achieved for the benzaldehyde substrate with the ligands 2
is comparable to the results of the diethylzinc addition to
benzaldehyde with these ligands.11 Despite the use of excess
diethylzinc, no 1-phenylpropanol, the product of the diethyl-
zinc addition to benzaldehyde, was observed, which is
consistent with the findings in the literature.6,8 The conditions
for the best ligand (Rp,S)-2 were further optimized by varying
reaction temperature and catalyst loading (Table 1).
Decreasing the reaction temperature to -20 or -30 °C
did not change the enantioselectivity of the catalysis.
However, at -20 °C using 5 mol % of ligand (Rp,S)-2,
selectivity rose to 90% ee but with a diminished isolated
yield of 57% (Table 1, entry 5). Hence, all further catalyses
were carried out at -30 °C with 2 mol % of ligand as a
compromise between selectivity and yield.
systems. To the best of our knowledge, in the alkenylzinc
addition to aldehydes there has been no report of a highly
enantioselective addition to these substrates. However, these
aldehydes are excellent substrates for the paracyclophane
ligands, giving virtually complete enantioselection for cy-
clohexylcarbaldehyde and pivalaldehyde (entries 4 and 5).
Bulky alkynes were examined next, and at first glance they
seemed to limit the wide applicability of the paracyclophane
ligands. Thus, with the internal alkyne 3-hexyne only 75%
ee was obtained (entry 6) and the sterically even more
demanding tert-butyl ethyne lead to the desired product in a
disappointing 64% ee (entry 7).
At this point, the other ligands (Figure 1) were reexamined
with the sterically demanding tert-butyl ethyne, but no
improvement could be made (entries 8-10). Obviously, the
greater steric bulk of the zinc species resulting from tert-
butyl ethyne substantially diminishes the stereoselection of
the catalyst. In a final attempt to decrease the steric demand
of the zinc species, dimethylzinc (3 equiv) was employed
as the transmetalation reagent. This should result in the
formation of the somewhat less demanding zinc species 5
(R′′ ) methyl in Scheme 2). Gratifyingly, this brought the
The scope of the asymmetric addition of alkenylzinc
reagents to aldehydes with 2 mol % of ligand (Rp,S)-2 is
illustrated in Table 2. Electron-withdrawing substituents on
Table 2. Alkenylzinc Addition to Various Aldehydes to Give
the Allyl Alcohols 7 According to Scheme 1a
entry
alkyne 3
aldehyde 4, R ) yield (%) ee (%)b
1
2
3
4
5
6
7
8
9
10
11
12
13
1-octyne
phenyl
71
88
62
80
89
86
78
90
88
83
88
84
48
86 (S)
97 (S)
91 (S)
>98 (R)c
>98 (R)d
75 (S)
Scheme 2. Equilibrium of Zinc Species
1-octyne
1-octyne
1-octyne
1-octyne
3-hexyne
tert-butyl ethyne 4-Cl-phenyl
tert-butyl ethyne 4-Cl-phenyl
tert-butyl ethyne 4-Cl-phenyl
tert-butyl ethyne 4-Cl-phenyl
tert-butyl ethyne 4-Cl-phenyl
4-Cl-phenyl
4-MeO-phenyl
cyclohexyl
tert-butyl
phenyl
64 (S)
62 (R)e
67 (R)f
54 (S)g
89 (S)h
88 (S)g
76 (S)g
3-hexyne
1-octyne
phenyl
phenyl
breakthrough for bulky alkynes. For tert-butyl ethyne now
89% ee was achieved with ligand (Rp,S)-2 (entry 11), and
also for the symmetrical internal alkyne 3-hexyne 88% ee
could be obtained under the changed reaction conditions
(entry 12).
However, the use of dimethylzinc was only beneficial for
the aforementioned bulky alkynes. Its application in the
reaction of the hydroborated 1-octyne with benzaldehyde lead
to diminished enantioselectivity and yield (76% ee and 48%
isolated yield, entry 13) as compared to the already men-
tioned results with diethylzinc (Table 1 and Table 2, entry
1).
a In the presence of 2 mol % of chiral ligand (Rp,S)-2 (unless otherwise
stated) at -30 °C. b Determined by HPLC (Chiracel OD column, entries
1-3, 13; Chiracel AD column entries 7-11; (S,S)-Welk-O 1 column entries
6, 12). See ref 12. c Determined by GC (Chirasil-dex). The other enantiomer
was not observed. d Determined by NMR of the camphanic ester derivative.
The other enantiomer was not observed. e (Sp,S)-1 was used as ligand.
f (Sp,S)-2 was used as ligand. g (Rp,S)-1 was used as ligand. h Dimethylzinc
was used instead of diethylzinc.
aromatic aldehydes are well tolerated and lead to increased
enantioselecivity of 97% ee for p-chlorobenzaldehyde (entry
2). The electron-rich p-methoxybenzaldehyde, which is a
difficult substrate for many ligands in the diethylzinc1 as well
as alkenylzinc8c addition, also provides a very good ee of
91% although with a diminished yield of 62% (entry 3).
Aliphatic and especially R-branched aliphatic aldehydes
belong to the most problematic substrates for nearly all ligand
We rationalize these results with the slow equilibration
of zinc species at temperatures below 0 °C. Oppolzer et al.
1
already mentioned H NMR studies of the species formed
by reaction of hydroborated alkynes with dimethylzinc.6a
Although the transmetalation proceeds quickly at -65 °C
to give two equilibrating alkenylzinc species (which they
assigned as monomeric and dimeric species), no equilibrium
between mixed (5 in Scheme 2) and symmetrical species
(5′ in Scheme 2) was observed up to 0 °C, where the zinc
species begin to decompose. Thus, the active zinc species 5
can be tuned by the choice of the transmetalation reagent.
(11) Dahmen, S.; Bra¨se, S. Chem. Commun. 2001, in press.
(12) The absolute configuration was assigned by comparison of the
optical rotation with the literature known compounds (S)-1-(4-chloro-
phenyl)-hept-2-en-1-ol [ref c, Supporting Information] and (S)-1-phenyl-
non-2-en-1-ol [ref a], respectively, and the assumption of a unanimous
reaction pathway for all other aldehyde substrates. The absolute configu-
ration of the allyl alcohol products 7 is consistent with the induction observed
in the diethylzinc addition to aldehydes with the ligands 1 and 2.
Org. Lett., Vol. 3, No. 25, 2001
4121