A. Bellomo et al. / Tetrahedron Letters 51 (2010) 4934–4936
4935
Table 2
OH
HO
Allylation reactions of 4-Me cyclohexanones 1 in 1 M aqueous solution of fructopyr-
anosides 3 and 4
O
OH
OH
O
Entry
Fructopyranoside
Time (h)
Product: Yield (%)
2-ax/2-eq
3
1
2
3
4
3
0.8
2.0
1.5
24.0
2c: 76
2c:75
2c:75
2c:75
3:1
4a
4b
4c
1.56:1
2.33:1
1.77:1
Hydrophobic regions
R2
to indium, as previously reported.15 Finally, the multi-substituted
cyclohexanone (entry 7) demonstrated to be the less reactive
(26 h for total consumption) and led to poor diastereoselectivity,
mostly due to bulky substituents that prevent the attack of the
nucleophile.
: R2 = allyl
4a
O
4b: R2 = pentenyl
4c: R2 = propyl
OH
O
OH
HO
OH
Figure 1. Structure and hydrophobic areas of facial amphiphilic fructopyranosides
3 and 4.
We then performed the same reaction but in 1 M aqueous solu-
tion of fructopyranosides and arbitrarily using the 4-methyl cyclo-
hexanone (Table 2) as a cheap model. Facial amphiphilic sugars 3
and 4 were prepared following known procedures17–19 and their
properties as original additives were evaluated. Once again, yields
obtained were good (around 75%) and slightly decreased as com-
carbohydrate in water and also in a hydrophobic region (Fig. 1)
proved to be able to localize and orientate numerous organic sub-
strates in aqueous solutions. Faster reactions as well as better sol-
ubility of the reactants were also observed. Thus such original
media may constitute alternative solvents for the development of
sustainable chemistry, by preventing the utilization of complex,
expensive, and sometimes toxic organometallic catalysts.12 Exten-
sion to the allylation reaction was therefore attempted.
Based on these considerations, initial experiments were per-
formed using water as solvent by simply generating in situ13 the
allyl indium reagents (Scheme 1) with a mixture of In(0) and allyl
bromide.14 All reaction mixtures were stirred at room temperature
and their aspects rapidly turned into a milky suspension. Once all
the cyclohexanones were consumed, the resulting homoallylic
alcohols were purified by flash chromatography and characterized
by NMR. Compounds 2-ax and 2-eq were differentiated with the
help of NOE experiments showing correlations between the meth-
pared to those described in water (88%). The allyl-b-
entry 1) demonstrated the fastest reaction time and also the better
selectivity of 3/1, identical to the one observed in water. Its -enan-
L-sugar 3
D
tiomer, also commercially available but more abundant in nature
and therefore 60,000 times cheaper, exhibited a completely differ-
ent behavior (entries 2–4). Short 3-carbon aglycons even allowed
us to greatly modify the selectivity (entries 2 and 4), more than
by their pentenyl counterparts (entry 3). The 1-propyl fructopyr-
anoside 4c led to a slower reaction (entry 4). Thus, the allyl deriv-
ative 4a proved to show the largest effect under our conditions,
with a diastereoselectivity decreased to 1.56/1 in 2 h (entry 2).
The scope of the effect of the aqueous b-D-fructopyranoside 4a
solution in the In-catalyzed allylation reaction was finally studied
on the different cyclohexanones 1. As described in Table 3, reaction
times and yields remained unchanged and therefore comparable to
the ones obtained in water. However, as far as the stereoselectivity
is concerned, cyclohexanones 1 can be separated in two groups: (i)
the 3-substituted ones for those the proportions of 2-ax are slightly
increased (entries 1 and 2), and (ii) the others, that is, the 2- and
4-substituted ones, for those the contrary effects were observed
(entries 3–6). In addition, the bulky cyclohexanone (entry 7) was
totally unreactive in our conditions, even after three days of
reaction.
Such observations may result from different orientations of the
substrates in the aqueous media thanks to hydrophobic interac-
tions with the carbohydrate entity. This hypothesis was evaluated
by using 1H NMR titrations of 4-Me cyclohexanone in 1 M solution
of fructopyranoside 4a in D2O (see Supplementary data). No signif-
icant movement for signals corresponding to the sugar moiety
were observed upon the addition of the ketone. Still, most signals
were broadened and specially for those corresponding to the allylic
positions of 4a. This effect might be a consequence of very weak
ylene of the allyl group and the hydrogens at the a-position of the
alcoholic carbon (H-2) or with other substituents in the cyclohex-
ane ring. Ratio of 2-ax/2-eq was determined through the integra-
tion of the corresponding signals of the allylic protons in the 1H
NMR spectra. Results are summarized in Table 1. In an overall
manner, isolated yields were good to excellent (between 78% and
93%) and comparable to those previously observed in conventional
organic solvent. It is noteworthy that reactions were fast, even fas-
ter than those realized in an organic solvent.15,16 Unsurprisingly, as
predicted by Cram’s rule, the indium reagent showed a preference
for the formation of the axial alcohol via an equatorial attack, thus
minimizing the steric hindrance of the 3,5-axial substituents. This
effect was more pronounced for 3-substituted cyclohexanones (en-
tries 1 and 2) than for 4-substituted ones (entries 3 and 4) and in
direct correlation with the size of the group (entry 3 compared
to 4). As trifluoromethyl groups are generally considered as isoster-
ic replacements of methyl ones, some electronic factors might also
be implied (entries 1 and 2). For the 2-substituted derivatives (en-
tries 5 and 6), the proportion of the axial alcohol was even greater,
presumably as a result of favorable sulfur or oxygen coordination
Table 3
Allylation reactions of cyclohexanones 1 in 1 M aqueous b-D-fructopyranosides 4a
solutions
Table 1
Allylation reactions of cyclohexanones 1 in water
Entry
Cyclohexanone 1
Time (h)
Product: Yield (%)
2-ax/2-eq
(R=)
Entry
Cyclohexanone 1
(R@)
Time (h)
Product: Yield (%)
2-ax/2-eq
1
2
3
4
5
6
7
3-Me
3-CF3
4-Me
4-t-Bu
2-OMe
2-SMe
2,2-Me-3-(S)-OH
6.0
0.5
2.0
2.0
8.0
2a: 91
2b: 95
2c: 75
2d: 90
2e: 90
2f: 60
2g: 0
6.70:1
9:1
1.56:1
3.16:1
19:1
4:1
1
2
3
4
5
6
7
3-Me
3-CF3
4-Me
4-t-Bu
2-OMe
2-SMe
2,2-Me-3-(S)-OH
6.0
1.0
2.5
1.0
8.0
7.0
26
2a: 91
2b: 87
2c: 88
2d: 90
2e: 93
2f: 79
2g: 78
5.25:1
6.70:1
3:1
3.80:1
24:1
7.0
72.0
n.a.
6.14:1
1.56:1
n.a.: not applicable.