Full Papers
These performances were only
slightly affected after decreasing
the catalyst loading to [1]0/
[metal] up to 50000 (entries 1,2
vs. 3,4). In contrast, the nature
of the solvent (entries 2 vs. 7–9)
and catalyst precursor (entries 2,
11, and 14) appeared more im-
portant; in fact, significantly
higher amounts of the hydroge-
nation product undecanenitrile
(4)—the only side-product ob-
served besides those from the
isomerization of 1 (see the dis-
cussion below)—were observed
with [Ir(acac)(cod)] as precursor,
in contrast to [IrCl(cod)]2 and
[IrCpMe(cod)] as precursors, for
reactions performed in toluene.
Use of [Ir(acac)(cod)] as precur-
sor gave better performance in
acetonitrile (entries 11 vs. 13).
Table 1. Hydroformylation–isomerization of 1/1-int promoted by Rh- and Ir-biphephos systems.[a]
Entry Precursor
[M]
[1]0/[M] Solvent
1[b] 1-int[b] 9-/8- 2+3[b] 2/3 4[b] Conv. 1[c] HF[d]
[%] [%]
undec [%]
[%] [%]
[%]
1[e]
2
3
4
5[f]
6[f]
7
8
9
10
11
12
13
14
[Rh(acac)(CO)2] 20000 toluene
[IrCl(cod)]2 20000 toluene
[Rh(acac)(CO)2] 50000 toluene
[IrCl(cod)]2 50000 toluene
[Rh(acac)(CO)2] 20000 toluene
0
0
0
16
22
21
22
7
81/19 79
86/14 73
89/11 75
92/8 65
98/2 46
99/1
99/1
99/1
99/1
97/3
98/2
99/1
99/1
99/1
99/1
5
5
4
5
2
2
5
5
4
4
100
100
100
92
83
77
79
75
92
69
73
74
80
77
70
60
78
78
76
75
80
83
80
8
45
82
0
53
14
[IrCl(cod)]2
[IrCl(cod)]2
[IrCl(cod)]2
[IrCl(cod)]2
[IrCl(cod)]2
20000 toluene
20000 tetrahydrofuran
20000 N-methyl-2-pyrrolidone 25 18
20000 acetonitrile
50000 acetonitrile
7
26
95/5
9
88/12 69
95/5 52
92/8 75
93/7 69
92/8 64
96/4 30
88/12 74
90/10 62
88/12 76
93/7 73
84/16 75
94/6 72
92/8 72
100
74
1
5
3
20
22
21
99
95
97
53
[Ir(acac)(cod)] 20000 toluene
[Ir(acac)(cod)] 50000 toluene
[Ir(acac)(cod)] 20000 acetonitrile
99/1 12
99/1 12
45 13
22
16 18
0
99/1
99/1
99/1
99/1
99/1
99/1
99/1
4
4
5
4
4
4
5
100
83
100
98
[IrCpMe(cod)]
20000 toluene
20000 toluene
15[g] Pd(acac)2
16[h] [RuCl2(PPh3)3] 20000 toluene
18[i]
19[i]
20[i]
0
2
1
8
4
19
21
20
16
19
[Rh(acac)(CO)2] 20000 bulk
[Ir(acac)(cod)] 20000 bulk
[RuCl2(PPh3)3] 20000 bulk
99
92
96
[a] Reaction conditions unless otherwise stated: 1/1-int (95:5, 5.0 mmol), [biphephos]0/[M]=20, solvent (5 mL),
P=20 bar CO/H2 (1:1), T=1208C, t=5 h (Rh), 20 h (Ir), 72 h (Ru, Pd). [b] Distribution [mol%] of remaining 1 and
internal alkenes 1-int, aldehydes 2 and 3, and hydrogenation product 4, as determined by NMR and GLC analy-
ses. [c] Conversion of 1. [d] Selectivity for hydroformylation products (2+3) vs. isomerization and hydrogena-
tion. [e] Results from Ref. [21]. [f] With Xantphos as ligand. [g] t=72 h. [h] t=48 h. [i] A minimal amount
(0.5 mL) of toluene (Rh and Ru) or acetonitrile (Ir) was used to introduce the catalyst precursors.
Both the Ir- and Rh-biphephos
systems catalyzed the parallel
isomerization of 1 and, at total
conversion of the terminal olefin,
induced formation of internal
that the Ir biphephos system was less active than the Rh bi-
phephos system (hence, the batch reactions were typically
conducted over 5 h with Rh and 20 h with Ir), but not dramati-
cally so. Indeed, the Ir biphephos system showed a non-opti-
mized hydroformylation turnover frequency (TOFHF) towards
12-oxododecanenitrile (2) of 770 mol(2)mol(Ir)À1 hÀ1 (entry 2)
that was only approximately 5 times lower than the TOFHF of
3320 hÀ1 for the analogous Rh-based system (entry 1). As men-
tioned above, Beller et al. disclosed recently that an [Ir(cod)-
acac]PPh3 (in which acac=acetylacetonato, cod=1,5-cyclo-
octadiene) catalyst was no more than 8 times slower than a Rh
catalyst in the hydroformylation of various olefins.[9a] Yet, the
activity of Ir-based catalysts in the hydroformylation of 1 was
very much dependent on the nature of the ligand: an experi-
ment performed with Ir-Xantphos proceeded significantly
more slowly (TOFHF =97 hÀ1, entry 6) than that with the Rh-
isomers (1-int) to a similar extent (ꢀ15–25% yield). Remarka-
bly, based on the results reported in Table 1, both the Rh- and
Ir-biphephos systems controlled the distribution of the internal
isomers formed, and eventually only 9-undecenitrile (major
product) and 8-undecenitrile (minor product) were generated.
The distribution of internal isomers of undecenitrile (1-int) was
determined readily by NMR analysis of the crude reaction mix-
ture. The typical signals for the internal isomers were observed
1
in H NMR (CDCl3) spectra at d=5.20–5.40 ppm for olefinic hy-
drogens and the presence of a distinct CH3CH2 triplet signal at
d=1.05 ppm evidenced isomerization products in which the
double bond had migrated over at least two carbons from the
terminal positions, that is 8-undecenitrile, 7-undecenitrile, and
possibly other internal isomers (see Figure S5). Distinct alkene
signals were also evidenced in 13C{1H} NMR spectra, enabling
determination of the trans/cis configurations of the internal
isomers (Figure 2, see also Figure S6).[22] As seen in Table 1, the
Ir-biphephos system allowed for slightly but noticeably better
control of the distribution of the internal isomers than the Rh-
biphephos catalyst, with systematically higher ratios of 9-/8-un-
decenitrile, whatever the conditions used (entries 1,3 vs. 2,4).
Thus, the internal olefins formed during the hydroformylation
of 1 were typically comprised of 90–96% 9-undecenitrile with
Ir-biphephos and 80–90% with Rh-biphephos.
Xantphos system (TOFHF =1950 hÀ1
, entry 5). Notably, the
[Pd(acac)2]/biphephos and [RuCl2(PPh3)3]/biphephos combina-
tions also proved effective, although they proceeded more
slowly than with Ir- and Rh-biphephos catalysts under the
given conditions, with TOFHF values of 210 mol(2)mol(Pd)À1 hÀ1
(entry 15) and 310 mol(2)mol(Ru)À1 hÀ1 (entry 16; see also
Table S2 in the Supporting Information). For these reasons, fur-
ther studies were focused on Ir and Rh systems.
Remarkably, the Ir- and Rh-biphephos catalysts performed
similarly well in terms of chemo- and regioselectivity in hydro-
formylation. The linear aldehyde product 2 was obtained in
73% yield with [IrCl(cod)]2/biphephos (entry 2) and in 79%
yield with [Rh(acac)(CO)2]/biphephos (entry 1), with l/b ratios
[2/10-methyl-11-oxoundecanenitrile (3)] of 99:1 in both cases.
A series of experiments were performed to evaluate the im-
portance of the excess of ligand in the hydroformylation–isom-
erization reaction of 1, regarding in particular the distribution
of internal isomers 1-int. The results are summarized in Table 2.
With [Rh(acac)CO2], [IrCl(cod)]2, and [Ir(acac)(cod)] as precur-
sors, a slight but noticeable positive effect on the selectivity
ChemCatChem 0000, 00, 0 – 0
3
ꢁ 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
These are not the final page numbers! ÞÞ