Increased Felkin-Anh selectivity in nucleophilic additions to α-chiral aldehydes using vinylalanes

Article abstract of DOI:10.1016/S0040-4039(02)00753-0

Vinylalanes, prepared from the zirconium-catalyzed carboalumination of alkynes, were added directly to aldehydes bearing an alpha chiral carbon to give good yields of the corresponding alcohols. The addition was more stereoselective than the addition of t

Full text of DOI:10.1016/S0040-4039(02)00753-0

TETRAHEDRON
LETTERS
Pergamon
Tetrahedron Letters 43 (2002) 4183–4185
Increased Felkin–Anh selectivity in nucleophilic additions to
a-chiral aldehydes using vinylalanes
Claude Spino,* Marie-Claude Granger, Luc Boisvert and Christian Beaulieu
Universite´ de Sherbrooke, De´partement de Chimie, Sherbrooke, Que´bec, Canada J1K 2R1
Received 15 March 2002; accepted 17 April 2002
Abstract—Vinylalanes, prepared from the zirconium-catalyzed carboalumination of alkynes, were added directly to aldehydes
bearing an alpha chiral carbon to give good yields of the corresponding alcohols. The addition was more stereoselective than the
addition of the corresponding vinyllithium or vinylmagnesium bromide. © 2002 Published by Elsevier Science Ltd.
The stereoselectivity of nucleophilic addition to chiral
aldehydes and ketones was rationalized thanks to con-
tributions by Felkin, Anh, and Eisenstein many years
ago.1,2 Although the Felkin–Anh rule has proven a
highly reliable tool for predicting the stereochemical
outcome of the irreversible addition of nucleophiles to
aldehydes and ketones, the selectivities are generally
modest for Grignard-type nucleophiles.3 This is espe-
cially true for aldehydes bearing two alpha n-alkyl
groups.
added to aldehyde 4 (Fig. 1) with superior diastereose-
lectivity to 8b and 8c (entry 1). Wipf and co-workers
reported that the vinylzinc corresponding to 2a (Table
1, AlMe2=ZnMe) added to aldehyde 4 to give a 5.6:1
ratio of alcohols.10 In the case of 8a, when
dichloromethane was replaced by tetrahydrofuran prior
to the addition of aldehyde 4, a 16:1 ratio of alcohols
was obtained. This suggests that the solvent is not a
major factor in the selectivity of addition of vinylalanes
to 4. Aldehyde 5 was converted to the corresponding
alcohol by vinylalane 8a with good stereoselectivity
(entry 3). The 2:1 ratio of diastereomers of the starting
aldehyde 5 was unchanged in the product. While the
Recently, we disclosed our findings that vinylalanes
added to menthylcarboxaldehyde 1 with high stereose-
lectivity, while the corresponding or similar vinyllith-
iums or vinylmagnesium bromides added with
selectivities of 2–3:1 (Table 1).4 The good chemical
yields obtained in theses additions were also a pleasant
surprise. A prior report by Newman5 indicated that the
direct addition of vinylalanes afforded alcohols in poor
yield. Perhaps this explains why vinylalanes are gener-
ally transformed to the corresponding vinylhalide first
and metalated again before addition to aldehydes or
ketones. As far as we know, the stereoselectivity of the
addition of vinylalanes and trialkylalanes onto chiral
aldehydes has not been probed in detail. We are aware
of only two reports of the stereoselective addition of
trivinylalane to Garner’s aldehyde giving a modest ratio
of 2:1.6,7
Table 1. Ratio and yield of 3 generated in the addition of
vinylalanes 2 to chiral aldehyde 1
Me
AlMe2
i-Pr
i-Pr
R
R
2
CHO
CH2Cl2, 0-20 °C
Me
Me
OH Me
1
3
Entry
2
R
3 (b:a)a
Yield (%)b
1
2
3
4
5
6
a
b
c
d
e
f
n-Bu
12:1
8:1
10:1
14:1
18:1
11:1
70
68
85
80
63
76
TBSO(CH2)3
HOCH2
c-C6H11
Ph
Curious to see if vinylalanes added selectively to other
chiral aldehydes,8 we initiated a comparative study of
their addition to various aldehydes with respect to
vinyllithium and magnesium (Table 2).9 Vinylalane 8a
PhCH2
a Ratios determined by GC.
b Isolated yield of pure mixtures of isomers. All isomers were separa-
ble by normal silica gel column chromatography.
* Corresponding author.
0040-4039/02/$ - see front matter © 2002 Published by Elsevier Science Ltd.
PII: S0040-4039(02)00753-0

4184
C. Spino et al. / Tetrahedron Letters 43 (2002) 4183–4185
Table 2. Ratio and yield of allylic alcohols 9 generated in
the addition of vinylmetals 8 to chiral aldehydes11
Table 3. Ratio and yield of allylic alcohols generated in
the addition of alkylmetals to chiral aldehyde 2
Me
M
Entry
Reagent
Solvent
Ratio
Yielda
R
R'
R'
R
CHO
1
2
3
4
AlMe3
AlMe3
MeMgBr
MeLi
CH2Cl2
THF
THF
2.6:1
–
3.8:1
10.6:1
87
0
52
47
8a M=AlMe2, R' = n-C5H11
8b M=MgBr, R' = Me
8c M=Li, R' = n-C5H11
OH Me
9
THF
Entry
R-CHO
9 from 8a
9 from 8b
9 from 8c
a % isolated yield.
1
2
3
4
5
6
7
1
4b
5c
6a
6b
7a
7b
15:1 (83)a
20:1 (85)
12:1 (65)
3:1 (87)
1:1 (84)
2:1 (46)
2:1 (62)
4:1 (60)
3:1 (45)
1.5:1 (46)
1:1 (58)
2:1 (12)
4:1 (47)
2:1 (62)
9:1 (55)
2.5:1 (34)
1.5:1 (56)
1:1 (57)
1:2 (11)
3:1 (49)
addition of vinyl Grignard reagents 8b and 8c, respec-
tively (Table 2, entry 2). This observation is puzzling
and suggests that the rational for the stereoselective
addition of vinylalanes may be more complex than first
anticipated. We initially believed that the size of the
metal in vinylalanes as compared to the corresponding
lithium or magnesium derivatives was the main factor
for the stereoselectivity of addition. Indeed, this argu-
ment has been used many times to explain the differ-
ences in Felkin–Anh selectivity between various
alkylmetals.1 However, our result with trimethylalu-
minum seems to contradict that hypothesis. Performing
the addition of trimethylaluminum in tetrahydrofuran
killed the reaction entirely (entry 2, Table 3).
2.5:1 (82)
a All ratios determined by GC and % isolated yield of pure 9 as
mixtures of isomers in parentheses. Isomers from aldehydes 1, 5 and
7a were separable by chromatography, all others were not.
b A ratio of 5.6:1 was obtained when a vinylzinc was added to 4; see
Ref. 10.
c The 2:1 ratio of diastereomers in the starting aldehyde 5 was
unchanged in product.
selectivity of addition of vinylalane 8a to aldehyde 6a,
bearing a n-alkyl chain, proceeded with modest selectiv-
ity (3:1), the same addition using vinyllithium or vinyl-
magnesium bromide gave a nearly equal mixture of
both diastereomers (entry 4). A b-chiral center had no
influence on the selectivity as shown by the non-
stereoselective addition of all vinylmetals to aldehyde
6b (entry 5).
In conclusion, we have shown that vinylalanes derived
from the zirconium-catalyzed carboalumination of
alkynes add stereoselectively to a-chiral aldehydes.
Other vinylmetals give lower selectivities and often
lower yields as well. Accessibility to a wide variety of
vinylalanes remains an issue to be solved. Trialkyl-
alanes are less selective, a result that poses a very
interesting mechanistic question. Investigations are con-
tinuing in our laboratories.
When a chelating group was present on the alpha
carbon, as in aldehydes 7a or 7b, the stereoselectivity
dropped, a result consistent with the findings of Garner
and Coleman (entries 6 and 7).4,5 Interestingly, the
sense of induction for aluminum and lithium was oppo-
site that of magnesium, though the differences were
small. The major isomer in the case of the vinylmagne-
sium halide addition corresponded to a Cram-chelate
addition while the other two vinylmetals furnished
Felkin–Anh products.
Acknowledgements
We thank the Merck-Frosst Centre for Therapeutic
Research (Canada), the Natural Sciences and Engineer-
ing Research Council of Canada, and the ACS
Petroleum Research Fund for financial support.
References
We also investigated the selectivity of addition of alkyl-
metals to aldehyde 4. Surprisingly, trimethylaluminum
added to aldehyde 4 with low stereoselectivity (Table 3,
entry 1) compared to vinylalane 8a (Table 2, entry 2).
This was not the case with methyllithium and methyl-
magnesium bromide (Table 3, entries 3 and 4), which
added to aldehyde 4 with selectivities similar to the
1. Che´rest, M.; Felkin, H.; Prudent, N. Tetrahedron Lett.
1968, 18, 2199–2204.
2. (a) Anh, N. T. Top. Curr. Chem. 1980, 88, 145–162; (b)
Anh, N. T.; Eisenstein, O. Nouv. J. Chim. 1977, 1, 61–70;
(c) Anh, N. T.; Eisenstein, O. Tetrahedron Lett. 1976, 26,
155–158.
Ph
CHO
CHO
CHO
Me
Z
Me
Me
OR
7a R=Me
7b R=TBDPS
Me
Me
Me
6a Z = CHO
4
5
6b Z = CH2CHO
Figure 1. Aldehydes used in the additions listed in Table 2.

C. Spino et al. / Tetrahedron Letters 43 (2002) 4183–4185
4185
10. Wipf, P.; Xu, W. Tetrahedron Lett. 1994, 35, 5197–5200.
11. Major isomer of 9 from 4: H NMR (300 MHz, CDCl3):
3. Mengel, A.; Reiser, O. Chem. Rev. 1999, 99, 1191–1223.
4. Spino, C.; Beaulieu, C. Angew. Chem., Int. Ed. Engl.
2000, 35, 1930–1932.
5. Newman, H. Tetrahedron Lett. 1971, 12, 4571–4572.
6. Garner, P.; Park, J. M.; Malecki, E. J. Org. Chem. 1988,
53, 4395–4398.
1
l 7.36–7.17 (m, 5H), 5.08 (dd, 1H, J=8.8, 0.6 Hz), 4.44
(dd, 1H, J=8.9, 6.2 Hz), 2.88 (qi, 1H, J=6.9 Hz), 1.91 (t,
2H, J=7.5 Hz), 1.53 (d, 3H, J=0.8 Hz), 1.41 (s, 1H),
1.34 (d, 3H, J=7.1 Hz), 1.33–1.07 (m, 6H), 0.87 (t, 3H,
J=7.2 Hz); IR (neat, cm−1): 3376; LRMS (m/z): 228
(M−H2O, 10), 86 (90), 84 (100); exact mass calcd for
C17H24 (M+−H2O): 228.1878, found: 228.1874. Major
7. Coleman, R. S.; Carpenter, A. J. Tetrahedron Lett. 1992,
33, 1697–1700.
8. Typical experimental procedure: Dichlorobiscyclopenta-
dienyl zirconium (0.558 g, 1.91 mmol) was dissolved in
dichloromethane (25 mL). Neat trimethylaluminum (0.22
mL, 2.29 mmol) was added dropwise and the yellowish
solution was stirred 10 min before cooling to 0°C. n-Hep-
tyne (0.10 mL, 0.76 mmol) was then added and the
mixture stirred for 24 h. The resulting solution containing
the vinylalane was cooled to −78°C and 2-phenylpropi-
onaldehyde 4 (78 mL, 0.59 mmol) in THF was added. The
reaction was monitored by TLC and warmed to 0°C
before a saturated aqueous potassium carbonate solution
was added slowly until gas evolution ceased. The white
suspension was taken up in 1N HCl and the mixture
vigorously stirred for 30 min. A typical work-up was then
performed, extracting with dichloromethane. Final purifi-
cation was done by silica gel flash column chromatogra-
phy. The ratios of products were determined by GC on
the crude product after identification of each GC peak
was ascertain by GC mass spec. on the purified mixture.
Ratios were corroborated by proton NMR. Isomers of 9
obtained from aldehydes 1, 5, and 7a were separable by
chromatography on silica gel.
1
isomer 9 from 5: H NMR (300 MHz, CDCl3): l 5.38 (s,
1H), 5.18 (d, 1H, J=8.6 Hz), 4.30 (t, 1H, J=8.5 Hz),
2.03–1.88 (m, 5H), 1.80–1.69 (m, 3H), 1.67 (d, 3H, J=0.8
Hz), 1.64 (s, 3H), 1.60–1.55 (m, 3H), 1.46–1.21 (m, 6H),
0.91–0.86 (m, 3H), 0.75 (d, 3H, J=7.0 Hz); 13C NMR
(CDCl3, 75 MHz): 139.6 (d), 133.9 (s), 126.2 (d), 120.8
(s), 70.0 (d), 43.6 (d), 39.7 (d), 33.9 (t), 31.5 (t), 30.9 (t),
28.1 (t), 27.4 (t), 26.4 (t), 23.4 (t), 22.5 (q), 16.6(q), 14.0
(q), 10.4 (q); IR (pur, cm−1): 3389; LRMS (m/z): 264
(M+, 5), 246 [(M+−H2O), 20], 152 (100), 141 (95), 95 (80),
94 (90); exact mass calcd for C18H32O: 264.2453, found:
1
264.2457. 9 from 6a H NMR (300 MHz, CDCl3): l 5.19
(d,1H, J=8.9 Hz), 5.13–5.06 (m, 1H), 4.20–4.14 (m, 1H),
2.08–1.90 (m, 4H), 1.67 (d, 3H, J=2.7 Hz), 1.66 (d, 3H,
J=0.8 Hz), 1.67–1.01 (m, 10H), 0.94 (d, 3H, J=6.6 Hz),
0.95–0.85 (m, 6H); IR (pur, cm−1): 3353; LRMS (m/z):
252 (M+), 141 (100); exact mass calcd for C17H32O:
252.2453, found: 252.2458. Major isomer of 9 from 7a:
1H NMR (300 MHz, CDCl3): l 5.13 (d, 1H, J=9.3 Hz),
4.25 (t, 1H, J=8.4 Hz), 3.44 (s, 3H), 3.08–3.02 (m, 1H),
2.02 (t, 2H, J=7.5 Hz), 1.70 (s, 3H), 1.68–1.10 (m, 13H),
0.93–0.86 (m, 6H); IR (neat, cm−1): 3439; LRMS (m/z):
225 (M−OH, 20), 141 (100), 101 (95); exact mass. calcd
for C15H29O: 225.2218, found: 225.2213.
1
9. All new compounds gave satisfactory H NMR, IR, and
mass spec. data.