Highly diastereoselective catalytic Meerwein-Ponndorf-Verley reductions

Article abstract of DOI:10.1021/jo052121t

Very practical synthesis of ephedrine analogues in high yields and enantiopurity was realized by a highly diastereoselective Meerwein-Ponndorf- Verley (MPV) reduction of protected α-amino aromatic ketones using catalytic aluminum isopropoxide. The high anti selectivity resulted from the chelation of the nitrogen anion to the aluminum. In contrast, high syn selectivity was obtained with α-alkoxy ketones and other compounds via Felkin-Ahn control.

Full text of DOI:10.1021/jo052121t

reduction in general has not attracted much attention as a
stereoselective reduction method.4 In the few examples of MPV
reductions with reported high diastereoselectivities, stoichio-
metric amounts of aluminum alkoxides were used.5 Therefore,
MPV reduction has seen very limited applications for diaste-
reoselective reductions in modern synthetic organic chemistry,
and various boron hydrides and aluminum hydrides are usually
the preferred agents. To realize the full potential of the MPV
reduction, it is important to develop catalytic and highly
stereoselectiVe processes. Here, we report such a process for
the syntheses of ephedrine analogues and monoprotected 1,2-
diols.
Highly Diastereoselective Catalytic
Meerwein-Ponndorf-Verley Reductions
Jingjun Yin,* Mark A. Huffman, Karen M. Conrad, and
Joseph D. Armstrong, III
Department of Process Research, Merck Research Laboratories,
P.O. Box 2000, Rahway, New Jersey 07065
jingjun_yin@merck.com
Ephedrine and its analogues (4) are widely used as chiral
auxiliaries6 and chiral resolution agents7 and also have important
implications in weight control and treatment of obesity.8 As
shown in Scheme 1,9 4 could be obtained by direct LiAlH4
reduction of 3, whose precursor 2 could be readily prepared
from commercially available enantiopure carbamate-protected
alanine 1 (X ) OH) via activation as an acyl chloride or a
Weinreb amide [X ) Cl or NMe(OMe)].10 PhMe2SiH-TFA
was the most effective condition for the key diastereoselective
reduction of 2 [diastereomeric ratio (d.r.) > 98:2],9b,c but the
relatively high cost of PhMe2SiH and use of TFA as solvent
limited the application of this method on a large scale.
We were interested in developing a practical reduction of
the aminoketone 2-3 and started by screening the commonly
used reducing agents for this reaction (Table 1). The bistrif-
luoromethyl ketone 2a was a good substrate because of its high
reactivity and good compatibility of CF3 groups to various
reducing agents.11
ReceiVed October 10, 2005
Very practical synthesis of ephedrine analogues in high yields
and enantiopurity was realized by a highly diastereoselective
Meerwein-Ponndorf-Verley (MPV) reduction of protected
R-amino aromatic ketones using catalytic aluminum isopro-
poxide. The high anti selectivity resulted from the chelation
of the nitrogen anion to the aluminum. In contrast, high syn
selectivity was obtained with R-alkoxy ketones and other
compounds via Felkin-Ahn control.
The TFA-PhMe2SiH procedure9b,c did give a very high d.r.
(entry 1), but the use of a cheaper silane Et3SiH resulted in
slow and incomplete reactions. Among various borohydrides
(entries 2-10), only NaB(OAc)3H gave a reasonable d.r. of 4.9:1
(entry 9); however, this reaction was sluggish, and attempts to
further optimize it failed. Simple aluminum hydrides such as
LiAlH4 and LiAl(OtBu)3H gave moderate undesired selectivities
The Meerwein-Ponndorf-Verley (MPV) reduction of ke-
tones and aldehydes has been known for 80 years.1 It generally
i
uses inexpensive and environmentally friendly PrOH as a
hydride source and aluminum alkoxides as catalysts, and the
reaction is chemoselective, easy to operate, and readily scalable.2
These apparent practical advantages make the MPV reduction
a particularly attractive green-chemistry approach for reduction
of carbonyl compounds. However, aluminum-mediated MPV
reduction is limited by its requirement of a stoichiometric
amount of the aluminum alkoxide catalyst and sometimes low
yields due to side reactions.2 A number of modified aluminum
catalyst systems have been designed for catalytic MPV reduc-
tions,3 but they either give side reactions due to increased Lewis
acidity3a or require complex ligands3b-d or pyrophoric starting
materials.3e Additionally, although enantioselective aluminum-
catalyzed MPV reduction has met with moderate success,2c MPV
(4) The MPV reduction was generally not considered to give good
diastereoselectivities. See ref 2a.
(5) (a) Lund, H. Ber. 1937, 70B, 1520. (b) Jackman, L. M.; Macbeth,
A. K.; Mills, J. A. J. Chem. Soc. 1949, 2641. (c) Banthorpe, D. V.; Davies,
H. ff. S. J. Chem. Soc. B 1968, 1356. (d) Mueller, H. K.; Schuart, J.;
Baborowski, H. J. Prakt. Chem. 1973, 315, 1045. (e) Hach, V. J. Org.
Chem. 1973, 38, 293. (f) Kuroboshi, M.; Ishihara, T. Bull. Chem. Soc. Jpn.
1990, 63, 1185. (g) Hilpert, H. EP 703209, 1996. (h) Satoh, H.; Yamamoto,
K. EP 934923, 1999. (i) Matsuo, K.; Matsumoto, S.; Inoue, K. WO 9855452,
1998. (j) Brown, J. D.; Cain, R. O.; Kopach, M. E. EP 963972, 1999. (k)
Urban, F. J.; Jasys, V. J. Org. Process Res. DeV. 2004, 8, 169.
(6) Selected examples: (a) Rueck, K. Angew. Chem., Int. Ed. Engl. 1995,
34, 433. (b) Anakabe, E.; Badia, D.; Carrillo, L.; Rodriguez, M.; Vicario,
J. L. Trends Org. Chem. 2001, 9, 29.
(7) Enantiomers, Racemates, and Resolutions; Jacques, J., Collet, A.,
Wilen, S. H., Eds.; Krieger: Malabar, FL,1991.
(1) (a) Meerwein, H.; Schmidt, R. Justus Liebigs Ann. Chem. 1925, 444,
221. (b) Ponndorf, W. Angew. Chem. 1926, 39, 138. (c) Verley, M. Bull.
Soc. Chim. Fr. 1925, 37, 871.
(2) Reviews: (a) Wilds, A. L. Org. React. 1944, 2, 178. (b) de Graauw,
C. F.; Peters, J. A.; van Bekkum, H.; Huskens, J. Synthesis 1994, 1007. (c)
Nishide, K.; Node, M. Chirality 2002, 14, 759.
(3) (a) Kow, R.; Nygren, R.; Rathke, M. W. J. Org. Chem. 1977, 42,
826. (b) Ko, B.-T.; Wu, C.-C.; Lin, C.-C. Organometallics 2000, 19, 1864.
(c) Konishi, K.; Makita, K.; Aida, T.; Inoue, S. J. Chem. Soc., Chem.
Commun. 1988, 643. (d) Ooi, T.; Miura, T.; Maruoka, K. Angew Chem.,
Int. Ed. 1998, 37, 2347. (e) Campbell, E. J.; Zhou, H.; Nguyen, S. T. Org.
Lett. 2001, 3, 2391.
(8) Selected examples: (a) Clapham, J. C. Curr. Drug Targets 2004, 5,
309. (b) Dulloo, A. G. Int. J. Obes. 2002, 26, 590.
(9) (a) Buckley, T. F., III.; Rapoport, H. J. Am. Chem. Soc. 1981, 103,
6157. (b) Fujita, M.; Hiyama, T. J. Am. Chem. Soc. 1984, 106, 4629. (c)
Fujita, M.; Hiyama, T. J. Org. Chem. 1988, 53, 5415.
(10) (a) Skiles, J. W.; Fuchs, V.; Miao, C.; Sorcek, R.; Grozinger, K.
G.; Mauldin, S. C.; Vitous, J.; Mui, P. W.; Jacober, S.; Chow, G.; Matteo,
M.; Skoog, M.; Weldon, S. M.; Possanza, G.; Keirns, J.; Letts, G.; Rosentha,
A. S. J. Med. Chem. 1992, 35, 641. (b) Kano, S.; Yokomatsu, T.; Iwasawa,
H.; Shibuya, S. Tetrahedron Lett. 1987, 28, 6331.
(11) Unlike many other ketones in Table 2, the product isomers 3a/5a
are readily separated by LC for easy analysis during screening.
10.1021/jo052121t CCC: $33.50 © 2006 American Chemical Society
Published on Web 12/16/2005
840
J. Org. Chem. 2006, 71, 840-843

TABLE 2. Diastereoselective MPV Reduction of Protected
Aminoketonesa
SCHEME 1
TABLE 1. Reducing Agents Screening
temp time conv
entry
reducing agents
solvent
(°C)
(h)
(%) 3a/5a
1
2
3
4
5
6
7
8
9
TFA-PhMe2SiH
NaBH4
Bu4NBH4
Zn(BH4)2
Zn(BH4)2
LiBH4
TFA
MeOH
THF
THF
MeOH
THF
THF
THF
THF
-5
-78
0
15
1
2
3
1
2
16
2
72
72
1
99
100
100
100
100
100
33
97
93
60
95
64
0.71
1.3
2.3
0.63
1.3
2
0.71
4.9
1.9
0.55
0.45
-78a
-78
-5
9-BBN
rt
-78
20
LiB(sBu)3H
NaB(OAc)3H
10 NaB(OAc)3H-HOAc DMAc
20
11 LiAl(OtBu)3H
12 LiAlH4
THF
THF
toluene
tol-THF -78
toluene
toluene
toluene
toluene
toluene
toluene
toluene
-40
-78
-78
0.5 100
100
13 DIBAL
2
12
8.4
11
8.5
14 DIBAL
0.5 100
0.7 74
0.5 45b
16
0.2 95
15 Al(iBu)3
-78
0
rt
16 Et2AlCl
17 Et2Zn
18 DIBAL-ZnCl2
19 DIBAL-BHT
20 DIBAL-IPA (1:1)
21 DIBA--EtOH (1:1)
30b
-78
10
-78a 16
100
13
20
17
9.8
3.7
29
20 120
80
20
20
20 120
20 120
96
96
85c
90c
10
22 DIBAL-MeOH (1:1) toluene
23 DIBAL-tBuOH (1:1) toluene
24 DIBAL-menthol (1:2) toluene
75
48 120
25 20% DIBAL-
toluene
20
3
12 equiv IPA
a Reaction conditions: A mixture of ketone (0.637 mmol, 1.0 equiv),
Al(OiPr)3 (0.2-0.6 equiv), and iPrOH (0.536 mL, 11 equiv) in toluene (0.80
mL) was heated at 50 °C. b Based on crude 1H NMR or LC. c Isolated yields.
d 99% ee from 99% ee SM. e Combined yield of both isomers.
18
18
62 118
26 0.2 equiv Al(OiPr)3-
11 equiv IPA
toluene
toluene
20
50
22
91
76
27 0.2 equiv Al(OiPr)3-
11 equiv IPA
18
100
transfer,13 but the reaction still proceeded with a very high d.r.
of 29 (entry 24), indicating a highly diastereoselective R-hydride
transfer from an aluminum alkoxide, i.e., MPV reduction.
Indeed, with a combination of 20% DIBAL and 12 equiv of
IPA,14 the MPV reduction proceeded in an excellent d.r. (120:
1) to 48% conversion at room temperature in 3 h but slowed
significantly afterward (entry 26).15 Use of commercial Al(Oi-
Pr)3 resulted in a slower reaction (entry 26), but complete
conversion was achieved by simply raising the temperature to
50 °C (entry 27). It is also critical to discover that no
racemization was observed. Thus, highly diastereoselective
MPV reduction of carbamate-protected aminoketone 2 was
developed using a catalytic amount of the simple and practical
a Warmed to room temperature (rt) at the end. b Some Et adduct. c Not
clean.
(entries 11 and 12). However, DIBAL (diisobutyl-
aluminum hydride) in toluene gave a good selectivity of 12:1
(entry 13). The d.r. dropped when THF was used as a cosolvent
(entry 14). Interestingly, Al(iBu)3 alone gave a high d.r. of 11:1
(entry 15), indicating a hydride transfer from the alkyl group.
Other alkylmetal reagents for reduction were less efficient
(entries 16 and 17).
Modification of DIBAL by adding ZnCl2 resulted in a slightly
lower d.r. (entry 18). Addition of BHT to DIBAL,12 however,
gave a slightly higher d.r. (entry 19), and this prompted us to
screen other cheaper proton sources to modify DIBAL. Although
(13) N-H of the carbamate on 2a provides the third equiv of a proton
to fully quench DIBAL’s two isobutyl groups and one hydride.
(14) An excess of IPA was required, as the reaction using 2 equiv of
Al(OiPr)3 without IPA was very sluggish. It has been proposed that IPA
can help convert the tetramer of aluminum isopropoxide to the much more
reactive form (∼2.8 oligomer). Shiner, V. J.; Whittaker, D. J. Am. Chem.
Soc. 1969, 91, 394.
(15) The fast initial reaction rate suggests the higher activity of the freshly
prepared Al(OiPr)3, as observed in ref 3e. The slow rate afterwards was
comparable to the rate using commercial Al(OiPr)3, suggesting the deactiva-
tion of the catalyst to a less active form during the reaction.
i
the reaction became very slow, addition of 1 equiv of PrOH
(IPA) gave an excellent d.r. of 20:1 (entry 20). Other alcohols
were less efficient (entries 21-23). Addition of 2 equiv of
menthol to DIBAL would leave no isobutyl group for hydride
(12) Iguchi, S.; Nakai, H.; Hayashi, M.; Yamamoto, H. J. Org. Chem.
1979, 44, 1363. BHT ) 2,6-di-tert-butyl-4-methylphenol.
J. Org. Chem, Vol. 71, No. 2, 2006 841

SCHEME 2
SCHEME 3
SCHEME 4
TABLE 3. Diastereoselective MPV Reduction without Chelation
Controla
Al(OiPr)3, which offers significant advantages over the TFA-
silane procedure in terms of costs, potential functional group
compatibility, environmental aspects, scalability, and ease of
operation/isolation/waste control.
The MPV reduction proved to be applicable to various
N-protected R-aminoketones (Table 2). With Cbz as the
protecting group, very high diastereoselectivities (97.6:2.4 to
>99:1) and excellent yields of the anti products 3 were obtained
for all ephedrine-type substrates. No racemization was observed
for an electron-deficient substrate (entry 1) or an electron-rich
one (entry 5). NaBH4 reduction in MeOH at room temperature
was carried out for each ketone to obtain the syn isomer 5 for
1
selectivity measurements of the MPV reductions by H NMR
or liquid chromatography (LC) (entries 1 and 4), and the 3/5
ratio with NaBH4 ranged from 2.1:1 to 3.8:1. Aryl groups with
electron-withdrawing substituents (entries 1,7, and 8) or electron-
donating substituents (entries 3-5) and with steric hindrance
(entries 4 and 5) were tolerated. In most cases, 20-30% of the
aluminum catalyst was enough, but the ketone with a pyridyl
group required up to 60% of Al(OiPr)3 for >95% conversion
(entry 6), likely due to partial deactivation of the catalyst by
the pyridine ring.16 The R-chloromethyl group in place of an
aryl group was also tolerated to give product 3i (entry 9), which
could be readily converted to R-aminoepoxide 6, an advanced
building block for aminodiol HIV protease inhibitors (Scheme
2).17
When a mixture of 3d/5d in a ratio of 2.9:1 was spiked into
the MPV reduction of 2b using 0.40 equiv of Al(OiPr)3, 3d/5d
was recovered in about the same ratio after 24 h at 50 °C,
suggesting that the high anti selectivity is kinetically controlled.18
The observed anti selectivity agrees with the chelation control
(Scheme 3). This is also supported by the fact that NTos in 2j,
a more electron-deficient and weaker chelating group, gave a
lower selectivity of 82:18 (entry 10).
a Reaction conditions: a mixture of racemic ketone (2.55 mmol, 1.0
equiv), Al(OiPr)3 (0.1-0.3 equiv), and iPrOH (2.15 mL, 11 equiv) in toluene
(3.2 mL) was heated at 50 °C. b Based on crude 1H NMR or LC. c Isolated
combined yield of both isomers; NMR yield in parentheses.
instead of the chelation-controlled anti product 9, the Felkin-
Ahn-controlled syn product 8 was the main product. This is
the only reported highly syn-selective reduction of 7.19 NaBH4
gave a high ratio of the anti product 9.20
The MPV reductions of other R-alkoxy ketones also gave
very high levels of Felkin-Ahn control (Table 3, entries 2 and
3). Only 10-30% of the catalyst was required. The d.r. increased
with the size of the alkoxy group. The opposite selectivities
observed for 2-methoxy cyclohexanone (14) and 2-phenyl
cyclohexanone (16) indicated an equatorial transfer of the
hydride and the preference for an axial 2-methoxy group vs an
equatorial 2-phenyl group in the cyclohexanone.
The drastic difference in the chelation abilities of a NHCbz
group and a MeO group in the MPV reduction was puzzling at
first but can be explained by the deprotonation of the NH to
generate a nitrogen anion for better chelation; the same model
To determine if other substituents such as an alkoxy group
have the same chelation control, benzoin monomethyl ether was
subjected to the MPV reduction (Scheme 4). Surprisingly,
(16) It was found that addition of a catalytic amount of bipyridine resulted
in a slower MPV reduction.
(17) (a) Bisacchi, G. S.; Ahmad, S.; Alam, M.; Ashfaq, A.; Barrish, J.;
Cheng, P. T. W.; Greytok, J.; Hermsmier, M.; Lin, P. F.; Merchant, Z.;
Skoog, M.; Spergal, S.; Zahler, R. Bioorg. Med. Chem. Lett. 1995, 5, 459.
For a review, see: (b) Venkatesan, N.; Kim, B. H. Curr. Med. Chem. 2002,
9, 2243.
(19) A moderate d.r. (8/9 ∼2:1) has been reported using other methods:
(a) Burk, M. J.; Harper, T. G. P.; Lee, J. R.; Karberg, C. Tetrahedron Lett.
1994, 35, 4963. (b) Shibata, I.; Yoshida, T.; Kawakami, T.; Baba, A.;
Matsuda, H. J. Org. Chem. 1992, 57, 4049.
(20) Most other reducing agents also gave high anti/syn ratios: (a) Davis,
F. A.; Haque, M. S.; Przeslawski, R. M. J. Org. Chem. 1989, 54, 2021. (b)
Miyai, T.; Inoue, K.; Yasuda, M.; Shibata, I.; Baba, A. Tetrahedron Lett.
1998, 39, 1929. (c) Abernethy, C. D.; Cole, M. L.; Davies, A. J.; Jones, C.
Tetrahedron Lett. 2000, 41, 7567.
(18) We also observed about the same d.r. at 3 h (∼60% conv) as that
at 18 h for the MPV reduction of 2a.
842 J. Org. Chem., Vol. 71, No. 2, 2006

Without the chelation control, the Al(OiPr)3-iPrOH system
functions as a bulky hydride source to give very high syn/anti
selectivities, especially for R-alkoxy ketones. Therefore, it is
important to note that, for diastereoselective reductions of
carbonyls, the highly practical catalytic MPV reduction is an
excellent option in both chelation-controlled and Felkin-Ahn-
controlled settings.
SCHEME 5
Experimental Section
Representative Procedure for MPV Reduction. A mixture of
ketone 2a (267 mg, 0.637 mmol, 1.0 equiv), Al(OiPr)3 (26 mg, 0.2
i
equiv), and PrOH (0.536 mL, 11 equiv) in toluene (0.80 mL, 1.3
mL/mmol) was heated at 50 °C under N2 for 15 h. The reaction
was cooled and quenched with 4 mL of 1 N HCl and 4 mL of
EtOAc. The organic layer was washed with 4 mL of water and
was proposed in the highly anti-selective reduction of chloro-
methyl ketones using LiAl(OtBu)3H in EtOH at -78 °C.21 To
gain more insight, the hydrogen in 2b was replaced with an
ethyl group (18) and subjected to the MPV reduction (Scheme
5). Interestingly, after hydrolysis to remove the protecting group,
the anti/syn ratio was a reversed 30:70, disfavoring the chelation-
controlled anti product 19.22,23 MPV reduction of 21 also
disfavored the chelation-controlled product 22 with an anti/syn
ratio of 22:78. All these support the model of chelation with a
nitrogen anion, which forms by proton exchange between the
NHCbz group and the aluminum alkoxide.
1
concentrated. Crude H NMR was taken to determine the diaste-
reoselectivity to be 98.3:1.7. The product was further purified by
a hexane tituration to give 260 mg (97% yield) of (1R,2S)-1-(3′,5′-
bistrifluoromethylphenyl)-2-benzyloxycarbonylamino-1-propanol (3a)
as a white solid with >99% enantiomeric excess (ChiralPak AD-H
(4.6 × 150 mm) column, 0.1:10:90 of TFA/iPrOH/heptane, 1.5 mL/
min at 30 °C; retention times for desired 1R,2S enantiomers of 2.6
1
min and for undesired, 4.7 min). Mp 141-142 °C. H NMR (400
MHz, CDCl3): δ 7.82 (d, J ) 9.6 Hz, 3 H), 7.33-7.40 (m, 5 H),
5.14 (s, 2 H), 5.05 (br s, 1 H), 4.95 (br d, J ) 4.4 Hz, 1 H), 4.06
(br s, 1 H), 3.40 (br s, 1 H), 1.01 (d, J ) 6.9 Hz, 3 H). 13C NMR
(100 MHz, CDCl3): δ 156.7, 143.5, 136.0, 131.5 (q, J ) 33.3 Hz),
128.6, 128.4, 128.1, 126.5 (m), 123.3 (q, J ) 273 Hz), 121.5 (m),
75.3, 67.3, 52.5, 13.9. Anal. Calcd for C19H17F6NO3: C, 54.16; H,
4.07; N, 3.32. Found: C, 54.15; H, 3.84; N, 3.29.
In summary, the catalytic MPV reduction of Cbz-protected
R-amino aromatic ketones shows very high anti/syn selectivity
and provides an exceedingly practical preparation of ephedrine-
type products in high diastereo- and enantiopurity. The reaction
likely proceeds through chelation with deprotonated carbamate.
Characteristic 1H NMR signals for the minor isomer: 1.21 (d, J
) 6.9 Hz, 3 H).
Supporting Information Available: Experimental procedures
and characterization data for reduction products (Tables 2 and 3)
and compounds 18-23. This material is available free of charge
via the Internet at http://pubs.acs.org.
(21) Hoffman, R. C.; Maslouh, N.; Cervantes-Lee, F. J. Org. Chem. 2002,
67, 1045.
(22) In this case, NaBH4 gave an anti/syn selectivity of 94:6.
(23) Direct MPV reduction on 2-(ethylamino)propiophenone hydrochlo-
ride to give 19/20 was unsuccessful possibly because of the reaction between
the amino group and the ketone.
JO052121T
J. Org. Chem, Vol. 71, No. 2, 2006 843