A general study of aryloxo and alkoxo ligands in the titanium-catalyzed intermolecular hydroamination of terminal alkynes

Article abstract of DOI:10.1002/ejoc.200500423

A general study of the regioselective hydroamination of terminal alkynes in the presence of Ti(NEt₂)₄ and different aryloxo and alkoxo ligands is presented. Depending on the ligand the regioselectivity towards the Markovnikov and the anti-Markovnikov addition product can be controlled. The experimentally observed isomer distribution is explained perfectly by detailed theoretical investigations which demonstrate that the regioselectivity is determined by the relative stability of the corresponding alkynetitanium π complexes. Wiley-VCH Verlag GmbH & Co. KGaA, 2005.

Full text of DOI:10.1002/ejoc.200500423

FULL PAPER
DOI: 10.1002/ejoc.200500423
A General Study of Aryloxo and Alkoxo Ligands in the Titanium-Catalyzed
Intermolecular Hydroamination of Terminal Alkynes
Annegret Tillack,[a] Vivek Khedkar,[a] Haijun Jiao,[a] and Matthias Beller*[a]
Keywords: Hydroamination / Markovnikov / anti-Markovnikov / Titanium complexes / DFT / Alkynes
A general study of the regioselective hydroamination of ter-
perfectly by detailed theoretical investigations which demon-
minal alkynes in the presence of Ti(NEt2)4 and different aryl- strate that the regioselectivity is determined by the relative
oxo and alkoxo ligands is presented. Depending on the li-
gand the regioselectivity towards the Markovnikov and the
anti-Markovnikov addition product can be controlled. The
experimentally observed isomer distribution is explained
stability of the corresponding alkynetitanium π complexes.
(© Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim,
Germany, 2005)
of their reactivity and availability titanium complexes have
probably attracted the most interest. Important progress in
the intermolecular hydroamination of alkynes with titanium
complexes has been reported by Bergman,[5] Doye,[6]
Odom,[7] us,[8] Schafer,[9] and others.[10] It is worth noting
that Ackermann et al. recently demonstrated that norbor-
nene reacts with various anilines in the presence of
TiCl4.[10d]
Among the various titanium catalysts that are known for
hydroamination, we introduced bis(aryloxo)-bis(dialkyl-
amido)titanium complexes, which are comparably air-stable
and easy to handle. These complexes catalyze the selective
Markovnikov hydroamination of terminal alkynes.[11] Re-
cently, we also reported that by addition of different phenol
ligands to Ti(NEt2)4 the regio- and chemoselectivity of hy-
droaminations of terminal alkynes can be controlled.[12]
Here, we present a full account of our selectivity studies.
Furthermore, we present theoretical calculations, which ex-
plain the observed selectivity.
Introduction
The addition of nitrogen and oxygen compounds across
carbon–carbon multiple bonds is an important subject in
organic synthesis.[1] In general, these processes are perfectly
suited to fulfil today’s requirements for sustainable pro-
cesses because of the availabilty of substrates, and 100%
atom efficiency. By using unsymmetrical olefins or alkynes
the addition of H–Nu (Nu = OR, NR2, etc.) can lead to
two isomeric products. Typically, most of the electrophilic
addition reactions follow the Markovnikov rule, in which
the branched compound is mainly produced. However,
often the linear isomer is the more desired product for in-
dustrial bulk applications. Hence, anti-Markovnikov func-
tionalizations, especially of aliphatic olefins, continue to be
a challenging goal for catalysis.
It is obvious that the introduction of tailor-made transi-
tion metal complexes led to significant advances over the
last decades in a number of functionalization reactions, e.g.
hydroformylation, hydrocarboxylation, hydrocyanation.
Nevertheless, control of chemo-, regio- and enantio-
selectivity for addition processes to unsaturated substrates
needs to be further improved, because this is the basis for
new applications and innovation in organic synthesis.
We have been involved in the development and explora-
tion of new methods for selective amination of olefins and
alkynes for some time now. Apart from carbonylative amin-
ation of olefins (so-called hydroaminomethylations),[2] the
direct hydroamination of olefins[3] and alkynes has attracted
our interest. Over the last decade a multitude of catalysts
have been explored for alkyne hydroaminations.[4] Because
Results and Discussion
Monodentate Alkoxo and Aryloxo Ligands in
Hydroamination of Alkynes
Recently, we investigated the control of regioselectivity
(Markovnikov or anti-Markovnikov) by using titanocene
catalysts with different ligands in the hydroamination of un-
symmetrical alkynes (Scheme 1).[8a]
Based on this work we were interested in investigating,
in more detail, the influence of monodentate and bidentate
alkoxo and aryloxo ligands on the regioselectivity of the
intermolecular hydroamination of terminal alkynes. Ini-
tially, the addition of sec-butylamine to 1-octyne was
studied as a model reaction in the presence of 11 mono-
dentate alkoxo and aryloxo ligands (Table 1). With regards
[a] Leibniz-Institut für Organische Katalyse an der Universität Ro-
stock e.V.,
Albert-Einstein-Str. 29a, 18059 Rostock, Germany
Fax: +49-381-1281-5000
E-mail: matthias.beller@ifok-rostock.de
Eur. J. Org. Chem. 2005, 5001–5012
© 2005 Wiley-VCH Verlag GmbH & Co. KgaA, Weinheim
5001

A. Tillack, V. Khedkar, H. Jiao, M. Beller
FULL PAPER
(NEt2)4 and alcohols or phenols 1–11 (Scheme 2). In gene-
ral, the hydroamination reaction was performed in toluene
at 100 °C for 24 h in the presence of 10 mol-% Ti(NEt2)4
and 20 mol-% of the corresponding ligand. Typically a
slight excess of amine (1:1.2 equiv.) was employed in order
to suppress oligomerization and polymerization of 1-oc-
tyne. Nevertheless, small amounts of dimers, oligomers and
polymers of the alkyne and also amine-alkyne dimers were
observed in the case of the less selective catalysts.
In a preliminary study[12] we have shown that the use of
sterically hindered bulky phenol 1 in the hydroamination of
Scheme 1. Hydroamination of terminal alkynes.
to practicalities, instead of using synthesized alkoxo- or 1-octyne with sec-butylamine gave an excellent yield (98%)
aryloxotitanium complexes, we formed the corresponding and a high Markovnikov selectivity (anti-M:M = 10:90).
titanium catalysts in situ from commercially available Ti- However, by employing sterically hindered alkoxo ligands 2
Table 1. Hydroamination of 1-octyne with sec-butylamine in the presence of Ti(NEt2)4/ligand.[a]
Entry
Ligand
Conversion [%]
Yield[b] [%]
anti-M:M ratio[c]
1
2
3
4
5
6
7
8
1
2
100
66
14
98
64
13
31
97
88
94
97
98
10:90
74:26
76:24
75:25
49:51
72:28
94:6
94:6
94:6
96:4
3
4
32
5
100
100
100
100
100
83
100
100
7
6
7
8
9[d]
10[e]
11
12
13
8
8
82
98 (70)
90
9
94:6 (96:4)
95:5
84:16
10
11
4
[a] Reaction conditions: 1.5 mmol 1-octyne, 1.8 mmol amine, 10 mol-% catalyst, Ti/L = 1:2, 2 mL toluene, 100 °C, 24 h. [b] Yield is
determined by GC analysis with dodecane or hexadecane as internal standard, isolated yield is given in parenthesis. [c] GC analysis used
to determine the regioisomers, the ratio of regiomers of the isolated product after distillation is given in parenthesis. [d] 85 °C. [e] 5 mol-
% catalyst.
Scheme 2. Alkoxo and aryloxo monodentate ligands used in our study.
5002
www.eurjoc.org
© 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2005, 5001–5012

Titanium-Catalyzed Hydroamination of Terminal Alkynes
FULL PAPER
and 3 a lower yield of the corresponding imines with poor ligand from 1 to 8, 9 or 10. Here, especially the behavior of
regioselectivity was obtained (Table 1, Entries 2–3). Also in the 2-morpholinophenol (10) is surprising. To the best of
the presence of ferrocenyl ligand 4 (benzyloxo type) hydro- our knowledge this is the first example where aminophenol
amination proceeded very slowly and gave only a 31% yield is used as a ligand in titanium-catalyzed aminations.
of the corresponding imines (Table 1, Entry 4). Compared
In general, only sterically hindered phenols showed sig-
to alkoxo and benzyloxo ligands, monodentate phenols gave nificant catalytic activity. By employing simple phenol or
good to excellent conversions and yields of the correspond- pentafluorophenol 11 as the ligand no activity was observed
ing imine. Surprisingly, when using 2,6-diphenlyphenol (5) due to the formation of stable tris- and tetrakisphenoxo ti-
the selectivity dropped significantly and a 1:1-mixture of tanium complexes (Table 1, Entry 13).
regioisomers was obtained. In the presence of 2-tert-butyl-
Next, we were interested in the generality of the observed
4,6-dimethylphenol (6) an 88% yield of imines was obtained regioselectivity effect (Table 2). Hence, the behavior of li-
with major anti-Markovnikov selectivity (anti-M:M = gands 1–2 and 5–10 was tested in the hydroamination reac-
72:28).
tion of 1-octyne with benzylamine, n-butylamine, tert-bu-
Excellent anti-Markovnikov selectivity (94:6 to 95:5) is tylamine, cyclooctylamine and aniline. In the presence of
observed in the presence of ligands 7, 8, 9 and 10. Thus, a ligand 1 excellent yields of the imines and a high Markovni-
switch from the Markovnikov product to the anti-Markov- kov selectivity was obtained with nonhindered amines such
nikov product is observed simply by changing the phenol as benzylamine, n-butylamine, cyclooctylamine and aniline
Table 2. Hydroamination of 1-octyne with different amines using Ti(NEt2)4 and monodentate ligands.[a]
Entry
Ligand
Amine
Conversion [%]
Yield[b] [%]
anti-M:M ratio[c]
1
2
3
1
1
1
1
1
1
1
2
2
5
5
5
5
5
6
6
7
7
7
8
8
8
8
8
9
9
9
10
10
benzylamine
n-butylamine
cyclooctylamine
tert-butylamine
tert-butylamine
aniline
100
100
100
50
99
99
99
50
58
99
96
9
22
86
96
18
98
40
44
20
52
92
66
72 (41)
82
94 (65)
44
86 (37)
70
88 (50)
30
40
49
20:80
25:75
14:86
74:26
88:12
22:78
20:80
44:56
92:8
43:57
48:52
74:26
16:84
30:70
70:30
92:8
84:16
4
5[d]
6
75
100
100
12
26
92
100
23
100
100
97
7[d]
8
aniline
benzylamine
tert-butylamine
benzylamine
n-butylamine
tert-butylamine
aniline
9
10
11
12
13
14[e]
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
aniline
benzylamine
tert-butylamine
benzylamine
cyclooctylamine
tert-butylamine
benzylamine
n-butylamine
cyclooctylamine
tert-butylamine
aniline
benzylamine
n-butylamine
tert-butylamine
benzylamine
tert-butylamine
22
100
100
98
100
100
100
54
100
100
100
33
94:6
99:1
86:14 (87:13)
92:8
94:6 (96:4)
99:1
34:66 (30:70)
88:12
91:9 (91:9)
99:1
98
89
93:7
98:2
[a] Reaction conditions: 1.5 mmol 1-octyne, 1.8 mmol amine, 10 mol-% catalyst, Ti/L = 1:2, 2 mL toluene, 100 °C, 24 h. [b] Yield is
determined by GC analysis with dodecane or hexadecane as internal standard, isolated yield is given in parenthesis. [c] GC analysis was
used to determine the regioisomers, the ratio of regiomers of the isolated product after distillation is given in parenthesis. [d] Ti:L = 1:1.
[e] (2,6-C6H5-C6H3O)2TiCl2 as catalyst; reaction conditions: 1.5 mmol 1-octyne, 1.5 mmol aniline, 0.9 mmol tBuNH2 as additive, 10 mol-
% catalyst, 2 mL toluene, 100 °C, 24 h.
Eur. J. Org. Chem. 2005, 5001–5012
© 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
5003

A. Tillack, V. Khedkar, H. Jiao, M. Beller
FULL PAPER
(Table 2, Entries 1–3 and 6–7). However, sterically hindered gands proceeds very slowly with a high anti-Markovnikov
tert-butylamine gave only a low yield with a high anti-Mar- selectivity. On the other hand when applying aniline, the
kovnikov selectivity (Table 2, Entries 4–5).
Markovnikov isomer is always obtained preferentially. A
When using ligand 2 the reaction with benzylamine gave more detailed discussion of this selectivity effect is pre-
a very low yield (9%) and a 1:1-mixture of regioisomers. sented in the area of theoretical investigations (see below).
Also tert-butylamine led to a low yield (22%), but with ex-
cellent anti-Markovnikov selectivity (anti-M:M = 98:2).
Bidentate Alkoxo and Aryloxo Ligands in Hydroamination
When employing ligand 5 with different aliphatic amines
of Alkynes
the observed regioselectivity is bad (Table 2, Entries 10–12).
The reaction of aniline, using the in situ formed aryloxo-
Next, we investigated the effect of bidentate alkoxo and
titanium complex from commercially available Ti(NEt2)4, aryloxo ligands (Scheme 3) in different titanium-catalyzed
and 5 gave an excellent yield (98%) with predominant Mar- hydroamination reactions. Here, we performed the hydro-
kovnikov selectivity (Table 2, Entry 13). At this point it is amination of 1-octyne with six different alkoxo and aryloxo
interesting to note that previous studies showed that the ligands (18–23) with simple nonhindered amines, e.g. ben-
reaction of anilines with internal alkynes using simple TiCl4 zylamine, sec-butylamine and sterically hindered tert-
as the catalyst gave less of the hydroaminated product butylamine (Table 3).
(Ͻ2%). However, Ackermann demonstrated that the cata-
The reaction of benzylamine in the presence of bidentate
lytic activity for such a reaction can be increased by ad- ligands 18–23 led preferentially to the Markovnikov imines,
dition of tert-BuNH2 (up to 96%).[10e] Thus, we also tested albeit with somewhat lower selectivity (anti-M:M = 40:60
the defined complex (2,6-C6H5-C6H3O)2TiCl2 in the pres- to 28:72) and very low yields (10–52%). Surprisingly, sec-
ence of tert-butylamine as the additive, as well as without butylamine led preferentially to the anti-Markovnikov iso-
this additive. Here, a 40% yield of imine is obtained with mer. Applying this amine compared to benzylamine the
the additive (Table 2, Entry 14) (without additive Ͻ1%), yields of the corresponding imines were good (Table 3, En-
which is less efficient compared to the halide-free aryloxoti- tries 2, 5, 8, 11 and 15), except for ligand 22 because of the
tanium complex. The hydroamination reaction using dif- precipitation of the Ti complex (Table 3, Entry 13). This
ferent aliphatic amines in the presence of ligands 6–10 gave result is in contrast with those from experiments in the pres-
preferentially the anti-Markovnikov product with good to ence of monodentate ligands for nonhindered amines.
excellent selectivity (anti-M:M = 70:30 to 99:1). However,
in the case of aniline a high Markovnikov selectivity was ence of ligands 18, 19 and 23 with low yields and high anti-
obtained by employing 8 (Table 2, Entry 24). Markovnikov selectivity. However, a bulky group on the or-
As expected tert-butylamine reacted slowly in the pres-
In general, the ligand has a significant impact on the tho-position of bidentate phenols 20 and 21 led to good
observed selectivity. However, the nature of the amine also yields of imines with excellent anti-Markovnikov selectivity
influences the regioselectivity and yields of the correspond- (74–88%, anti-M:M = 97:3 to 99:1). In general, when apply-
ing imines. For example the reaction with sterically hin- ing bidentate phenol ligands, oligomers and polymers of 1-
dered tert-butylamine in the presence of all the tested li- octyne were observed as side products.
Scheme 3. Bidentate alkoxo and aryloxo ligands used in catalytic hydroamination reactions.
5004
www.eurjoc.org
© 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2005, 5001–5012

Titanium-Catalyzed Hydroamination of Terminal Alkynes
FULL PAPER
Table 3. Hydroamination of 1-octyne with different amines using Ti(NEt2)4 and bidentate ligands.[a]
Entry
Ligand
Amine
Conversion [%]
Yield[b] [%]
anti-M:M ratio[c]
1
2
3
4
5
6
7
8
18
18
18
19
19
19
20
20
20
21
21
21
22
23
23
23
benzylamine
14
72
46
10
63
39
52
88
36
30
88
86
48
91
74
5
33:67
68:32
98:2
33:67
62:38
86:14
30:70
56:44
97:3
28:72
76:24
99:1
76:24
40:60
76:24
94:6
sec-butylamine
tert-butylamine
benzylamine
sec-butylamine
tert-butylamine
benzylamine
sec-butylamine
tert-butylamine
benzylamine
sec-butylamine
tert-butylamine
sec-butylamine
benzylamine
81
100
100
52
90
97
9
10
11
12
13[d]
14
15
16
91
100
100
5
70
99
22
70
36
sec-butylamine
tert-butylamine
58
[a] Reaction conditions: 1.5 mmol 1-octyne, 1.8 mmol amine, 10 mol-% catalyst, Ti/L = 1:1, 2 mL toluene, 100 °C, 24 h. [b] Yield is
determined by GC analysis with dodecane or hexadecane as internal standard. [c] GC analysis was used to determine the regioisomers.
[d] Ti complex precipitates.
Comparison of Catalytic Systems
been proven to give the best regioselectivity with this amine
in the hydroamination reaction. Using the in situ catalyst
system Ti(NEt2)4/1 a significant change in regioselectivity
was observed by varying the Ti/L ratio from 1:1 to 1:2
(Table 5, Entries 1 and 2). Further increasing the ratio up
to 1:4 (Ti/1) did not influence the regioselectivity (Table 5,
Entry 3). The sterically less hindered ligands 5, 7 and 8,
when compared to 1, did not work at all when higher con-
centration of ligands were used (Table 5, Entries 6, 7, 10,
and 14). All the ligands that were tested showed lower re-
gioselectivity for the 1:1 ratio (Ti/L), and the best results
were obtained for a 1:2 Ti/L ratio.
The influence of the ligand concentration is explained by
the fact that the sterically hindered ligand 1 does not form
tris- and tetrakis(aryloxo)titanium complexes at higher li-
gand ratios. However, tris- and tetrakis(aryloxo)titanium
complexes were formed quickly by employing sterically less
hindered ligands 5, 7 and 8.
Next, we were interested in comparing the in situ gener-
ated catalyst from commercially available Ti(NEt2)4 and
aryloxo ligand 1 with the previously described bis(2,6-di-
tert-butyl-4-methylphenoxo)-bis(dimethylamido)titanium
complex 24 in the hydroamination of 1-octyne with dif-
ferent amines. Catalytic experiments were performed with
both systems at 100 °C for 24 h (Table 4).
By employing n-butylamine, benzylamine, sec-bu-
tylamine and cyclooctylamine, similar yields of the corre-
sponding imine and almost similar regioselectiviy are ob-
served with both catalytic systems (Table 4, Entries 1–8).
Interestingly, the reaction of tert-butylamine with 1-octyne
in the presence of the defined complex 24 did not work
(Table 4, Entry 9). However, by using the in situ system [Ti-
(NEt2)4/1] under the same reaction conditions the corre-
sponding imine is obtained in 50% yield with high anti-
Markovnikov selectivity (anti-M:M = 74:26) (Table 4, Entry
10). This effect might be explained by the formation of dif-
ferent Ti(OAr)n(NHtBu)4–n species under the in situ condi-
tions. In agreement with this the reaction of tert-butylamine
in the presence of a 1:1-catalyst-to-ligand-ratio gave a better
Hydroamination of Internal Alkynes
We also studied the feasibility of aryloxo ligands in the
regioselectivity compared to a 1:2 ratio (Table 4, Entries 10 hydroamination of an unsymmetrical internal alkyne. Here,
and 11). Also, by applying aniline the concentration of the the hydroamination reaction of methylphenylacetylene with
ligand influenced the regioselectivity (Table 4, Entries 12 benzylamine was performed using Ti(NEt2)4 and four dif-
and 13). Thus, we were interested in studying the effect of ferent monodentate aryloxo ligands at 120 °C for 24 h
the ligand concentration in more detail. Table 5 summarizes (Table 6). Surprisingly, by employing ligand 1, an excellent
the influence of the titanium to ligand ratio (Ti/L) in the yield (99%) of the corresponding imine and a high anti-
hydroamination of 1-octyne with sec-butylamine. Here, we Markovnikov selectivity (anti-M:M = 94:6) was obtained.
used aryloxo monodentate ligands 1, 5, 7 and 8, which have However, in the presence of 5 or 6 only low yields of imine
Eur. J. Org. Chem. 2005, 5001–5012
© 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
5005

A. Tillack, V. Khedkar, H. Jiao, M. Beller
FULL PAPER
Table 4. Intermolecular hydroamination of 1-octyne with different amines using Ti(NEt2)4 and ligand 1 in comparison with [Ti(NMe2)2-
(2,6-tBu2-4-Me-C6H2O)2] (24) as catalyst.[a]
Entry
Ligand or Complex (24)
Amine
Conversion [%]
Yield[b] [%]
anti-M:M ratio[c]
1
2
3
4
5
6
7
8
24
1
24
1
24
1
24
1
24
1
1
24
1
1
benzylamine
benzylamine
n-butylamine
n-butylamine
sec-butylamine
sec-butylamine
cyclooctylamine
cyclooctylamine
tert-butylamine
tert-butylamine
tert-butylamine
aniline
100
100
100
100
100
100
100
100
n.r.
50
99 (65)
99
97 (58)
99
98 (71)
98
99 (82)
99
–
50
58
99
25:75 (24:76)
20:80
28:72 (28:72)
25:75
8:92 (16:84)
10:90
16:84 (22:78)
14:86
9
–
10
11[d]
12
13
14[d]
74:26
88:12
4:96
22:78
75
100
100
100
aniline
aniline
99
96
20:80
[a] Reaction conditions: 1.5 mmol 1-octyne, 1.8 mmol amine, 10 mol-% catalyst, Ti/L = 1:2, 2 mL toluene, 100 °C, 24 h, reaction condi-
tions not optimized. [b] Yield is determined by GC analysis with dodecane or hexadecane as internal standard, isolated yield is given in
parenthesis. [c] GC analysis was used to determine regioisomers, the ratio of regiomers of the isolated product after distillation is given
in parenthesis. [d] Ti:L = 1:1.
Table 5. Hydroamination of 1-octyne with sec-butylamine in the presence of Ti(NEt2)4 and monodentate ligands using different Ti:L
ratios.[a]
Entry
Ligand
Ti:L
Conversion [%]
Yield[b] [%]
anti-M:M ratio[c]
1
2
3
4
1
1
1
5
5
5
5
7
7
7
8
8
8
8
1:1
1:2
1:4
1:1
1:2
1:3
1:3
1:1
1:2
1:3
1:1
1:2
1:3
1:4
100
100
100
100
100
n.r.
n.r.
100
100
n.r.
100
100
65
98
98
99
94
97
–
23:77
10:90
8:92
66:34
49:51
–
5
6
7[d]
8
–
–
88
94
-
94
97
63
–
87:13
94:6
-
87:13
94:6
96:4
–
9
10
11
12
13
14
n.r.
[a] Reaction conditions: 1.5 mmol 1-octyne, 1.8 mmol amine, 10 mol-% catalyst, 2 mL toluene, 100 °C, 24 h. [b] Yield is determined by
GC analysis with dodecane or hexadecane as internal standard. [c] GC analysis was used to determine the regioisomers. [d] Reaction
with complex [Ti(NMe2)(2,6-Ph2-C6H3O)3] (25).
5006
www.eurjoc.org
© 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2005, 5001–5012

Titanium-Catalyzed Hydroamination of Terminal Alkynes
Table 6. Hydroamination of methylphenylacetylene with benzylamine using Ti(NEt2)4 and different monodentate ligands.[a]
FULL PAPER
Entry
Ligand
Conversion [%]
Yield[b] [%]
anti-M:M ratio[c]
1
2
3
4
1
5
6
8
100
24
42
99 (42)
24
16
94:6 (91:9)
74:26
65:35
100
51
88:12
[a] Reaction conditions: 1.5 mmol 1-octyne, 1.8 mmol amine, 10 mol-% catalyst, Ti/L = 1:2, 2 mL toluene, 120 °C, 24 h. [b] Yield is
determined by GC analysis with dodecane or hexadecane as internal standard, isolated yield is given in parenthesis. [c] GC analysis was
used to determine the regioisomers, the ratio of regiomers of the isolated product after distillation is given in parenthesis.
and low regioselectivity were observed (Table 6, Entries 2– Both complexes should be in equilibrium during their for-
3). In the presence of 8 the reaction gave 100% conversion mation, and their equilibrium constant (K) is determined
but only 51% yield of the corresponding imine and moder- by their relative Gibbs free energy (∆G = –RTlnK), which
ate regioselectivity (Table 6, Entry 4).
is also the difference between the two competing reaction
energies on the basis of the active catalyst (ArO)2Ti(=N–
CH2Ph) and propyne. The consequence of this is that the
more stable π complex should be higher in concentration
and should also lead to the more dominant product. Since
it was not possible to perform the corresponding frequency
calculation, it was also not possible to obtain the thermal
correction for the Gibbs free energy, therefore, the calcu-
lated relative energies (∆H) are taken as approximate Gibbs
free energies (∆G) to determine the equilibrium constant
and ratio of anti-Markovnikov and Markovnikov π com-
plexes. The working temperature used in the experiment is
373 K. The optimized structures are shown in Figure 1 and
Figure 2, and the calculated energies are given in Table 7.
Selected structural parameters are given in the Table 8 and
Table 9.
Calculations on the Aryloxotitanium-Catalyzed
Hydroamination of Terminal Alkynes
As discussed above, the hydroamination of terminal al-
kynes catalyzed by aryloxotitanium complexes shows high
regioselectivity, which depends significantly on the changes
of the substituent at the benzene ring of the aryloxo ligands.
For example, hydroamination of alkynes with benzylamine
shows regioselectivity in favor of the Markovnikov product
by using 2,6-di-tert-butyl-4-methylphenol ligand 1, while
the anti-Markovnikov product is obtained by using 2,6-di-
isopropylphenol ligand 8. Both ligands differ only in the
substitution at the 2,6-positions. However, does this compa-
rably small difference determine the regioselectivity?
Recently, we studied the combined experimental and
theoretical investigation in the regioselectivite hydroamin-
ation of terminal alkynes by using (η5-C5H5)-
Ti(=NR)(NHR) as the active catalyst.[8a] We have found
that this reaction is highly regioselective, i.e. substituted ani-
lines favor Markovnikov products and tert-butylamine fa-
vors the anti-Markovnikov product. The regioselectivity is
determined by the electrostatic attractive and steric repul-
sive effects in the preformed π complexes. Neither the kin-
etic nor thermodynamic parameters occurring during the
formation of the [2+2]-cycloaddition step determine the re-
gioselectivity.
In our model calculations, we have used propyne as the
terminal alkyne, and benzylamine as the amine, and the
aryloxo ligand (ArO) is modeled by 2,6-diisopropyl-phen-
oxo 8Ј and 2,6-di-tert-butyl-phenoxo ligands 1Ј. On the
Scheme 4. anti-Markovnikov and Markovnikov coordination of
propyne.
For the reaction with 2,6-diisopropylphenoxo 8Ј as the
basis of the experimental findings, the active catalyst has ligand, both the anti-Markovnikov π complex (AM-1) and
been considered to be the bis-aryloxotitaniumimido com- the Markovnikov one (M-1) have been optimized at the
plex, (ArO)2Ti(=N–CH2Ph), which coordinates with the at- B3LYP level of theory with the LANL2DZ and
tacking propyne to form the π complexes in both Markov- LANL2DZ(d) basis sets. The optimized structures are
nikov and anti-Markovnikov ways, as shown in Scheme 4. shown in Figure 1. At B3LYP/LANL2DZ(d), the former is
Eur. J. Org. Chem. 2005, 5001–5012
© 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
5007

A. Tillack, V. Khedkar, H. Jiao, M. Beller
FULL PAPER
Figure 1. Optimized structures for the π complexes and the cyclo-
addition products with 2,6-diisopropylphenoxo ligand 8Ј (hydrogen
atoms are omitted for clarity).
Figure 2. Optimized structures for the π complexes and the cyclo-
addition products with 2,6-di-tert-butylphenoxo ligand 1Ј (hydro-
gen atoms are omitted for clarity).
computed to be lower in energy than the latter by 1.25 kcal/
mol, and, therefore, AM-1 should be more dominant over Both [2+2]-cycloaddition products (AM-P1 and M-P1) are
M-1. This should also be the case for subsequent products. close in energy (0.76 kcal/mol) at B3LYP/LANL2DZ(d).
This energy difference gives a percentage ratio of 84 to 16
In contrast to 2,6-di-tert-butyl-phenoxo 1Ј as ligand, it is
for anti-Markovnikov to Markovnikov products, and this only possible to get the π complexes optimized at B3LYP/
ratio matches the experimental finding (86:14) perfectly. LANL2DZ, and further optimization at B3LYP/LANL(d)
Table 7. Computed total (Etot, au) and relative energies (∆H, kcal/mol) of the π complexes and [2+2]-cycloaddition products.
Isomers
Etot (B3LYP/LANL2DZ)
Etot [B3LYP/LANL2DZ(d)]
∆H
AM-1
M-1
AM-P1
M-P1
AM-2
M-2
–1585.92067
–1585.91997
–1585.94721
–1585.94292
–1743.09715
–1743.10942
–1743.14414
–1743.14566
–1586.24169
–1586.23997
–1586.27616
–1586.27737
[–1743.45068][a]
[–1743.46108][a]
–1743.50695
–1743.50798
0.00
1.25
0.00
–0.76
[0.00][a]
[–6.35][a]
0.00
AM-P2
M-P2
–0.65
[a] Single-point energies at B3LYP/LANL2DZp//B3LYP/LANL2DZ.
Table 8. B3LYP/LANL2DZp optimized structures (in Å and º) for the π complexes and the cycloaddition products with 2,6-diiso-
propylphenoxo ligands.[a]
AM-1
AM-P1
M-1
M-P1
Ti–C1
Ti–C2
Ti–C1/C2
Ti–N1
Ti–O1
2.429 [2.490]
2.370 [2.415]
4.799 [4.905]
1.701 [1.706]
1.838 [1.834]
1.841 [1.836]
178.0 [177.5]
167.3 [171.0]
146.6 [156.4]
176.6 [175.1]
2.197 [2.250]
1.963 [1.991]
4.160 [4.241]
1.882 [1.888]
1.811 [1.809]
1.806 [1.804]
158.3 [155.7]
178.5 [176.6]
170.2 [176.2]
179.5 [179.6]
2.317 [2.328]
2.545 [2.630]
4.862 [4.958]
1.706 [1.711]
1.837 [1.833]
1.832 [1.828]
170.1 [170.2]
169.2 [171.5]
153.1 [159.4]
–6.9 [–6.3]
1.941 [1.969]
2.241 [2.296]
4.182 [4.265]
1.896 [1.901]
1.810 [1.808]
1.804 [1.804]
152.6 [150.6]
175.2 [175.7]
176.3 [178.5]
0.1 [0.1]
Ti–O2
Ti–N1–C3
Ti–O1–C4
Ti–O1–C5
N1–Ti–C1–C2
[a] B3LYP/LANL2DZ values are given in square brackets.
5008
www.eurjoc.org
© 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2005, 5001–5012

Titanium-Catalyzed Hydroamination of Terminal Alkynes
FULL PAPER
Table 9. B3LYP/LANL2DZp optimized structures (in Å and º) for the π complexes and the cycloaddition products with 2,6-di-tert-
butylphenoxo ligands.[a]
AM-2
AM-P2
M-2
M-P2
Ti–C1
Ti–C2
Ti–C1/C2
Ti–N1
Ti–O1
[2.536]
[2.542]
[5.078]
[1.697]
[1.864]
[1.858]
[174.9]
[163.9]
[150.5]
[167.4]
2.204 [2.257]
1.963 [1.991]
4.167 [4.248]
1.888 [1.891]
1.818 [1.816]
1.821 [1.818]
157.9 [155.4]
178.3 [178.2]
175.6 [175.7]
–178.8 [–178.8]
[2.231]
[2.530]
[4.761]
[1.720]
[1.845]
[1.840]
[173.3]
[174.9]
[171.6]
[4.5]
1.940 [1.965]
2.252 [2.307]
4.192 [4.272]
1.904 [1.909]
1.817 [1.814]
1.820 [1.817]
152.8 [150.7]
179.6 [179.3]
176.4 [178.3]
1.9 [1.7]
Ti–O2
Ti–N1–C3
Ti–O1–C4
Ti–O1–C5
N1–Ti–C1–C2
[a] B3LYP/LANL2DZ values are given in square brackets.
leads directly to the [2+2]-product, indicating the rather low kov isomer (AM-2) (4.761 vs. 5.078 Å, Table 9). Thus, the
activation barrier for the cycloaddition step. In order to Ti–C distance correlates nicely with the relative stability of
make comparisons, single-point energy calculations at the these π complexes. This clearly shows that the observed re-
B3LYP/LANL2DZ(d) level with the B3LYP/LANL2DZ gioselectivity is determined by the steric difference between
optimized geometries have been carried out. The optimized tert-butyl and isopropyl substituents at the 2,6-position of
structures are shown in Figure 2. The computed energy dif- the aryloxo ligand.
ferences indicate that the anti-Markovnikov π complex
In summary, we presented a detailed study on the tita-
(AM-2) is computed to be higher in energy than the Mar- nium-catalyzed hydroamination of terminal alkynes in the
kovnikov one (M-2) by 6.53 kcal/mol, and this energy dif- presence of 17 different alkoxo and aryloxo ligands. By
ference completely favors the Markovnikov product over using the “right” ligand all the reactions shown can be per-
the anti-Markovnikov product (Ͼ99%). However, this cal- formed in good to excellent yield and with high regioselec-
culated result agrees only qualitatively with the experimen- tivity. Interestingly, by slight variations of the sterics of the
tal finding of the same trend, and quantitatively overesti- aryloxo ligands the selectivity can be controlled and a selec-
mated the regioselectivity, since the observed regioselectiv- tivity switch can even be obtained. Such a control of re-
ity is 80 to 20. As expected, both cycloaddition products gioselectivity is still rare for addition processes to unsatu-
(AM-P2 and M-P2) are close in energy (0.65 kcal/mol).
rated compounds. Computational studies demonstrate that
Apart from the agreement between theory and experi- the observed change of regioselectivity is determined by ste-
ment, it is interesting to understand the driving force for ric factors.
the regioselectivity. Since both aryloxo ligands differ only
in the substituents in the 2,6-positions, e.g. tert-butyl and
isopropyl, it is expected that their regioselectivity should be
directly related with the steric effect of these substitutes.
Experimental Section
From a structural point of view, there are many possible General Remarks: All reactions were carried out under argon.
Chemicals were obtained from Aldrich, Fluka, Acros and Strem
and unless otherwise noted were used without further purification.
Amines were distilled from CaH2. Alkynes were degassed, flushed
with argon and stored over molecular sieves (4 Å). Absolute sol-
vents were purchased from Fluka. All operations were carried out
under an argon atmosphere. All compounds were characterized by
explanations for the reduction in strain energy caused by
the steric effect, (a) the C–O–Ti angles of the aryloxo li-
gands, (b) the Ti–N–C angle of the imido ligand, (c) the
bulky substituents at 2,6-positions of the aryloxo ligands,
(d) the interaction between the propyne methyl group and
the phenyl group of imido ligands in the Markovnikov π
complexes, and (e) the interaction between the propyne
methyl group and the tert-butyl or isopropyl groups of the
aryloxo ligands. All these factors work together, and their
1
1H NMR, 13C NMR, MS and IR spectroscopy. H and 13C NMR
spectra were recorded with a Bruker ARX 400 spectrometer. The
1H and 13C NMR chemical shifts are reported relative to the center
of solvent peak [CDCl3: 7.25 (1H), 77.0 (13C); [D8]THF: 1.73 (1H),
net effect is reflected by the distances of Ti to the alkyne 25.2 (13C)]. EI mass spectra were recorded with an AMD 402 spec-
trometer (70 eV, AMD Intectra GmbH). IR spectra were recorded
with a Nicolet Magna 550. Elemental analyses were determined by
C/H/N/S-Analysator 932 (Leco). GC analysis was performed with
a Hewlett Packard HP 6890 chromatograph with a 30 m HP5 col-
umn. All yields reported in Tables 1–6 refer to GC yields using
dodecane or hexadecane as an internal standard. We have described
the imine N-tert-butyl-octylidene-amine (15a) previously.[8c] The ti-
tanium catalysts bis(2,6-di-tert-butyl-4-methylphenoxo)bis(dimeth-
ylamido)titanium ([Ti(NMe2)2(2,6-tBu2-4-Me-C6H2O)2]) (24)[11b]
and bis(2,6-di-phenylphenoxo)titanium dichloride ([TiCl2(2,6-Ph2-
triple bond or sum of the two Ti–C distances, i.e.; the
shorter the distances, the stronger the interaction, and the
more stable the π complex.
On this basis, it is clearly seen that the more stable anti-
Markovnikov π complex (AM-1) of the isopropyl-substi-
tuted aryloxo ligand has a shorter Ti–C distance than the
less stable Markovnikov isomer (M-1) (4.905 vs. 4.958 Å,
Table 8). For the tert-butyl-substituted aryloxo ligand, the
Ti–C distance of the more stable Markovnikov π complex
(M-2) is shorter than that of the less stable anti-Markovni- C6H3O)2])[13] were prepared following literature procedures.
Eur. J. Org. Chem. 2005, 5001–5012
© 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
5009

A. Tillack, V. Khedkar, H. Jiao, M. Beller
FULL PAPER
Tris(2,6-diphenylphenoxo)(dimethylamido)titanium ([Ti(NMe2)(2,6-
Ph2-C6H3O)3]) (25): 2,6-Diphenylphenol (3.50 g, 14.1 mmol) was
added to tetrakis(dimethylamido)titanium (1.05 g, 4.7 mmol) in
toluene (75 mL). While stirring at room temperature for 2 days
orange microcrystals precipitated. The resulting orange precipitate
was collected by filtration, washed with toluene (10 mL), and dried
under vacuum. M.p. 112–117 °C (dec.); yield: 1.91 g (49%). 1H
N-Benzyloctylideneamine/N-benzyl-2-octylideneamine (ratio 87:13)
(13a/13b): 2,6-Diisopropylphenol was used as the ligand. Colorless
oil; b.p. 97–98 °C/0.1 mbar; 41% (0.98 g) isolated yield (72% GC
yield). Data for anti-Markovnikov product (13a): 1H NMR
(CDCl3, 400 MHz): δ = 7.78 (tt, J = 5.0, 1.4 Hz, 1 H), 7.23–7.33
(m, 5 H), 4.56 (s, 2 H), 2.27–2.35 (m, 2 H), 1.51–1.62 (m, 2 H),
1.21–1.39 (m, 8 H), 0.88 (m, 3 H) ppm. 13C NMR (CDCl3,
NMR ([D8]THF, 400 MHz): δ = 6.95–7.20 (m, 39 H), 1.38 (s, 6 100 MHz): δ = 166.4, 139.4, 128.4, 127.9, 126.8, 65.1, 36.0, 31.7,
H) ppm. 13C NMR ([D8]THF, 100 MHz): δ = 159.8, 140.4, 133.4,
130.7, 130.2, 128.7, 127.3, 121.8, 45.6 ppm. MS (EI, 70 eV) m/z
(rel. intensity): 827 (21) [M+], 583 (57), 582 (100), 538 (71), 537
29.2, 29.0, 26.0, 22.6, 14.0 ppm. MS (EI, 70 eV) m/z (rel. intensity):
217 (1) [M+], 147 (23), 146 (19), 133 (52), 132 (23), 91 (100), 65
(16), 41 (14), 28 (25). FT IR (neat): 1668 (C=N) cm–1. HRMS:
(95), 293 (24), 291 (16), 246 (49), 245 (17). FT IR (nujol): 3050, Calcd. for C15H23N: 217.18304; found 217.18328.
3024, 2777, 1598, 1494, 1408, 1234, 905, 756, 720, 700 cm–1.
C56H45NO3Ti (827.83): calcd. C 81.25, H 5.48, N 1.69; found C
80.72, H 5.30, N 1.66.
N-Benzyl-2-octylideneamine/N-Benzyloctylideneamine (ratio 76:24)
(13b/13a): Complex [Ti(NMe2)2(2,6-tBu2-4-Me-C6H2O)2] was used.
Colorless oil; b.p. 98 °C/0.1 mbar; 65% (1.56 g) isolated yield (99%
GC yield). Data for Markovnikov product (13b): 1H NMR (CDCl3,
400 MHz): δ = 7.20–7.35 (m, 5 H), 4.51 + 4.48 (s, 2 H), 2.27–2.35
(m, 2 H), 2.07 + 1.89 (s, 3 H), 1.51–1.63 (m, 2 H), 1.31 (m, 6 H),
0.88 (m, 3 H) ppm. 13C NMR (CDCl3, 100 MHz): δ = 171.8 +
171.4, 140.5, 128.3, 127.7 + 127.6, 126.4, 55.0 + 54.6, 42.9, 31.6 +
31.5, 29.4 + 29.1, 26.6 + 26.3, 22.5, 17.5, 14.0. MS (EI, 70 eV)
m/z (rel. intensity): 217 (2) [M+], 160 (15), 147 (88), 146 (57), 91
(100). FT IR (neat): 1663 (C=N) cm–1.
General Procedure for Hydroamination Reactions: In an Ace-pres-
sure tube under an argon atmosphere the ligand (2,6-diisopro-
pylphenol, 2,6-di-sec-butylphenol or 2,6-di-tert-butyl-4-methylphe-
nol) (2.2 mmol) was dissolved in 7.5 mL toluene. Amine
(13.5 mmol), 1-octyne (11.3 mmol) and Ti(NEt2)4 (1.1 mmol) were
added to this solution. When preparing the Markovnikov isomers
as the main product, complex [Ti(NMe2)2(2,6-tBu2-4-Me-
C6H2O)2] (1.1 mmol) was used and dissolved in 7.5 mL toluene.
The pressure tube was fitted with a Teflon cap and heated at 100 °C
for 24 h in an oil bath. Afterwards all volatiles were removed and
the product was isolated by fractional distillation in vacuo. The
imine products were isolated as a mixture of anti-Markovnikov and
Markovnikov products. The ratio of the regioisomers was deter-
mined by GC analysis. After isolation (via distillation) this ratio
might differ because of the different boiling points of the regioiso-
mers (see Tables 1, 2, 4, and 6). The Markovnikov product is a
mixture of the E- and Z-isomers; thus, some NMR resonances were
present as two signals (indicated below by “+”), representing both
isomers.
N-Butyloctylideneamine/N-Butyl-2-octylidene-2-amine (ratio 91:9)
(14a/14b): 2,6-Di-sec-butylphenol was used as the ligand. Colorless
oil; b.p. 57–58 °C/0.13 mbar; 50% (1.05 g) isolated yield (88% GC
yield). Data for anti-Markovnikov product (14a): 1H NMR
(CDCl3, 400 MHz): δ = 7.59 (tt, J = 5.0, 1.4 Hz, 1 H), 3.32 (dt, J
= 7.1, 1.0 Hz, 2 H), 2.16–2.22 (m, 2 H), 1.44–1.58 (m, 4 H), 1.20–
1.33 (m, 10 H), 0.88 (t, J = 7.3 Hz, 3 H), 0.84 (t, J = 7.1 Hz, 3
H) ppm. 13C NMR (CDCl3, 100 MHz): δ = 164.8, 61.1, 35.8, 32.8,
31.7, 29.2, 29.0, 26.1, 22.6, 20.3, 14.0, 13.8 ppm. MS (EI, 70 eV)
m/z (rel. intensity): 183 (0.7) [M+], 182 (1), 140 (10), 112 (25), 99
(32), 84 (100), 70 (19), 57 (69), 56 (40), 55 (17), 43 (16), 42 (26), 41
(46), 29 (42). FT IR (neat): 1671 (C=N) cm–1. HRMS: Calcd. for
C12H25N: 183.19870: found 183.19790.
N-(sec-Butyl)octylideneamine/N-(sec-Butyl)-2-octylidene-2-amine
(ratio 96:4) (12a/12b): 2,6-Di-sec-butylphenol was used as the li-
gand. Colorless oil; b.p. 41–43 °C/0.1 mbar; 70% (1.47 g) isolated
yield (98% GC yield). Data for anti-Markovnikov product (12a):
1H NMR (CDCl3, 400 MHz): δ = 7.57 (t, J = 5.2, 1 H), 2.88 (sext,
J = 6.5 Hz, 1 H), 2.20 (m, 2 H), 1.46 (quin, J = 7.5 Hz, 4 H), 1.17–
1.35 (m, 8 H), 1.11 (d, J = 6.3 Hz, 3 H), 0.85 (m, 3 H), 0.76 (t, J
= 7.5 Hz, 3 H) ppm. 13C NMR (CDCl3, 100 MHz): δ = 162.9, 67.9,
35.7, 31.7, 30.4, 29.2, 29.0, 26.4, 22.6, 22.4, 14.0, 11.0 ppm. MS
(EI, 70 eV) m/z (rel. intensity): 183 (1) [M+], 182 (2), 154 (38), 112
(24), 99 (100), 84 (44), 71(15), 70 (37), 69 (30), 57 (50), 56 (44), 55
(33), 44 (52), 43 (29), 42 (27), 41 (75), 29 (47), 28 (40), 27 (32). FT
IR (neat): 1669 (C=N) cm–1. HRMS: Calcd. for C12H25N:
183.19870; found 183.19762.
N-Butyl-2-octylidene-2-amine/N-butyloctylideneamine (ratio 72:28)
(14b/14a): Complex [Ti(NMe2)2(2,6-tBu2-4-Me-C6H2O)2] was used.
Colorless oil; b.p. 45–46 °C/0.1 mbar; 58% (1.22 g) isolated yield
(97% GC yield). Data for Markovnikov product (14b): 1H NMR
(CDCl3, 400 MHz): δ = 3.15–3.25 (m, 2 H), 2.19 (m, 2 H), 1.95 +
1.76 (s, 3 H), 1.42–1.62 (m, 4 H), 1.20–1.39 (m, 8 H), 0.81–0.95 (m,
6 H) ppm. 13C NMR (CDCl3, 100 MHz): δ = 170.3 + 169.8, 51.1
+ 50.5, 42.9, 33.3 + 33.0, 32.1 + 31.6, 29.4 + 29.1, 27.0 + 26.7,
22.5, 20.7, 16.8, 13.99, 13.96 ppm. MS (EI, 70 eV) m/z (rel. inten-
sity): 183 (2) [M+], 182 (3), 140 (21), 126 (43), 113 (41), 98 (100),
84 (23), 71 (80), 70 (16), 57 (22), 56 (21), 55 (10), 42 (36), 41 (23),
29 (12). FT IR (neat): 1662 (C=N) cm–1.
N-(sec-Butyl)-2-octylidene-2-amine/N-(sec-Butyl)octylideneamine
(ratio 84:16) (12b/12a): Complex [Ti(NMe2)2(2,6-tBu2-4-Me-
N-Cyclooctyloctylideneamine/N-Cyclooctyl-2-octylidene-2-amine
(ratio 96:4) (16a/16b): 2,6-Diisopropylphenol was used as the li-
C6H2O)2] was used. Colorless oil; b.p. 49–50 °C/0.1 mbar; 71% gand. Colorless oil; b.p. 95 °C/0.14 mbar; 65 % (1.76 g) isolated
(1.49 g) isolated yield (98 % GC yield). Data for Markovnikov
product (12b): 1H NMR (CDCl3, 400 MHz): δ = 3.34 + 3.28 (sext,
J = 6.5 Hz, 1 H), 2.13–2.24 (m, 2 H), 1.94 + 1.77 (s, 3 H), 1.39–
1.52 (m, 4 H), 1.21–1.33 (m, 6 H), 1.02 + 1.03 (d, J = 6.3 Hz, 3 H),
0.82–0.89 (m, 3 H), 0.73–0.81 (m, 3 H) ppm. 13C NMR (CDCl3,
100 MHz): δ = 167.8 + 167.4, 56.4 + 55.9, 43.1, 32.1 + 31.7, 31.0
+ 30.9, 29.5 + 29.0, 27.1 + 26.9, 22.5, 21.7 + 21.2, 16.5, 14.0,
11.1 ppm. MS (EI) m/z (%): 183 (4) [M+], 182 (6), 154 (48), 140
yield (94% GC yield). Data for anti-Markovnikov product (16a):
1H NMR (CDCl3, 400 MHz): δ = 7.55 (t, J = 5.4 Hz, 1 H), 3.07
(m, 1 H), 2.16 (m, 2 H), 1.65–1.77 (m, 4 H), 1.41–1.58 (m, 12 H),
1.20–1.32 (m, 8 H), 0.85 (m, 3 H) ppm. 13C NMR (CDCl3,
100 MHz): δ = 161.5, 71.3, 35.7, 34.2, 31.7, 29.1, 29.0, 27.2, 26.3,
25.6, 24.1, 22.6, 14.0 ppm. MS (EI, 70 eV) m/z (rel. intensity): 237
(3) [M+], 236 (2), 194 (14), 166 (32), 153 (42), 152 (37), 138 (95),
128 (23), 124 (17), 110 (37), 96 (23), 82 (50), 69 (78), 67 (30), 56
(20), 126 (40), 113 (88), 112 (21), 98 (100), 85 (39), 84 (66), 70 (32), (44), 55 (68), 44 (56), 43 (57), 42 (19), 41 (100), 29 (43). FT IR
58 (31), 57 (26), 55 (14), 42 (59), 41 (31), 29 (15). FT IR (neat):
(neat): 1668 (C=N) cm–1. HRMS: Calcd. for C16H31N: 237.24565:
found 237.24508.
1661 (C=N) cm–1.
5010
www.eurjoc.org
© 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2005, 5001–5012

Titanium-Catalyzed Hydroamination of Terminal Alkynes
FULL PAPER
N-Cyclooctyl-2-octylidene-2-amine/N-Cyclooctyloctylideneamine
(ratio 78:22) (16b/16a): Complex [Ti(NMe2)2(2,6-tBu2-4-Me-
C6H2O)2] was used. Colorless oil; b.p. 88 °C/0.1 mbar; 82% (2.21 g)
isolated yield (99 % GC yield). Data for Markovnikov product
holz (all from the IfOK, Rostock) for their excellent technical and
analytical support.
[1] a) J. Seayad, A. Tillack, M. Beller, Angew. Chem. Int. Ed. 2004,
43, 3368; b) F. Alonso, I. P. Beletskaya, M. Yus, Chem. Rev.
2004, 104, 3079.
1
(16b): H NMR (CDCl3, 400 MHz): δ = 3.37–3.52 (m, 1 H), 2.12–
2.21 (m, 2 H), 1.91 + 1.77 (s, 3 H), 1.60–1.76 (m, 2 H), 1.41–1.59
(m, 12 H), 1.20–1.36 (m, 8 H), 0.81–0.93 (m, 3 H) ppm. 13C NMR
(CDCl3, 100 MHz): δ = 165.9 + 165.6, 59.9 + 59.3, 43.0, 34.3 +
34.0, 31.7 + 31.6, 29.2 + 29.0, 27.1, 26.9 + 26.8, 26.0 + 25.9, 24.5,
22.5, 16.4, 14.0 ppm. MS (EI, 70 eV) m/z (rel. intensity): 237 (10)
[M+], 222 (16), 194 (25), 180 (91), 167 (48), 166 (42), 154 (24), 152
(100), 128 (68), 124 (18), 110 (23), 96 (22), 82 (27), 70 (21), 69 (53),
58 (34), 55 (37), 41 (36). FT IR (neat): 1659 (C=N) cm–1.
[2] For an excellent review see: a) P. Eilbracht, L. Bärfacker, C.
Buss, C. Hollmann, B. E. Kitsos-Rzychon, C. L. Kranemann,
T. Rische, R. Roggenbuck, A. Schmidt, Chem. Rev. 1999, 99,
3329. Recent examples from our group: b) A. Moballigh, R.
Jackstell, M. Beller, Tetrahedron Lett. 2004, 45, 869; c) M.
Ahmed, A. M. Seayad, R. Jackstell, M. Beller, Angew. Chem.
Int. Ed. 2003, 42, 5615; d) A. M. Seayad, K. Selvakumar, M.
Ahmed, M. Beller, Tetrahedron Lett. 2003, 44, 1679; e) M.
Ahmed, A. M. Seayad, R. Jackstell, M. Beller, J. Am. Chem.
Soc. 2003, 125, 10311; f) A. M. Seayad, M. Ahmed, H. Klein,
R. Jackstell, T. Gross, M. Beller, Science 2002, 297, 1676; g) B.
Zimmermann, J. Herwig, M. Beller, Angew. Chem. 1999, 111,
2515; Angew. Chem. Int. Ed. 1999, 38, 2372.
[3] Rhodium-catalyzed amination of olefins: a) H. Trauthwein, A.
Tillack, M. Beller, Chem. Commun. 1999, 2029; b) M. Beller, H.
Trauthwein, M. Eichberger, C. Breindl, J. Herwig, T. E. Müller,
O. R. Thiel, Chem. Eur. J. 1999, 5, 1306. For base-catalyzed
hydroaminations of styrenes see: c) K. Kumar, D. Michalik, I.
Garcia Castro, A. Tillack, A. Zapf, M. Arlt, T. Heinrich, H.
Böttcher, M. Beller, Chem. Eur. J. 2004, 10, 746; d) M. Beller,
C. Breindl, Chemosphere 2001, 43, 21; e) C. G. Hartung, C.
Breindl, A. Tillack, M. Beller, Tetrahedron 2000, 56, 5157; f)
M. Beller, C. Breindl, T. H. Riermeier, M. Eichberger, H.
Trauthwein, Angew. Chem. 1998, 110, 3571; Angew. Chem. Int.
Ed. 1998, 37, 3389; g) M. Beller, C. Breindl, Tetrahedron 1998,
54, 6359.
N-(2-Octylidene)aniline/N-(Octylidene)aniline (ratio 70:30) (17b/
17a): 2,6-Diisopropylphenol was used as the ligand. Colorless oil;
b.p. 79–80 °C/0.12 mbar; 37 % (0.85 g) isolated yield (86% GC
yield). Data for Markovnikov product (17b): 1H NMR (CDCl3,
400 MHz): δ = 7.24–7.30 (m, 2 H), 6.98–7.04 (m, 1 H), 6.65–6.70
(m, 2 H), 2.37–2.42 + 2.08–2.13 (m, 2 H), 2.14 + 1.76 (s, 3 H),
1.60–1.70 + 1.42–1.50 (m, 2 H), 1.14–1.42 (m, 6 H), 0.83–0.90 (m,
3 H) ppm. 13C NMR (CDCl3, 100 MHz): δ = 172.7 + 172.2, 151.6
+ 151.1, 128.8 + 128.7, 122.9 + 122.8, 119.5, 41.7 + 34.0, 31.7 +
31.4, 29.1 + 29.0, 26.8 + 26.3, 22.6 + 22.4, 19.4, 14.0 + 13.9 ppm.
MS (EI, 70 eV) m/z (rel. intensity): 203 (14) [M+], 188 (4), 146 (30),
133 (93), 132 (93), 119 (41), 118 (69), 93 (31), 92 (29), 77 (100), 51
(31), 43 (22), 42 (25), 41 (30), 39 (22), 29 (27), 28 (20), 27 (25). FT
IR (neat): 1662 (C=N) cm–1. HRMS: Calcd. for C14H21N:
203.16740; found 203.16682.
N-Benzyl-(1-methyl-2-phenylethylidene)amine/N-Benzyl(1-phenyl-
propylidene)amine (ratio = 91:9) (26a/26b): 2,6-Di-tert-butyl-4-
methylphenol was used as the ligand. Colorless oil; b.p. 104 °C/
0.13 mbar; 42% (1.05 g) isolated yield (99% GC yield). Data for
anti-Markovnikov product (26a): 1H NMR (CDCl3, 400 MHz): δ
= 7.12–7.25 (m, 10 H), 4.41 (s, 2 H), 3.55 (s, 2 H), 1.73 (s, 3 H) ppm.
13C NMR (CDCl3, 100 MHz): δ = 169.7, 140.3, 137.5, 129.1, 128.5,
128.4, 127.7, 126.5, 55.4, 49.7, 17.0 ppm. MS (EI, 70 eV) m/z (rel.
intensity): 223 (6) [M+], 132 (21), 91 (100), 65 (16). FT IR (neat):
1658 (C=N) cm–1. HRMS: Calcd. for C16H17N: 223.13609; found
223.13533.
[4] For transition metal-catalyzed intermolecular hydroaminations
of alkynes see: a) T. Shimada, G. B. Bajracharya, Y. Yamam-
oto, Eur. J. Org. Chem. 2005, 59; b) D. P. Klein, A. Ellern, R. J.
Angelici, Organometallics 2004, 23, 5662; c) G. V. Shanbhag,
S. B. Halligudi, J. Mol. Catal. A: Chem. 2004, 222, 223; d) L. L.
Anderson, J. Arnold, R. G. Bergman, Org. Lett. 2004, 6, 2519;
e) S. Breitenlechner, M. Fleck, T. E. Müller, A. Suppan, J. Mol.
Catal. A: Chem. 2004, 214, 175; f) F. Pohlki, S. Doye, Chem.
Soc. Rev. 2003, 32, 104; g) E. Mizushima, T. Hayashi, M.
Tanaka, Org. Lett. 2003, 5, 3349; h) M. Beller, C. Breindl, M.
Eichberger, C. G. Hartung, J. Seayad, O. Thiel, A. Tillack, H.
Trauthwein, Synlett 2002, 1579; i) T. Shimada, Y. Yamamoto,
J. Am. Chem. Soc. 2002, 124, 12670; j) C. G. Hartung, H.
Trauthwein, A. Tillack, M. Beller, J. Org. Chem. 2001, 66, 6339;
k) M. Tokunaga, Y. Wakatsuki, J. Synth. Org. Chem. Jpn. 2000,
58, 587; l) M. Tokunaga, M. Eckert, Y. Wakatsuki, Angew.
Chem. Int. Ed. 1999, 38, 3222; m) Y. Uchimaru, Chem. Com-
mun. 1999, 1133; n) I. Kadota, A. Shibuya, L. M. Lutete, Y.
Yamamoto, J. Org. Chem. 1999, 64, 4570; o) T. E. Müller, M.
Beller, Chem. Rev. 1998, 98, 675; p) A. M. Baranger, P. J.
Walsh, R. G. Bergman, J. Am. Chem. Soc. 1993, 115, 2753; q)
P. J. Walsh, A. M. Baranger, R. G. Bergman, J. Am. Chem. Soc.
1992, 114, 1708; r) J. Barluenga, A. Aznar, R. Liz, R. Rodes,
J. Chem. Soc., Perkin Trans. 1 1980, 2732. For lanthanide- and
actinide-catalyzed intermolecular hydroaminations of alkynes:
s) S. Hong, T. J. Marks, Acc. Chem. Res. 2004, 37, 673; t) J.-S.
Ryu, G. Y. Li, T. J. Marks, J. Am. Chem. Soc. 2003, 125, 12584;
u) J. Wang, A. K. Dash, M. Kapon, J.-C. Berthet, M. Ephriti-
khine, M. S. Eisen, Chem. Eur. J. 2002, 8, 5384; v) T. Straub,
A. Haskel, T. G. Neyround, M. Kapon, M. Botoshansky, M. S.
Eisen, Organometallics 2001, 20, 5017; w) Y. Li, T. J. Marks, J.
Am. Chem. Soc. 1998, 120, 1757; x) Y. Li, T. J. Marks, Organo-
metallics 1996, 15, 3770. For base-catalyzed intermolecular hy-
droaminations see: y) J. Seayad, A. Tillack, C. G. Hartung, M.
Beller, Adv. Synth. Catal. 2002, 344, 795; z) D. Tzalis, C. Kora-
din, P. Knochel, Tetrahedron Lett. 1999, 40, 6193.
Computational Details: In order to understand the origin of the
observed regioselectivity by employing aryloxotitanium complexes
in the hydroamination of terminal alkynes, density functional
theory[14] calculations have been performed. All structures have
been optimized at the B3LYP density functional level of theory in
combination with the LANL2DZ basis set, further calculations
have been done with the extended L2NL2DZ basis set including a
set of polarization functions [LANL2DZ(d)].[15] Due to the large
size and low symmetry, it has not been possible to carry out the
frequency calculation for characterizing the optimized structures as
energy minimum structures. However, all optimized structures are
considered to be energy minima since they all have positive eigen-
values from the Hessian calculations.[16] All calculations have been
done with the Gaussian 98 program.[17]
Acknowledgments
This work has been supported by the German Bundesland Meck-
lenburg-Vorpommern, the BMBF (Bundesministerium für Bildung
und Forschung) and the Fonds der Chemischen Industrie (FCI).
We thank Professor Dr. M. Michalik, Dr. W. Baumann, Mrs. C.
Mewes, Mrs. H. Baudisch, Mrs. A. Lehmann, and Mrs. S. Buch-
[5] a) J. S. Johnson, R. G. Bergman, J. Am. Chem. Soc. 2001, 123,
2923; b) B. F. Straub, R. G. Bergman, Angew. Chem. Int. Ed.
Eur. J. Org. Chem. 2005, 5001–5012
© 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
5011

A. Tillack, V. Khedkar, H. Jiao, M. Beller
FULL PAPER
2001, 40, 4632; c) J. L. Polse, R. A. Andersen, R. G. Bergman,
J. Am. Chem. Soc. 1998, 120, 13405.
[6] a) A. Heutling, R. Severin, S. Doye, Synthesis 2005, 1200; b)
A. Heutling, F. Pohlki, I. Bytschkov, S. Doye, Angew. Chem.
Beller, Tetrahedron Lett. 2004, 45, 3123; c) V. Khedkar, A. Til-
lack, M. Beller, Org. Lett. 2003, 5, 4767.
[12] A. Tillack, V. Khedkar, M. Beller, Tetrahedron Lett. 2004, 45,
8875.
Int. Ed. 2005, 44, 2951; c) S. Doye, Synlett 2004, 1653; d) A. [13] a) S. A. Waratuke, M. G. Thorn, P. E. Fanwick, A. P. Rothwell,
Heutling, F. Pohlki, S. Doye, Chem. Eur. J. 2004, 10, 3059; e)
F. Pohlki, I. Bytschkov, H. Siebeneicher, A. Heutling, W. A.
König, S. Doye, Eur. J. Org. Chem. 2004, 1967; f) I. Bytschkov,
S. Doye, Eur. J. Org. Chem. 2003, 935; g) A. Heutling, S. Doye,
J. Org. Chem. 2002, 67, 1961; h) F. Pohlki, A. Heutling, I.
Bytschkov, T. Hotopp, S. Doye, Synlett 2002, 799; i) F. Pohlki,
S. Doye, Angew. Chem. Int. Ed. 2001, 40, 2305; j) E. Haak, I.
Bytschkov, S. Doye, Angew. Chem. Int. Ed. 1999, 38, 3389.
[7] a) A. L. Odom, Dalton Trans. 2005, 225; b) Y. Li, Y. Shi, A. L.
Odom, J. Am. Chem. Soc. 2004, 126, 1794; c) Y. Shi, C. Hall,
J. T. Ciszewski, C. Cao, A. L. Odom, Chem. Commun. 2003,
584; d) Y. Shi, T. Ciszewski, A. L. Odom, Organometallics
2002, 21, 5148; e) Y. Shi, J. T. Ciszewski, A. L. Odom, Organo-
metallics 2001, 20, 3967.
[8] a) A. Tillack, H. Jiao, I. Garcia Castro, C. G. Hartung, M.
Beller, Chem. Eur. J. 2004, 10, 2409; b) I. Garcia Castro, A.
Tillack, C. G. Hartung, M. Beller, Tetrahedron Lett. 2003, 44,
3217; c) A. Tillack, I. Garcia Castro, C. G. Hartung, M. Beller,
Angew. Chem. Int. Ed. 2002, 41, 2541.
[9] a) Z. Zhang, L. L. Schafer, Org. Lett. 2003, 5, 4733; b) C. Li,
R. K. Thomson, B. Gillon, B. O. Patrick, L. L. Schafer, Chem.
Commun. 2003, 2462.
[10] a) C. Lorber, R. Choukroun, L. Vendier, Organometallics 2004,
23, 1845; b) B. D. Ward, A. Maisse-Francois, P. Mountford,
L. H. Gade, Chem. Commun. 2004, 704; c) L. Ackermann, L. T.
Kasper, C. J. Gschrei, Chem. Commun. 2004, 2824; d) L. Acker-
mann, L. T. Kasper, C. J. Gschrei, Org. Lett. 2004, 6, 2515; e)
L. Ackermann, Organometallics 2003, 22, 4367; f) T.-G. Ong,
G. P. A. Yap, D. S. Richeson, Organometallics 2002, 21, 2839;
g) J. E. Hill, R. D. Profilet, P. E. Fanwick, I. P. Rothwell, An-
gew. Chem. Int. Ed. Engl. 1990, 29, 664.
I. P. Rothwell, J. Am. Chem. Soc. 1999, 121, 9111; b) J. R.
Dilworth, J. Hanich, J. Beck, J. Strähle, J. Organomet. Chem.
1986, 315, C9–C12.
[14] a) A. D. Becke, J. Chem. Phys. 1993, 98, 5648; b) P. J. Stevens,
F. J. Devlin, C. F. Chablowski, M. J. Frisch, J. Phys. Chem.
1994, 98, 11623.
[15] a) P. J. Hay, W. R. Wadt, J. Chem. Phys. 1985, 82, 299; b)
T. H.Dunning Jr., P. J. Hay, Modern Theoretical Chemistry
(Ed.: Schaefer, H. F. III.), Plenum: New York, 1976, p. 1. For
polarization functions see: c) S. Huzinaga, J. Anzelm, M. Klo-
bukowski, E. Radzio-Andzelm, Y. Sakai, H. Tatewaki,
Gaussian Basis Sets for Molecular Calculations, Elsevier: Am-
sterdam, 1984.
[16] J. B. Foresman, E. Frisch, Exploring Chemistry with Electronic
Structure Methods: A Guide to Using Gaussian, 2nd ed.;
Gaussian, Inc. Pittsburgh PA, 1996.
[17] M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria,
M. A. Robb, J. R. Cheeseman, V. G. Zakrzewski, J. A. Mont-
gomery, R. E. Stratmann, J. C. Burant, S. Dapprich, J. M. Mil-
lam, A. D. Daniels, K. N. Kudin, M. C. Strain, O. Farkas, J.
Tomasi, V. Barone, M. Cossi, R. Cammi, B. Mennucci, C. Pom-
elli, C. Adamo, S. Clifford, J. Ochterski, G. A. Petersson, P. Y.
Ayala, Q. Cui, K. Morokuma, D. K. Malick, A. D. Rabuck, K.
Raghavachari, J. B. Foresman, J. Cioslowski, J. V. Ortiz, B. B.
Stefanov, G. Liu, A. Liashenko, P. Piskorz, I. Komaromi, R.
Gomperts, R. L. Martin, D. J. Fox, T. Keith, M. A. Al-Laham,
C. Y. Peng, A. Nanayakkara, C. Gonzalez, M. Challacombe,
P. M. W. Gill, B. G. Johnson, W. Chen, M. W. Wong, J. L.
Andres, M. Head-Gordon, E. S. Replogle, J. A. Pople, Gaussian
98, Gaussian, Inc., Pittsburgh PA, 1998.
Received: June 10, 2005
Published Online: October 19, 2005
[11] a) V. Khedkar, A. Tillack, M. Michalik, M. Beller, Tetrahedron
2005, 61, 7622; b) V. Khedkar, A. Tillack, M. Michalik, M.
5012
www.eurjoc.org
© 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2005, 5001–5012