1694
F. Luo et al. / Tetrahedron Letters 51 (2010) 1693–1695
Table 2
S.cerevisiae
OH
OH
Bioreduction of 1g–j with dry baker’s yeasta
R
R
1a-1f
Yieldc (%)
eed (%)
½ ꢀ
a 2D5
2a-2f
Entry
1 (loading)
Product
1
2
3
1g (50 mg)
1h (50 mg)
1h (27 mg)
1h (27 mg)
1i (27 mg)
1j (27 mg)
2g
2h
2h
2h
2i
99
41
90
96
95
27
81
nd
87
88
82
nd
ꢁ14.5°
S.cerevisiae
OH
OH
nd
X
nd
F
F
2g-2i
X
4b
5b
6b
ꢁ10.9°
ꢁ43.6°
nd
1g-1i
2j
Scheme 3.
a
b
c
BY (2.0 g), pH 7.5 PBS (20 mL), at 30 °C, 48 h.
0.1 mL of n-butanol (5.5 ꢂ 10ꢁ5 mol lꢁ1) was added.
Determined by GC.
Table 1
Bioreduction of 1a–g with dry baker’s yeasta
d
Determined by chiral HPLC using Chiralpak AS-H column.
Entry
Substrate
Time (h)
Product
Yieldb (%)
1
2
3
4
5
6
7
1a
1b
1c
1d
1e
1f
24
24
24
24
24
48
24
2a
2b
2c
2d
2e
2f
99
57 (58)
20
9
0
5
was carried out under the conditions as noted in Table 2, and the
reaction time was prolonged to 48 h owing to the slow rate. The
ee values for products 2g–i were determined by chiral HPLC. As a
consequence, 1g was completely converted into 2g with 81% ee
(Table 2, entry 1). Under the same conditions, 1h with an elec-
tron-donating methoxy group was reduced slowly and only 41%
of 2h was generated after 48 h (Table 2, entry 2). However, the
yield of 2h was raised to 90% when the amount of 1h was reduced
from 50 mg to 27 mg (entry 3). In order to improve the solubility of
the solid substrate 1h during the reaction, a certain amount of
n-butanol was added as the co-solvent. In this case, 1h was almost
completely converted into 2h (entry 4). Similar results were also
observed in the case of p-fluorocinnamyl alcohol 1i (Table 2, entry
5). It should be emphasized that a marked difference in reactivity
was observed between 1i and p-chlorocinnamyl alcohol 1j. In the
latter case, only 27% of 2j was yielded after 48 h (Table 2, entry
6). As reported, the Hammett constant for p-Cl is 0.23, and that
for p-F is 0.06. These results indicated that the electronic property
of b-substituents of allyl alcohols also had a considerable effect on
their bioreduction. Obviously, the electron-withdrawing b-substi-
tuent would retard the biocatalytic dehydrogenation of allyl alco-
hols to some extent. Moreover, the configuration of the new
compounds 2h and 2i was assigned to be S by determining their
optical rotation degrees and by characterizing the structures of
their Mosher esters. As shown in Table 2, all the three products
2g, 2h, and 2i are levorotatory although their rotation degrees
are different. The analysis of 1H and 19F NMR spectra data of the
Mosher esters prepared from the optically active compounds 2g,
2h, and 2i and their corresponding racemates clearly indicated that
the configurations of 2h and 2i were the same as that of 2g.11
Next, the effect of pH value of the phosphate buffer solution was
studied. As shown in Table 3, the conversion rate for 1h was some-
what dependent on the pH value, and a relatively low conversion
was observed in the pH 6.5 phosphate buffer (entry 1). However,
the effect of pH value on the enantioselectivity was ignorable.
Another interesting result was observed by changing the amount
of baker’s yeast. When the amount was reduced from 2.0 g to
0.5 g, 1h could also be quantitatively converted into the product
2h, and the ee value increased from 88% to 92% (Table 3, entry
1g
2g
61 (99)
a
1 (50 mg), pH 7.5 PBS (20 mL), baker’s yeast (2.0 g) at 30 °C.
Yields determined by GC, and the data in parentheses for 48 h.
b
carefully, indicating that the saturated alcohol 2 was the main
product. The reactions are shown in Scheme 3.
The product yields were determined by GC and calculated based
on the amount of the substrate 1 added. All the results are summa-
rized in Table 1.
First the steric effect of a-substituents of cinnamyl alcohols was
investigated. As shown in Table 1, cinnamyl alcohol 1a was con-
sumed completely within 24 h, giving the product 2a almost quan-
titatively (entry 1). When the substrate 1b (R = CH3) was used, only
57% of 2b was detected due to the low conversion rate (Table 1, en-
try 2). With the increase of the size of R group, the conversion shar-
ply declined (Table 1, entries 3–5). For example, 1e (R = C4H9) was
too sterically hindered to be reduced within 24 h.
On the other hand, the electronic effect of
a-substituents was
also assessed by using 2-halocinnamyl alcohols 1f and 1g as the
substrates. Although the size of bromine atom was near to ethyl
group, the yield for 2f was only 5% (Table 1, entry 6), which was
much lower than that for 2c in the parallel experiments. Similarly,
when hydrogen atom was replaced by fluorine atom with near size,
only 61% of 1g was converted into the saturated alcohol 2g after
24 h (Table 1, entry 7). Moreover, a marked reactivity difference
was also observed between 1g and 1b. When the reaction was pro-
longed to 48 h, 1g was quantitatively converted into the saturated
alcohol 2g (Table 1, entry 7), but the yield for 2b was almost kept
constant (Table 1, entry2). These results strongly indicated that
dehydrogenation of allyl alcohols was not only very sensitive to
the steric hindrance of R group, but was also affected by its elec-
tronic effect to a certain content. The inhibitive role for halogen
atoms could be attributed to their high negative inductive effect
that decreased the electron density on C@C double bond of 1f
and 1g. In addition, the stereochemistry for this bioreduction of
1b or 2-methylcinnamaldehyde mediated by baker’s yeast had
been studied, and the (S)-isomer was preferentially afforded.9 In
fact, the formation of (S)-isomer was almost preferred in the biore-
duction of other allyl alcohols.10 Similarly, the absolute configura-
tion of 2g was determined to be S by comparing its optical rotation
with the literature data3b, whereas the configurations of 2c, 2d,
and 2f were not determined for their low conversions.
Table 3
Effect of pH and the amount of baker’s yeast on bioreduction of 1ha
Entry
pH
BY (g)
Yield (%) for 2h
ee (%)
1
2
3
4
5
6
6.5
7.0
7.5
7.0
7.0
7.0
2.0
2.0
2.0
1.0
0.5
0.3
75
96
96
96
96
91
86
88
88
91
92
90
Apart from the marked effect of a-substitutes, the role for sub-
stituents at b-site was also estimated. Thus the bioreduction of
several fluorinated cinnamyl alcohols 1g–j with different substitu-
ents on benzene ring was studied. The reduction reaction of 1g–j
a
1h (27 mg), PBS (20 mL), 30 °C, 48 h; 0.1 mL of n-butanol added. Yields deter-
mined by GC, and ee by chiral HPLC.
Luo, Fan
