Communication
Based on our experience in asymmetric ion pair and hydro-
Table 1. Optimization of the reactions conditions for the enantioselective
allylic alkylation of (E)-1a.
[
17]
gen-bond catalysis, we decided to examine the enantioselec-
tive synthesis of the desired chiral chromans employing
a chiral Brønsted acid as catalyst. However, in all the cases
tested, the compounds were obtained with poor enantiomeric
ratios. After these discouraging results, we attempted the
[a]
I
chiral counterion strategy using chiral Au–phosphate com-
[
18–20]
plexes as reaction promoters,
but in these cases enantio-
meric ratios remained low (up to 64.5:35.5). Finally, we decided
to employ chiral gold–phosphine complexes in the intramolec-
ular allylic alkylation reaction to generate the quaternary ste-
reogenic center at C-2 of the chroman core in an asymmetric
fashion. Indeed, gold catalysis has attracted widespread atten-
tion in organic synthesis over the past years, promoting
[
21]
a great number of enantioselective reactions. Specifically, al-
lylic alcohols have been employed as substrates for some
[
22]
[23,24]
enantioselective
or diastereoselective
alkylation reac-
tions, but the formation of a quaternary carbon center through
an allylic alkylation reaction still constitutes a major challenge.
Herein, we describe an intramolecular enantioselective allylic
alkylation reaction catalyzed by gold–phosphine complexes to
synthesize chiral vinyl chromans as precursors for vitamin E
and analogues.
[
b]
[c]
Entry
Ligand
Solvent
Yield [%]
e.r.
1
2
3
4
5
6
7
8
9
L1 (5%)
L2 (5%)
L3 (5%)
L4 (5%)
L5 (5%)
L6 (5%)
L7 (5%)
L2 (5%)
L2 (5%)
L2 (5%)
L2 (5%)
L2 (2.5%)
L2 (1.25%)
toluene/CH Cl (4:1)
2
2
2
2
2
2
2
2
2
2
90
96
94
89
93
96
91
95
92
84
94
96
98
70:30
86:14
85:15
79:21
57:43
toluene/CH
toluene/CH
toluene/CH
toluene/CH
toluene/CH Cl (4:1)
toluene/CH
toluene
Et O
2
TBME
CpME
toluene
toluene
Cl
Cl
Cl
Cl
(4:1)
(4:1)
(4:1)
(4:1)
[
25,26]
With allylic alcohol (E)-1a
as a model substrate, we car-
ried out a series of experiments to identify the best chiral
67.5:32.5
67:33
2
2
2
2
[
27]
Cl
(4:1)
phosphine ligand (Table 1).
For the first set of reactions,
92.5:7.5
91.5:8.5
91.5:8.5
92.5:7.5
93:7
a mixture of toluene/CH Cl (4:1) was used as solvent owing to
2
2
the low solubility of (E)-1a in pure toluene. Several ligands
with different electronic and steric properties were tested and
generally better results were obtained with the SEGPHOS
ligand family (L1–L4) (Table 1, entries 1–4), (R)-DM-SEGPHOS
10
11
[
[
d]
12
3
e,f]
1
92.5:7.5
(
L2) being the most promising one (Table 1, entry 2). Then, we
[a] Reaction conditions: (E)-1a (0.057 mmol), solvent (1 mL) at rt under Ar
atmosphere. [b] Yield of isolated product after flash chromatography.
focused on improving the enantioselectivity of the reaction by
modifying the solvent. Surprisingly, even if the substrate was
not initially fully dissolved, toluene gave better results (Table 1,
entry 8) affording the desired compound with 95% yield and
an e.r. of 92.5:7.5. Nevertheless, it should be pointed out that
good enantioselectivity could also be obtained if different
ether solvents were used (Table 1, entries 9–11). We also evalu-
ated different temperatures (lower or higher), but only a drop
in enantioselectivity was observed. The addition of different
silver salts and additives did not improve the results. Finally,
we found that the catalyst loading could be reduced to
[
c] Determined by chiral supercritical fluid chromatography (SFC) by
using a Chiralcel OD-H column. [d] 2.5 mol% of DMS·AuCl and AgOTf.
e] 1.25 mol% of DMS·AuCl and AgOTf. [f] (E)-1a (0.23 mmol), solvent
[
(4 mL) under Ar atmosphere. DMS=dimethylsulfide, TBME=tert-butylme-
thylether, CPME=cyclopentylmethylether.
as did less substituted compounds (Table 2, entries 6, 8–9). Ex-
cellent yields and slightly better levels of enantioselectivity
were obtained when the methyl group present in the sub-
strate was replaced with the longer ethyl group (Table 2, en-
tries 4 and 5). Finally, we decided to investigate the influence
of the configuration of the double bond of the substrate, and
observed that the opposite enantiomer of the chiral chroman
was formed with lower selectivity.
2.5 mol% without observing any influence on the yield or
enantiomeric ratio of chroman 2a (Table 1, entry 12).
Having established an optimal protocol for the reaction,
[28]
the scope of the methodology with regard to the allylic alco-
hol was studied (Table 2). The reaction proceeded well with
most of the alcohols, furnishing the corresponding vinyl chro-
mans 2a–i with excellent yields and good levels of enantiose-
lectivity.
With all these results in hand, we decided to apply the
methodology for the synthesis of biologically active and valua-
ble compounds, including vitamin E. The absolute configura-
tion of the chromans obtained by the newly developed asym-
metric gold-catalyzed allylic alkylation is the appropriate one
to synthesize the desired natural products.
Thus, compatibility of the process with a wide range of sub-
stituted aromatic rings is given. As specified in Table 2, the use
of substrates with different alkoxy groups (OMe, OEt, OBn) in
the para position of the phenol ring gave good results
Thus, we started to explore the preparation of a- and g-to-
copherol (4 and 5), the most active isoforms, by employing
a cross-metathesis between vinyl chroman 2 and olefin 3.
The lipophilic alkyl chain 3 was obtained in 49% overall
yield following a three-step synthetic approach, as described
(Table 2, entries 1–3). A bulkier compound led to the isolation
of the desired adduct 2g with similar results (Table 2, entry 7),
Chem. Eur. J. 2014, 20, 13913 – 13917
13914
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim