The cationic Rh/DPPF14 system was also tested, but the
expected cyclopentanone was not obtained (entry 4). Since
the highest activity was achieved with dppe as ligand, which
forms a five-membered ring upon coordination at the metal,
the reaction was next carried out using (R,R)-Me-Duphos.15
Satisfactorily, the intramolecular hydroacylation using (R,R)-
Me-Duphos, 5 mol % of catalyst, in acetone at reflux afforded
cyclopentanone 7 in good yield and excellent enantioselec-
tivity (entry 5). When (S,S)-Me-Duphos was employed, the
reaction proceeded yielding the opposite enantiomer ent-7
(entry 6). In both cases, the yield and the enantioselectivity
were identical, and the results were reproducible.
At this point, the stereoselective reduction of 7 was
essential to obtain the nucleoside 5 with the appropriate
configuration. Initially, the reduction of the cyclopentanones
7 and ent-7 was carried out with DIBAL-H in CH2Cl2 at
-78 °C affording the epimeric mixture of cyclopentanols
6a (trans) and 12 (cis), and ent-6a (trans) and ent-12 (cis),
respectively, in quantitative yield but low stereoselectivity
(cis/trans ) 2:1). We therefore considered two methodolo-
gies to obtain the alcohol derivative with the required
configuration: (a) a dynamic kinetic resolution (DKR) of the
diastereomeric mixture of alcohols16,17 and (b) a directed
reduction of the cyclopentanone using sodium triacetoxyboro-
hydride.
Table 2. Dynamic Kinetic Resolution of 6a (trans) + 12 (cis)a
entry time (h) convn (%)b 7 (%)b 13 (%)b 13 de (%)b,c dEd
1e
24
48
48
72
72
51
77
53
73
93
51
54
25
11
--
--
23
28
62
93
--
80
--
11
>50
>50
>150
2f
3g
4g
5g,h
>95i
>95i
>95i
a Reactions were performed on a 0.03 mmol scale with 15 mg of enzyme,
0.09 mmol of p-ClPhOAc, and 4 mol % of [Ru] in 0.5 mL of toluene at 70
b
°C. Determined by H NMR. c Diastereomeric excess. d Diastereomeric
ratio. e 1.5 mg of enzyme was added. f 0.015 mmol of 2,4-dimethyl-3-
pentanol was added. g H2 gas (1 atm) was used. h 6 mol % of Ru was added.
i The other diastereomer was not observed.
1
with excellent diastereoselectivities (>99%, entries 4 and
5). However, after 48 h, 25% of ketone 7 was still detected
(entry 3), and after 72 h, the ketone concentration was only
reduced to 11% (entry 4). To quench ketone formation, the
catalyst loading was increased to 6 mol %. Under these
conditions, the analysis of the reaction residue revealed the
absence of ketone, and the obtained conversion rate and
diastereomeric ratio were excellent (dE >150, entry 5).
Similar results were obtained starting from ent-6a, but in
this case the major isomer was acylated affording compound
14 (Scheme 3). These results are in agreement with Ka-
zlauskas’ rule.17
Initially, we optimized the conditions for the kinetic
resolution process. Candida antarctica (Novozyme 435,
N-435) and Pseudomonas cepacia (PSC) lipases were
screened with different acyl donors. The best conditions
found were employing PSC and p-chlorophenyl acetate18
in toluene at 70 °C which provided high conversions and
diastereoselectivities (dE ) 100).
On the basis of our preliminary KR results, the KR of
cyclopentanol 6a + 12 was carried out in the presence of
the Shvo’s catalyst20 to perform a DKR process. The results
are summarized in Table 2. The first run was performed using
6a + 12, enzyme PSC, Shvo’s catalysts, and p-ClPhOAc,
but no acetylated product was obtained (entry 1). However,
large amounts of the corresponding ketone 7 formed during
the hydrogen transfer process were observed (entry 1).
Hence, two hydrogen sources were tested to direct the
reaction equilibrium back toward cyclopentanols 6a + 12.
The addition of 0.015 mmol of 2,4-dimethyl-3-pentanol
improved the reaction providing a 23% conversion of the
acetylated product.
Scheme 3. Dynamic Kinetic Resolution of ent-6a + ent-12
The second approach involved the reduction of alde-
hydes and ketones using NaBH(OAc)3.21,22 For this
purpose, the tert-butyldiphenylsilyl group in 7 was depro-
tected by adding 1.5 equiv of TBAF in THF to afford 15 in
80% yield (Scheme 4). Compound 15 was diastereoselec-
tively reduced using this reagent in different solvents such
After 48 h, 77% of the reagents were converted into
products, and the diastereomeric ratio was dE ) 11 (entry
2); however, the ketone concentration was still high (54%).
To improve these results, H2 gas (1 bar) was added. H2
effectively inhibited ketone formation, and 13 was obtained
Scheme 4. Diastereoselective Synthesis of 16
(11) (a) Kundu, K.; McCullagh, J. V.; Morehead, A. T., Jr. J. Am. Chem.
Soc. 2005, 127, 16042. (b) Tanaka, M.; Imai, M.; Fujio, M.; Sakamoto, E.;
Takahashi, M.; Eto-Kato, Y.; Wu, X. M.; Funakoshi, K.; Sakai, K.;
Suemune, H. J. Org. Chem. 2000, 65, 5806.
(12) Craig, D.; Henry, G. D. Eur. J. Org. Chem. 2006, 3558.
(13) Aloise, A. D.; Layton, M. E.; Shair, M. D. J. Am. Chem. Soc. 2000,
122, 12610.
(14) (a) Stemmler, R. T.; Bolm, C. AdV. Synth. Catal. 2007, 349, 1185.
(b) Kokubo, K.; Matsumasa, K.; Nishinaka, Y.; Miura, M.; Nomura, M.
Bull. Chem. Soc. Jpn. 1999, 72, 303.
Org. Lett., Vol. 10, No. 21, 2008
4737