Full Paper
ligand 16[10c] indicated that asymmetric hydrogenation follows
the generally accepted “unsaturated mechanism”[20] instead of
the “dihydride”[21] one.
Table 2. Optimization of the Rh-mediated asymmetric hydroformylation
of vinyl acetate with P–OP ligands 7a and 7d.[a]
Assuming that 1,1-P–OP ligands have the same reactivity
than their 1,2-analogues, the two front-right-hand octants
would be electronically disfavored for placement of the Ca and
Cb olefin carbon atoms of the alkene (double-bond coordina-
tion trans to the phosphino group is more favorable).[10c] Steric
hindrance in the transition state of the rate- and stereodeter-
mining step (oxidative addition in P–OP-mediated hydrogena-
tions)[10c] arises in the quadrant originally occupied by the Ca
olefin carbon.[20] Therefore, placement of this carbon atom in
the more accessible lower-left octant—far from the steric con-
gestion of the binaphthyl group—is strongly favored in hydro-
genations mediated by rhodium complexes of ligand 7a (see
the left side of Figure 5). The authors would like to emphasize
that these octant diagrams only represent a simplified view of
the hydrogenation process, or a mnemonic rule, that mainly
serves for predicting the sense of stereoinduction: hydrogena-
tion of a-(acylamino)acrylate (14a), dimethyl itaconate (14b),
the a-arylenamides 14c and 14e, and 1-phenylvinyl acetate
(14d) afforded R-, S-,[22] R- and R-configured products, respec-
tively (see Table 1) in high enantioselectivities.[23] On the con-
trary, in ligand 7b, placement of the Ca olefin carbon atom in
the front-upper-left octant—far from the steric congestion of
the binaphthyl group—is favored (see the right side of
Figure 5), which accounts for obtaining hydrogenation prod-
ucts with opposite configurations to those obtained with
ligand 7a (see Table 1). Additionally, enantioselectivities ob-
tained with the rhodium catalysts derived from 7b are consis-
tently lower (and in favor of the other enantiomer) with re-
spect to those observed for 7a (see Table 1 in the columns cor-
responding to ligands 7a and 7b). This effect reached its maxi-
mum when ligand 7b was used with substrate 14a (see
entry 1 in Table 1). Its substituent at the b-position (R=Ph)
sterically interacts with the phenyl substituent of the phos-
phine group in the front-lower-right octant (see the right side
of Figure 5), which could account for the enormous drop in
enantioselectivity obtained with the ligands presenting
matched and mismatched effects (i.e. 7a and 7b, respectively).
Entry Lig. L/[Rh] ratio S/C ratio T [8C] Conv. [%][b] b:l ratio[c] ee [%][c]
1
2
3
4
5[d]
6
7a 1.1
7a
7d 1.1
200
200
100
200
200
200
100
40
40
40
40
40
60
25
53
0
>99
>99
94
92:8
n.d.
29 (S)
n.d.
4
99:1
99:1
99:1
99:1
99:1
72 (R)
74 (R)
74 (R)
69 (R)
75 (R)
7d
7d
7d
4
4
4
>99
61
7
7d 1.1
[a] Hydroformylations were run in a parallel reactor under the specified
conditions. [b] Determined by 1H NMR spectroscopic analysis. [c] Deter-
mined by GC analysis on chiral stationary phases. Absolute configurations
were assigned by comparison of the elution orders in GC with reported
data, as indicated in the Supporting Information. [d] Incubation time:
23 h at 408C under CO/H2 (1:1, 10 bar).
Interestingly, incubation of the catalyst was not required,
which meant that future catalyst screening experiments would
be operationally easier (compare entries 4 and 5 in Table 2). Re-
garding temperature, the authors considered that 408C provid-
ed the optimal balance between enantioselectivity and conver-
sion (entries 5–7 in Table 2). Thus, based on the aforemen-
tioned results, the reaction conditions indicated in entries 1 or
4 in Table 2 were chosen as the optimal ones for future hydro-
formylation studies.
The authors then systematically tested the set of available
1,1-P–OP ligands in the hydroformylation of three benchmark
substrates: vinyl acetate (18a), styrene (18b), and (allyloxy)tri-
methylsilane (18c) (Table 3).[24] Concerning reactivity, for all
substrates except styrene (18b), the rhodium complexes de-
rived from ligands 7d–e (those of which contain a 3,3’-substi-
tuted-octahydrobinol-derived phosphite fragment) performed
better than did those derived from 7a–b or 10a–b.
Regarding the regio- and enantioselectivity, the best results
were achieved with ligand 7d (74% ee, see entry 1 in Table 3).
Moreover, the branched hydroformylation products obtained
with P–OP ligands 7d and 7e had opposite configurations.
This finding indicates that stereodifferentiation by the catalyst
is predominantly controlled by the phosphite fragment.
The authors also investigated the asymmetric hydroformyla-
tions of the challenging 1,2-disubstituted alkenes 18d and
18e (Table 3). Briefly, in terms of catalytic performance, the
rhodium complexes derived from the 1,1-P–OP ligands gener-
ally followed the same trend for substrates 18d–e as that pre-
viously observed for substrates 18a–c. Ligands 7d and 7e pro-
vided the best results in terms of activity and regio- and ste-
reoselectivity. For instance, hydroformylation of 18d with
ligand 7d yielded the 2-carbaldehyde 19d as the major isomer
with good conversion (83%) and high enantioselectivity (up to
80% ee; see entry 4 in Table 3), which is close to the highest
reported stereoselectivities in the literature (90% ee by Landis
Asymmetric hydroformylation
The authors also evaluated the catalytic properties of the pre-
pared set of 1,1-P–OP ligands in rhodium-mediated asymmetric
hydroformylations. A first round of experiments was complet-
ed, in which the aim was to identify optimal reaction condi-
tions using 1,1-P–OP ligands 7a or 7d, and vinyl acetate (18a)
as a model substrate. The reactions were run using catalysts
that had been preformed in situ by mixing [Rh(k2O,O’-
acac)(CO)2] with either ligand.
As reflected in Table 2, the ligand/[Rh] ratio affected the re-
action differently for each ligand: increasing it from 1.1 to 4
proved detrimental to catalytic activity for ligand 7a (compare
entries 1 and 2 in Table 2), yet did not affect catalytic activity
for ligand 7d (compare entries 3 and 4 in Table 2).
Chem. Eur. J. 2014, 20, 1 – 11
7
ꢂ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ