Catalyst-controlled diastereoselection in the hydrogenation of heterocycloalkyl ketones

Article abstract of DOI:10.1002/adsc.201100398

α-Substituted chiral ketones that have small steric and electronic differences around the reaction sites are difficult substrates to reduce with high diastereoselectivity. Metal hydride reduction of 2-(4-benzoylmorpholinyl) phenyl ketone and 3-(1-tert-butoxycarbonylpiperidinyl) phenyl ketone using sodium borohydride, zinc borohydride, and potassium tri-sec-butylborohydride as reducing agents affords the syn- and anti-alcohols in a lower than 80:20 ratio. Hydrogenation of these ketones with a catalyst system of RuCl ₂(BIPHEP)(DMEN) and potassium tert-butoxide in 2-propanol results in the syn-alcohols with ≥ 99:1 selectivity [BIPHEP=2,2′- bis(diphenylphosphino)biphenyl, DMEN=N,N-dimethylethylenediamine]. The marked difference in the diastereoselectivity suggests that the stereoselection in this hydrogenation is primarily regulated by the structure of the catalyst's reaction field ("catalyst-controlled diastereoselection") but not the internal stereocontrol of the substrates. This chemistry is applied to the asymmetric hydrogenation through dynamic kinetic resolution with a RuCl ₂[(S)-BINAP][(R)-DMAPEN]/potassium tert-butoxide catalyst [BINAP=2,2′-bis(diphenylphosphino)-1,1′-binaphthyl, DMAPEN=2-dimethylamino-1-phenylethylamine]. A series of aryl heterocycloalkyl ketones has been converted to the alcohols in excellent diastereo- and enantioselectivities. The modes of catalyst-controlled diastereoselection and enantioselection are interpreted by using transition-state molecular models. (S,S)-Reboxetine, a selective norepinephrine uptake inhibitor, was synthesized from one of product alcohols. Copyright

Full text of DOI:10.1002/adsc.201100398

COMMUNICATIONS
DOI: 10.1002/adsc.201100398
Catalyst-Controlled Diastereoselection in the Hydrogenation of
Heterocycloalkyl Ketones
Masaya Akashi,b Noriyoshi Arai,a Tsutomu Inoue,b and Takeshi Ohkumaa,*
a
Division of Chemical Process Engineering, Faculty of Engineering, Hokkaido University, Sapporo, Hokkaido 060-8628,
Japan
Fax : (+81)-11-706-6598; e-mail: ohkuma@eng.hokudai.ac.jp
Odawara Research Center, Nippon Soda Co., Ltd., 345 Takada, Odawara 250-0280, Japan
b
Received: May 17, 2011; Published online: August 10, 2011
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/adcs.201100023.
Abstract: a-Substituted chiral ketones that have
small steric and electronic differences around the
reaction sites are difficult substrates to reduce with
high diastereoselectivity. Metal hydride reduction of
2-(4-benzoylmorpholinyl) phenyl ketone and 3-(1-
tert-butoxycarbonylpiperidinyl) phenyl ketone using
sodium borohydride, zinc borohydride, and potassi-
um tri-sec-butylborohydride as reducing agents af-
fords the syn- and anti-alcohols in a lower than
80:20 ratio. Hydrogenation of these ketones with a
Diastereoselective reduction of a-substituted chiral
ketones to the 1,2-syn- or 1,2-anti-alcohols is a funda-
mental and indispensable reaction in modern organic
synthesis.[1] Metal hydride reducing agents have made
notable contributions to advance this important class
of chemistry. The stereoselective outcome has been
explained by using transition-state molecular models,
such as the Felkin–Anh model and the chelation-con-
trolled model (Figure 1).[1–3] In both cases the metal
hydride reagent is recognized as just a hydride, and it
reacts within the region of the substrate molecules.
Therefore, the intramolecular stereoinduction princi-
pally depends on the substrate structure (substrate
control), but not the shape of the metal hydrides (re-
agent control).[4] High diastereoselectivity in this reac-
tion is achieved only when the substrates have a-sub-
stituents (L, M, S, and X) with marked differences in
size and/or electronic properties. Thus, we thought
that a new conceptual approach is required to devel-
op a general method for the diastereoselective reduc-
tion of a-substituted ketones.
catalyst system of RuCl2ACTHNUGRTNENUG(BIPHEP)ACHUTNGTRENN(GUN DMEN) and
potassium tert-butoxide in 2-propanol results in the
syn-alcohols with ꢀ 99:1 selectivity [BIPHEP=2,2’-
bis(diphenylphosphino)biphenyl, DMEN=N,N-di-
methylethylenediamine]. The marked difference in
the diastereoselectivity suggests that the stereose-
lection in this hydrogenation is primarily regulated
by the structure of the catalystꢀs reaction field
(“catalyst-controlled diastereoselection”) but not
the internal stereocontrol of the substrates. This
chemistry is applied to the asymmetric hydrogena-
tion through dynamic kinetic resolution with a
RuCl2[(S)-BINAP][(R)-DMAPEN]/potassium tert-
butoxide catalyst [BINAP=2,2’-bis(diphenylphos-
phino)-1,1’-binaphthyl, DMAPEN=2-dimethylami-
no-1-phenylethylamine]. A series of aryl heterocy-
cloalkyl ketones has been converted to the alcohols
in excellent diastereo- and enantioselectivities. The
modes of catalyst-controlled diastereoselection and
enantioselection are interpreted by using transition-
state molecular models. (S,S)-Reboxetine, a selec-
tive norepinephrine uptake inhibitor, was synthe-
sized from one of product alcohols.
We selected heterocycloalkyl ketones 1a and 1c as
substrates to examine the diastereoselective ability of
reducing agents [Eq. (1)]. The small electronic and
Keywords: diastereoselectivity; enantioselectivity;
hydrogenation; ketones; reboxetine; ruthenium
Figure 1. Typical transition-state models in the diastereose-
lective hydride reduction of a-substituted chiral ketones. L,
M, and S=large-, medium-, and small-sized substituents, re-
spectively. X=a group able to coordinate to metal. Met=
metal.
Adv. Synth. Catal. 2011, 353, 1955 – 1960
ꢁ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1955

Masaya Akashi et al.
COMMUNICATIONS
Table 1. Diastereoselective metal hydride reduction of ketones 1.[a]
Entry
1
Reagent[b]
Conditions
Yield [%][c]
2:3[d]
1
2
3
4
5
6
a
a
a
c
c
c
NaBH4 (1.0)
MeOH, 08C, 2 h
THF, À788C, 3 h
THF, À788C, 3 h
MeOH, 08C, 3 h
THF, À788C, 3 h
THF, À788C, 3 h
99
93
79
99
29
92
60:40
80:20
70:30
53:47
59:41
49:51
Zn
KB(s-Bu)3H (1.2)
NaBH4 (1.0)
Zn(BH4)2 (1.2)
KB(s-Bu)3H (1.2)
ACHTUGNERTN(NUNG BH4)2 (1.2)
AHCTUNGTRENNUNG
AHCTUNGTRENNUNG
AHCTUNGTRENNUNG
[a]
Reactions using 1 mmol of 1.
The molar equivalent of reagents to 1 is shown in the parentheses.
Isolated yield.
[b]
[c]
[d]
Determined by HPLC analysis.
À
À
steric differences between the X(b)CH2(g) (X=O, considered to proceed through the transition state
d+
dÀ
d+
CH2) and CH2(b)NR(g) (R=Bz, Boc) moieties of (TS), in which the HdÀ Ru
N
H
quadrupole of
À
À
À
À
À
the heterocycles makes it difficult to achieve high ste- the Ru catalyst interacts with the Cd+=OdÀ dipole of
reoselectivity. Table 1 lists the results of reduction of the ketone (the metal-ligand cooperative TS). The
1a and 1c with the typical metal hydride reagents, carbonyl moiety is placed with a definite direction.
NaBH4,[5] Zn(BH4)2,[6] and KB(sec-C4H9)3H.[7] In the The reaction site is surrounded by the phosphine-
ACHTUNGTRENNUNG ACHTUNGTRENNUGN
reduction of 1a, a 2a (syn)/3a (anti) mixture in a ratio ligand skeleton, which is schematically illustrated as a
ranging from 60:40 to 80:20 was obtained (entries 1– blue wedge in this figure. By contrast with the metal
3). The relatively high selectivity of 80:20 observed in hydride reduction, the ketonic substrate is hydrogen-
the reaction with ZnACHTNUGRTNEUNG(BH4)2 may have been caused by ated within the field of the catalyst. We assumed that
the formation of the chelate intermediate. 3-Piperi- the reaction field could appropriately fix the confor-
dinyl ketone 1c was reduced to give an almost 1:1 mation of 1, and then the two diastereomeric faces of
mixture of 2c and 3c (entries 4–6).
the carbonyl group are differentiated by the bulkiness
We have reported the asymmetric hydrogenation of around the b and g positions, but not the a position
ketones catalyzed by chiral diphosphine/diamine-Ru close to the reaction site. In the hydrogenation of 1c,
complexes.[8] A range of aromatic, hetero-aromatic, the g nitrogen atom with a bulky substituent (R) is
cyclic, and unsaturated ketones as well as some kinds predominantly placed at the position facing the
of aliphatic ketones is hydrogenated with high reactiv- “open” side of the reaction field to afford the syn-al-
ity and enantioselectivity. Our recent mechanistic cohol 2c selectively. This is because the other side of
studies on the hydrogenation of ketones revealed that the reaction field is “shielded” by the walls of the
the RuH2(diphosphine)
cies.[8–10] As shown in Figure 2, this hydrogenation is pected to give 2a in a same manner.
We selected the hydrogenation of 1a in order to op-
ACHTUNGTNER(NUNG diamine) is the active spe- phosphine ligands. The hydrogenation of 1a was ex-
timize the catalyst structure so that it would exhibit
excellent diastereoselectivity (Table 2).[11] The reac-
tion with RuCl2[PACTHNUTRGNEUNG(C6H5)3]2(EN) (4a) at a substrate-to-
catalyst molar ratio (S/C) of 100 in t-C4H9OK
(30 mM) containing 2-propanol under 10 atm of H2 at
258C afforded 2a and 3a in a 94:6 ratio (entry 1). A
much higher syn-selectivity than that observed in the
metal hydride reductions shown in Table 1 was ob-
tained. When the phosphine ligand P
Ru complex was replaced with P(4-CH3C6H4)3 (elec-
tron rich; 4b) or P[3,5-(CH3)2C6H3]3 (sterically hin-
ACHTNUGERTN(NGNU C6H5)3 of the
Figure 2. A transition state image in the diastereoselective
hydrogenation of 1c with diphosphine/diamine-Ru(II) cata-
lysts. The diastereotopic environment constructed by the li-
N
ACHTUNGTRENNUNG
gands of the catalyst is illustrated with blue lines. R=Boc. dered; 4c), the stereoselectivity was somewhat de-
1956
asc.wiley-vch.de
ꢁ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2011, 353, 1955 – 1960

Catalyst-Controlled Diastereoselection in the Hydrogenation of Heterocycloalkyl Ketones
Table 2. Diastereoselective hydrogenation of ketones 1.[a]
The marked contrast in stereoselectivity between
metal hydride reductions (Table 1) and the hydroge-
nation catalyzed by the 5b/t-C4H9OK system (Table 2,
entries 6–9) suggests that the diastereoface selection
in this hydrogenation is primarily regulated by the
structure around the reaction site of the catalyst
(“catalyst-controlled diastereoselection”).
We next applied this chemistry to the asymmetric
hydrogenation of aryl heterocycloalkyl ketones 1 and
6 through dynamic kinetic resolution (Table 3).[11,12]
RuCl2[(S)-BINAP][(R)-DMAPEN]
[(S,R)-11],
a
chiral analogue of the Ru complex 5b, was selected as
the pre-catalyst.[13,14] When (Æ)-1a (4.45 g, 15 mmol)
was hydrogenated with (S,R)-11 (S/C=5000) in t-
C4H9OK containing 2-propanol under 50 atm of H2
for 20 h, the syn-alcohol (1R,2S)-2a (2a/3a= >99:1)
was obtained in 99% ee quantitatively (entry 1). Both
the diastereoface and enantioface selections of 1a
were successfully performed by the (S,R)-11/t-
C4H9OK catalyst with a rapid stereomutation at the a
chiral center of 1a. Almost perfect stereoselection
was achieved in the reaction of 1c (S/C=10 000,
10 atm H2, 20 h). The complete conversion of this re-
action with an S/C of 20 000 was achieved under an
H2 pressure of 50 atm without loss of stereoselectivity
(entry 3). The reactions of a 3-chlorophenyl ketone 1e
and a thienyl ketone 1f predominantly afforded the
syn-alcohols (1S,2S)-2e and (1S,2S)-2f, respectively
(entries 4 and 5). The 2-(1-tert-butoxycarbonylpiperi-
dinyl) ketone 1g was hydrogenated with the (S,R)-11/
t-C4H9OK catalyst followed by cyclization to furnish
the anti-product (1R,2R)-10 (9/10=1:>99) in 97% ee
(entry 6). The high level of anti-selectivity suggests
Entry
1
Ru cat. S/C[b] H2 [atm] Yield [%][c] 2:3[d]
1
2
3
4
5
a
a
a
a
a
a
a
c
c
4a
4b
4c
4d
5a
5b
5b
5b
5b
100
100
100
100
100
1000
5000
1000
5000
10
10
10
10
10
10
50
10
50
97
96
96
94
96
97
96
100
96
94:6
88:12
90:10
72:28
96:4
>99:1
99:1
>99:1
>99:1
6[e]
7[f,g]
8[e]
9[f]
[a]
Unless otherwise stated, reactions were conducted at
258C using 1 (1.0 mmol, 1.0M) in 2-propanol containing
a Ru complex and t-C4H9OK (30 mM) for 20 h. Com-
plete conversion was observed in all cases.
Substrate-to-catalyst molar ratio.
Isolated yield.
Determined by HPLC analysis.
4 mmol of 1 was used.
15 mmol of 1 was used.
[b]
[c]
[d]
[e]
[f]
À
À
that the NBoc(b)CH2(g) group of 1g is precisely
recognized by this catalyst system in a manner similar
[g]
À
À
to the CH2(b)NBoc(g) moiety of 1c. The N-benzoyl
analogue 1h was also converted to the anti-alcohol
(1R,2R)-3h (2h/3h=1:>99) in 96% ee without cycli-
0.5M solution of 1 was used.
creased (entries 2 and 3). The combination of zation (entry 7). Interestingly, the hydrogenation of 2-
(C6H5)3 with N,N-dimethylethylenediamine tetrahydrofuranyl ketone 6b with the (S,R)-11/t-
PACHTUNGTRENNUNG
(DMEN) instead of the unsubstituted ethylenedia- C4H9OK catalyst exhibited high syn- and enantiose-
mine (EN) significantly decreased the diastereoselec- lectivity, revealing that the catalyst can efficiently
À
À
tivity to 72:28 (4d; entry 4). The higher selectivity of discriminate
between
O(b)CH2(g)
and
À
À
96:4 was obtained when the biphenyl diphosphine
ligand
CH2(b)CH2(g) groups of the tetrahydrofuran ring
2,2’-bis(diphenylphosphino)biphenyl (entry 8). In contrast, the reaction of 2-(1-benzoylpyr-
(BIPHEP) was coupled with EN (5a; entry 5). Finally, rolidinyl) ketone 6h gave the anti-alcohol (1R,2R)-8h
the RuCl2A(BIPHEP)(DMEN) (5b)/t-C4H9OK catalyst (7h/8h=2:98) in 94% ee (entry 10). A 3-pyrrolidinyl
C
ACHTUNGTRENNUNG
system achieved an excellent syn-selectivity of ꢀ99:1 ketone 6d was hydrogenated to afford selectively
(entries 6 and 7). The hydrogenations with an S/C of (1S,2S)-7d (syn/anti=75:25) in 98% ee (entry 9).
1000 under 10 atm H2 and with an S/C of 5000 under
50 atm H2 were completed within 20 h. The 3-piperi- tones with the BINAP/diamine-Ru(II) catalyst has re-
dinyl ketone 1c was also hydrogenated with the 5b/t- vealed that the trans-RuH2A(BINAP)(diamine) is the
C4H9OK catalyst to give the syn-alcohol 2c predomi- active species in the catalytic cycle.[8–10] Figure 3 illus-
nantly (entries 8 and 9). Thus, the optimized catalyst trates molecular models of trans-RuH2[(S)-BINAP]-
As we described above, the hydrogenation of ke-
C
ACHTUNGTRENNUNG
AHCTUNGTRENNUNG
system
can
precisely
differentiate
the [(R)-DMAPEN] based on an X-ray structure of the
À
À
À
À
CH2(b)NBoc(g) moiety from the CH2(b)CH2(g)
RuCl2 precursor (S,R)-11 that was previously reported
as well as the O(b)CH2(g) groups of ketones 1. by our group.[14c] The skewed five-membered chelate
À
À
Adv. Synth. Catal. 2011, 353, 1955 – 1960
ꢁ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
1957

Masaya Akashi et al.
COMMUNICATIONS
d+
dÀ
d+
HdÀ Ru
Hax structure is much smaller than
d+ d+
À
À
N
À
Table 3. Stereoselective hydrogenation of ketones via dy-
namic kinetic resolution.[a]
that of the HdÀ Ru
Heq moiety. The hydro-
dÀ
À
À
N
À
genation of ketone 1a proceeds via six-membered TS,
TSA, TSB, or TSC, in which the HdÀ Ru
N
Hax
d+
dÀ
d+
À
À
À
quadrupole of the catalyst interacts with the ketonic
Cd+=OdÀ dipole (see also Figure 2). The C-benzoyl
and N-benzoyl groups are placed at the equatorial po-
sitions of the morpholine ring with a chair conforma-
tion. TSA, which affords (1R,2S)-2a (syn), has the
most preferable structure, in which the N-benzoyl-
morpholine group fits well with the “V-shape chan-
À À
nel” of BINAPꢀs Phax P Pheq moiety, and the planar
phenyl ring of 1a is faced on the amino Meeq group.
The N-benzoyl group is directed to the outside of the
reaction field and has little influence on the formation
of this TS. On the other hand, TSB, which yields
(1R,2R)-3a (anti), suffers significant steric repulsion
À
between the “wall” of the P Phax and the bulky N-
benzoyl group directed to inside of the reaction field.
TSC, which gives (1S,2R)-2a or (1S,2S)-3a, is also un-
favorable due to destabilization by the notable non-
bonded repulsion between the narrow BINAPꢀs V-
shape channel and the widely spread phenyl ring of
1a. Thus, the reaction environment can precisely
select the diastereoface of 1a, 1c, 1e, and 1f just by
recognizing the g-substituent, which is impossible for
the metal hydride reductions. Furthermore, the enan-
tioface selection is nearly perfect. The diastereoselec-
tion in the reaction of the 2-piperidinyl ketones, 1g
and 1h, is controlled by recognition of the b-substitu-
ent, resulting in the anti-products, 3g and 3h, exclu-
sively. The mode of enantioselection is the same as
that of 1a.[10] The stereoselective outcome of the hy-
drogenation of ketones with five-membered heterocy-
clic rings, 6b, 6d, and, 6h, is rationalized in the same
manner as described above. The hydrogenation of 6b
(X=O, Y =CH2) afforded 7b (syn) and 8b (anti) in a
95:5 ratio, while the reaction of 6h (X=NBz, Y=
CH2) furnished a 2:98 mixture of 7h and 8h (see
Table 3, entries 8 and 10). These results suggest that
the catalyst can strictly differentiate the size of the X
group as NBz>CH2 >O. The 75:25 diastereomeric
ratio for the hydrogenation of 6d means that discrimi-
nation between the b-NBoc moiety and the b-CH2
group of the 3-pyrrolidinyl ring is possible in the cata-
lytic system (see Table 3, entry 9).
Entry Ketone S/C[b] Yield [%][c] syn:anti[d] ee [%][e]
1[f,g]
2[h]
3[f,i]
4[g]
5[j]
1a
1c
1c
1e
1f
1g
1h
6b
6d
6h
5000
10000
20000
5000
500
1000
200
500
99
99
97
>99:1
>99:1
>99:1
>99:1
>99:1
1:>99[l]
1:>99
95:5
99
>99
>99
99
99
97
99
95
6
100[k]
95
7[j,m]
8
96
97
100
100
99[n]
98[n]
94
9
1000
200
75:25
2:98
10[j,o]
[a]
Unless otherwise stated, reactions were conducted under
10 atm of H2 at 258C using ketone (3–4 mmol, 1.0M) in
2-propanol containing (S,R)-11 and t-C4H9OK (30 mM)
for 20 h. Complete conversion was observed in all cases.
Substrate-to-catalyst molar ratio.
Isolated yield.
Determined by HPLC analysis.
Data of the major diastereomer. Determined by chiral
GC or HPLC analysis.
Under 50 atm of H2.
15 mmol of 1 was used.
30 mmol of 1c was used.
40 mmol of 1c was used.
0.5M solution of ketone was used.
Yield of cyclized products 9 and 10.
Ratio of 9 and 10.
Reaction in 2-propanol:CH2Cl2 =5:1.
The ee values of 8b and 8d (anti-isomers) were both
88%.
[b]
[c]
[d]
[e]
[f]
[g]
[h]
[i]
The isolated hydrogenation product (1R,2S)-2a
AHCTUNGTREG(NNUN >99% ee) was readily converted to (S,S)-reboxetine
[j]
[k]
[l]
succinate [(1S,2S)-13] (>99% ee) in four steps
through stereoinversion of the C-1 position
(Scheme 1).[15] (S,S)-Reboxetine is the active enantio-
mer of racemic reboxetine, a selective norepinephrine
uptake inhibitor.[16] The stereoselective hydrogenation
of 1a provides one of the most efficient and practical
procedures for the synthesis of this important com-
[m]
[n]
[o]
Reaction in 2-propanol:CH2Cl2 =6:1.
ring with the (R)-DMAPEN has two diastereotopic pound. (1R,2R)-2e is also a useful synthetic inter-
amino protons, Hax and Heq. The torsion angle of the mediate of a potent anti-hypertensive agent.[17]
1958
asc.wiley-vch.de
ꢁ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2011, 353, 1955 – 1960

Catalyst-Controlled Diastereoselection in the Hydrogenation of Heterocycloalkyl Ketones
Figure 3. Molecular models of the (S,R)-RuH2 complex (*=Ru) derived from 11 and diastereomeric TSs in the hydrogena-
tion of (Æ)-1a through dynamic kinetic resolution. The structures are simplified for clarity. Het=2-(4-benzoylmorpholinyl).
In summary, the RuCl2ACTHNUTRGENN(UG BIPHEP)AHCUTNGTREN(NGUN DMEN) (5b)/t- chemistry is applied to the asymmetric hydrogenation
C4H9OK catalyst system hydrogenates heterocycloalk- via dynamic kinetic resolution by using the (S)-
yl phenyl ketones 1a and 1c in ꢀ99:1 diastereoselec- BINAP/(R)-DMAPEN-Ru(II) catalyst. A series of
tivity. Reduction of these ketones with typical metal aryl heterocycloalkyl ketones 1 and 6 is converted to
hydrides results in an 80:20 diastereoselectivity in the the alcohols in excellent diastereo- and enantioselec-
best case. The contrasting results suggest that the ster- tivities. The modes of catalyst-controlled diastereose-
eoselection in this hydrogenation is primarily regulat- lection as well as enantioselection are interpreted by
ed by the structure of the catalystꢀs reaction field using transition-state molecular models. (S,S)-Reboxe-
(“catalyst-controlled
diastereoselection”).
This tine, a selective norepinephrine uptake inhibitor, is
readily synthesized from the product alcohol (1R,2S)-
2a.
Experimental Section
Typical Procedure for the Asymmetric
Hydrogenation of 1c with (S,R)-11
Ru complex (S,R)-11 (1.9 mg, 2 mmol),[14] ketone 1c (11.57 g,
40 mmol), and 2-propanol (38.8 mL) were placed in a 200-
mL SUS autoclave with a stirring propeller. A solution of t-
C4H9OK in 2-propanol (1M, 1.2 mL, 1.2 mmol) was added
to the autoclave, and the contents in the autoclave were de-
gassed. The vessel was pressurized with hydrogen to 50 atm,
and the reaction mixture was stirred at 258C for 20 h. After
carefully venting the hydrogen gas, the solvent was removed
under reduced pressure. The diastereomeric ratio was deter-
mined to be >99:1 by HPLC analysis of the reaction mix-
Scheme 1. Preparation
of
(S,S)-reboxetine
succinate
ture: GL Science Inertsil ODS-3 column (4.5ꢂ250 mm),
408C, CH3CN:H2O:10% H3PO4 =600 mL:400 mL:1 mL, 1.0
mLminÀ1, 215 nm, retention time (tR) of 3c: 9.6 min, tR of
2c: 11.0 min. The residue was purified by silica-gel column
chromatography (hexane:ethyl acetate=2:1 to 1:1) giving
[(1S,2S)-13]. Conditions: a) CH3SO2Cl, (C2H5)3N, CH2Cl2,
08C; b) 2-ethoxyphenol, Cs2CO3, dioxane, reflux (83% in
two steps); c) DIBAL, toluene, À788C; d) succinic acid, 2-
propanol, reflux (74% in two steps).
Adv. Synth. Catal. 2011, 353, 1955 – 1960
ꢁ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
1959

Masaya Akashi et al.
COMMUNICATIONS
(S,S)-2c; yield: 11.2 g (97%, >99% ee). The ee value of 2c
was determined by HPLC analysis (CHIRALPAK AD-H
column (4.6ꢂ250 mm), 208C, hexane:EtOH:i-PrOH=
97:2:1, 1.0 mLminÀ1, 215 nm): tR of (R,R)-2c: 17.3 min
(<0.5%), tR of (S,S)-2c: 24.4 min (>99.5%); [a]2D8: À30.0
(c=1.04, methanol). 1H NMR (300 MHz, DMSO-d6,
1408C): d=1.16–1.46 (m, 3H, including s, 9H at 1.37), 1.54–
1.68 (m, 2H), 2.59–2.71 (m, 2H), 3.75–3.84 (m, 1H), 4.01–
4.10 (m, 1H), 4.33 (dd, 1H, J=6.7, 4.6 Hz), 4.66 (d, 1H, J=
4.6 Hz), 7.18–7.31 (m, 5H). The procedure and physical data
of products are described in detail in the Supporting Infor-
mation.
Petersen, L. R. Sita, Angew. Chem. 1985, 97, 1–31;
Angew. Chem. Int. Ed. Engl. 1985, 24, 1–30.
[5] H. I. Schlesinger, H. C. Brown, A. E. Finholt, J. Am.
Chem. Soc. 1953, 75, 205–209.
[6] T. Oishi, T. Nakata, Acc. Chem. Res. 1984, 17, 338–344.
[7] C. A. Brown, J. Am. Chem. Soc. 1973, 95, 4100–4102.
[8] Recent reviews: a) T. Ohkuma, Proc. Jpn. Acad. Ser. B
2010, 86, 202–219; b) R. Noyori, T. Ohkuma, Angew.
Chem. 2001, 113, 40–75; Angew. Chem. Int. Ed. 2001,
40, 40–73.
[9] a) C. A. Sandoval, T. Ohkuma, K. MuÇiz, R. Noyori, J.
Am. Chem. Soc. 2003, 125, 13490–13503. See also:
b) K. Abdur-Rashid, S. E. Clapham, A. Hadzovic, J. N.
Harvey, A. J. Lough, R. H. Morris, J. Am. Chem. Soc.
2002, 124, 15104–15118; c) R. J. Hamilton, S. H. Ber-
gens, J. Am. Chem. Soc. 2006, 128, 13700–13701.
[10] See the Supporting Information for details.
[11] For diastereoselective hydrogenation of cyclic and acy-
clic ketones with BINAP/diamine-Ru(II) catalysts, see:
T. Ohkuma, H. Ooka, M. Yamakawa, T. Ikariya, R.
Noyori, J. Org. Chem. 1996, 61, 4872–4873.
Acknowledgements
This work was supported by a Grant-in-Aid from the Japan
Science and Technology Agency (JST) and the Japan Society
for the Promotion of Science (JSPS) (No. 21350048).
[12] For stereoselective synthesis via dynamic kinetic resolu-
tion, see: R. Noyori, M. Tokumaga, M. Kitamura, Bull.
Chem. Soc. Jpn. 1995, 68, 36–56.
References
[1] Selected reviews: a) H. C. Brown, S. Krishnamurthy,
Tetrahedron 1979, 35, 567–607; b) E. L. Eliel, in: Asym-
metric Synthesis, Vol. 2, (Ed.: J. D. Morrison), Academ-
ic Press, New York, 1983, pp 125–155; c) N. Greeves,
in: Comprehensive Organic Synthesis, Vol. 8, (Eds.:
B. M. Trost, I. Fleming), Pergamon Press, Oxford, 1991,
pp 1–24; d) E. Juaristi, Introduction to Stereochemistry
and Conformational Analysis, Wiley, New York, 1991,
pp 178–198; e) A. P. Davis, in: Methods of Organic
Chemistry – (Houben-Weyl), 4th edn., Vol. E21d,
(Eds.; G. Helmchen, R. W. Hoffmann, J. Mulzer, E.
Schaumann), Thieme, Stuttgart, 1995, pp 3988–4048;
f) A. Mengel, O. Reiser, Chem. Rev. 1999, 99, 1191–
1223.
[2] For the Felkin–Anh model, see: a) M. Chꢃrest, H.
Felkin, N. Prudent, Tetrahedron Lett. 1968, 2199–2204;
b) N. T. Anh, Top. Curr. Chem. 1980, 88, 145–162. See
also: c) D. J. Cram, F. A. A. Elhafez, J. Am. Chem. Soc.
1952, 74, 5828–5835; d) M. N. Paddon-Row, N. G.
Rondan, K. N. Houk, J. Am. Chem. Soc. 1982, 104,
7162–7166.
[13] BINAP=2,2’-bis(diphenylphosphino)-1,1’-binaphthyl;
DMAPEN=2-dimethylamino-1-phenylethylamine.
[14] a) N. Arai, H. Ooka, K. Azuma, T. Yabuuchi, N.
Kurono, T. Inoue, T. Ohkuma, Org. Lett. 2007, 9, 939–
941; b) N. Arai, K. Azuma, N. Nii, T. Ohkuma, Angew.
Chem. 2008, 120, 7567–7570; Angew. Chem. Int. Ed.
2008, 47, 7457–7460; c) H. Ooka, N. Arai, K. Azuma,
N. Kurono, T. Ohkuma, J. Org. Chem. 2008, 73, 9084–
9093.
[15] For example, see: a) E. Brenner, R. M. Baldwin, G.
Tamagnan, Org. Lett. 2005, 7, 937–939; b) K. E. Hane-
gar, M. Cebula, Org. Process Res. Dev. 2007, 11, 354–
358; c) T.-X. Mꢃtro, G. Pardo, J. Cossy, J. Org. Chem.
2008, 73, 707–710.
[16] Racemic reboxetine is produced commercially as an
antidepressant, see: K. E. Hanegar, C. T. Ball, C. M.
Horvath, K. D. Maisto, S. E. Mancini, Org. Process Res.
Dev. 2007, 11, 346–353.
[17] Z. Xu, S. Cacatian, J. Yuan, R. D. Simpson, L. Jia, W.
Zhao, C. M. Tice, P. T. Flaherty, J. Guo, A. Ishchenko,
S. B. Singh, Z. Wu, B. M. McKeever, B. B. Scott, Y.
Bukhtiyarov, J. Berbaum, J. Mason, R. Panemangalore,
M. G. Cappiello, R. Bentley, C. P. Doe, R. K. Harrison,
G. M. McGeehan, L. W. Dillard, J. J. Baldwin, D. A.
Claremon, Bioorg. Med. Chem. Lett. 2010, 20, 694–699.
[3] For the chelation-controlled model, see: a) D. J. Cram,
K. R. Kopecky, J. Am. Chem. Soc. 1959, 81, 2748–2755;
b) M. T. Reetz, Angew. Chem. 1984, 96, 542–555;
Angew. Chem. Int. Ed. Engl. 1984, 23, 556–569.
[4] Reagent control has been recognized as not important
for diastereoselection, see: S. Masamune, W. Choy, J. S.
1960
asc.wiley-vch.de
ꢁ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2011, 353, 1955 – 1960