechanism. Apparently, a
higher concentration of water leads to increased hydrolysis of
the imine to yield the ketone, which can also be hydrogenated
by the catalyst. Thus, we reduced the amount of water in the
reaction solution by suspending molecular sieves above the
reaction mixture in a Teflon basket (Table 1, entries 5 and 7).
Angew. Chem. Int. Ed. 2010, 49, 8126 –8129
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
8127