Asymmetric synthesis of α-chiral hydroxyalkylphosphines by a catalytic enantioselective reduction of acylphosphines

Article abstract of DOI:10.1021/ol5024757

Enantioselective reduction of acylphosphines, after precomplexation with borane, proceeded smoothly in the presence of a chiral oxazaborolidine catalyst and catecholborane. α-Hydroxyalkylphosphine products were obtained as phosphine-borane complexes in good yield and enantioselectivity. One of the products of the enantioselective reduction was successfully applied as an optically active phosphine ligand for asymmetric catalysis after suitable derivatization.

Full text of DOI:10.1021/ol5024757

Letter
pubs.acs.org/OrgLett
Asymmetric Synthesis of α‑Chiral Hydroxyalkylphosphines by a
Catalytic Enantioselective Reduction of Acylphosphines
Minoru Hayashi,* Hiroyuki Ishitobi, Yutaka Matsuura, Takashi Matsuura, and Yutaka Watanabe
Department of Materials Science and Biotechnology, Graduate School of Science and Engineering, Ehime University, 3 Bunkyo-cho,
Matsuyama 790-8577, Japan
S
* Supporting Information
ABSTRACT: Enantioselective reduction of acylphosphines,
after precomplexation with borane, proceeded smoothly in
the presence of a chiral oxazaborolidine catalyst and
catecholborane. α-Hydroxyalkylphosphine products were ob-
tained as phosphine−borane complexes in good yield and
enantioselectivity. One of the products of the enantioselective
reduction was successfully applied as an optically active phosphine ligand for asymmetric catalysis after suitable derivatization.
he development of new chiral phosphine ligands is a
fascinating research topic since asymmetric reactions
cleanly afforded the expected alcohol 3a as a phosphine−
borane complex without cleavage of the P−C/P−O bond
(Scheme 1).10 At least 2 equiv of borane was required due to
T
catalyzed by a chiral phosphine−transition-metal complex
provide a number of advantages in the organic synthesis of
chiral molecules.1 Most chiral phosphines have a chirality
derived from optically active natural compounds, whether it is
embedded into the skeleton or indirectly transferred during
resolution processes. On the other hand, catalytic asymmetric
synthesis of a chiral phosphine itself is a promising and
challenging subject that can lead to the efficient preparation of
various chiral phosphines.2 Chiral phosphines having a chiral
carbon atom at the α position to the phosphorus atom
generally show good performance in selectivity when they are
used as a chiral ligand.3 Hydroxyalkylphosphines are one of the
promising precursors of chiral phosphines.4 However, only a
few preparations of α-chiral hydroxyalkylphosphines through
asymmetric addition of a phosphorus nucleophile to carbonyl
compounds have been reported.5 α-Chiral hydroxyalkylphos-
phines could also be accessed if the carbonyl group of
acylphosphines were reduced with enantioselectivity. Actually,
there are a few examples of the asymmetric reduction of
acylphosphines and α-ketophosphonates.6
Herein, we report a successful asymmetric reduction of
acylphosphines for the preparation of chiral α-hydroxyalkyl-
phosphine derivatives. A successful application of one of these
products as a chiral ligand in a Pd-catalyzed asymmetric
allylation of an amine is also described.
Acylphosphines having a P(III) atom are conveniently and
quantitatively synthesized by the reaction of acyl halides with
silylphosphines.7 Acylphosphines are key intermediates for the
synthesis of low coordinate phosphorus compounds.8 However,
they have rarely been used in the synthesis of trivalent
phosphines, while their oxides have been successfully and
widely used as photochemical radical initiators and functional
additives for UV curing materials and inkjet ink.9
Scheme 1. Reduction of 1a with BH3
consumption of borane by complexation with the phosphine.
Use of other reagents, such as LiAlH4 and DIBAL, caused
partial cleavage of the P−C bond.
It is advantageous to obtain the product as a phosphine−
borane complex since the BH3 group serves as a useful
protective group for a trivalent phosphine preventing it from
undergoing oxidation during organophosphine synthesis.10a,11
Since complexation of the phosphorus atom with borane
confers high electrophilicity on the adjacent carbonyl carbon,
the real species to be reduced is considered as the
acylphosphine−borane complex. However, an NMR examina-
tion of the reduction of acetylphosphine 1b with BH3·THF (1.0
equiv) revealed that a two-step sequence of complexation and
reduction proceeded at the same time to give a mixture of the
products including hydroxyethylphosphine−borane 3b, acetyl-
phosphine−borane 2b, together with noncomplexed 1b (Figure
1d). Since reaction of a single species is preferable for
asymmetric versions of the reduction, selective complexation
of the phosphorus with BH3 prior to the reduction was
investigated using several borane reagents. In contrast to the
above result with BH3·THF, BH3·SMe2 was found to be
suitable for the stepwise reaction; an equimolar amount of BH3·
SMe2 selectively afforded the complex 2b without reduction
(Figure 1b), and the reduction slowly proceeded to give 3b
Initial attempts were made to search for a suitable reducing
agent for acylphosphines. Among tested reducing agents in the
reduction of benzoyldiphenylphosphine 1a, only BH3 reagent
Received: August 21, 2014
© XXXX American Chemical Society
A
dx.doi.org/10.1021/ol5024757 | Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
Table 1. Enantioselective Reduction of 1b under Various
a
Conditions
b
c
d
e
entry
cat.
borane
conditions
yield (%)
% ee
f
1
2
A
A
A
A
B
C
D
A
A
A
A
A
BMS
23 °C, 6 h
23 °C, 1 h
23 °C, 6 h
−42 °C, 3 h
23 °C, 6 h
23 °C, 6 h
23 °C, 6 h
23 °C, 18 h
−78 °C, 6 h
23 °C, 6 h
−42 °C, 6 h
−78 °C, 6 h
91
91
84
51
88
86
85
84
77
80
81
70
85
54
27
0
f
BMS
g
f
3
BMS
f
4
BMS
f
5
BMS
26
11
32
74
3
f
6
BMS
f
7
BMS
8
PB
PB
Figure 1. 162 MHz 31P NMR experiments of the stepwise
complexation and reduction of acetylphosphine 1b in CH2Cl2.12
9
h
10
11
12
CB
87
96
99
h
CB
h
CB
after the addition of another BH3·SMe2 (Figure 1c). These
results clearly indicate that in situ precomplexation using BH3·
SMe2 can make the reduction process simple as a reduction of
the complex 2b.
a
Unless otherwise noted, 1 equiv of BH3·SMe2 was added prior to the
b
catalytic reduction. Unless otherwise noted, 20 mol % of catalyst was
used. Catalyst: A, Me-CBS-oxazaborolidine ((S)-4) = (S)-(−)-B-
methyl-4,4-diphenyl-1,3,2-oxazaborolidine; B, (−)-diphenylprolinol +
additional BH3·SMe2 (1 equiv); C, Ph-CBS-oxazaborolidine = (S)-
(−)-B-phenyl-4,4-diphenyl-1,3,2-oxazaborolidine; D, acyloxyborane =
The CBS reduction is one of the most effective
enantioselective reductions of a carbonyl group with borane
reagents using a chiral oxazaborolidine catalyst.13 Therefore, we
investigated the CBS reduction of acylphosphine−borane
complexes prepared in situ using excess BH3·SMe2 reagent
for both complexation and reduction. Initial investigation of the
enantioselective reduction of 1b was performed by using 20
mol % (S)-(−)-B-methyl-CBS-oxazaborolidine (S)-4 with 2.5
equiv BH3·SMe2. Reaction at 23 °C for 6 h gave the 1-
hydroxyethylphosphine−borane complex (S)-3b in high yield
and selectivity (Table 1, entry 1). Then, several catalysts,
reagents, and conditions were surveyed for optimizing the
enantioselective reduction of 1b (Table 1). Poor enantiose-
lectivity was observed when lower amounts of the catalyst were
used (entry 3). Other oxazaborolidines and related catalysts
were not effective for the present reduction; all catalysts gave
the alcohol in good yield but with low-to-moderate
enantioselectivity (entries 5−7).
During this optimization, we found anomalous behavior for
the present reaction in regard to enantioselectivity. It is
unusual, but quite important, that the enantioselectivity of the
present reduction depends on the reaction period; enantiose-
lectivity gradually improves as the reaction period increases
(entries 1 and 2), despite the same chemical yield being
observed in each reaction. In addition, the reaction at low
temperature gave racemic alcohol 3b (entry 4), probably
because the catalytic reaction was slower than the noncatalytic
reduction at this temperature (noncatalytic reduction was
confirmed to occur slowly, even at −42 °C, to give the racemic
alcohol).14 Moreover, enantioselectivity was induced by catalyst
(S)-4 from racemic 1-hydroxyethylphosphine−borane once
prepared in situ.15 Both the time dependency of selectivity and
the induction of chirality from a racemic product cannot be
simply interpreted by the mechanistic investigations reported
for ketone reduction, such as dimerization of oxazaborolidine
catalyst at low temperature.14 These results show that the
present reduction does not involve a simple kinetic control of
the facial selection, but other mechanisms of chirality
induction.16
c
(−)-proline + additional BH3·SMe2 (1 equiv). Borane reagent (1.5
equiv): BMS = BH3·SMe2; PB = pinacolborane; CB = catecolborane.
d
e
Isolated yield. Determined by HPLC analysis using Daicel Chiralpak
f
g
IA column. BH3·SMe2 (2.5 equiv) was added in one portion. 5 mol
% of catalyst was used. Catecholborane (1.8 equiv) was used.
h
Other borane sources were investigated as reductants for
enantioselective reduction of acetylphosphine−borane 2b,
prepared in situ by the reaction of 1b with 1 equiv of BH3·
SMe2 in CH2Cl2 (entries 8−12). These results clearly showed
that catecholborane was found to be the best for the
enantioselective reduction of acylphosphine−borane 2b.
Temperature dependencies were also observed in these
investigations, though they depended on the reductant used.
Reduction of acetylphosphine by catecholborane showed
different dependencies from other cases; this seemed to involve
simple kinetic control, since the reaction proceeded at lower
temperature in better yield and selectivity.
By applying the two-step reactions with BH3·SMe2 (complex-
ation) and catecholborane (reduction) with Me-CBS catalyst,
several acylphosphines were successfully reduced to the
corresponding α-hydroxyalkylphosphine−borane complexes in
good yield and enentioselectivity (Table 2).
The results shown in Table 2 indicate that the induction of
selectivity is related to electronic effects in the substituent R.
Acylphosphines with an electron-withdrawing substituent on
the carbonyl group (1b−d and 1j) tend to show better
selectivity at low temperature, probably because of a rapid
reaction, due to high reactivity of the electronically activated
carbonyl group resulting in kinetic control of the catalytic
reaction predominantly. The stereoselectivity of the present
reaction is in accordance with that expected in the CBS
reduction of prochiral ketones.13 Oxalyldiphosphine 1m gave
the corresponding diphosphine derivative, though the ee and
chemical yield were poor. On the other hand, several
alkanoylphosphines (1e, 1f) and aroylphosphines (1a, 1h, 1i)
gave opposite results; they gave better selectivity at higher
B
dx.doi.org/10.1021/ol5024757 | Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
Table 2. Enantioselective Reduction of Acylphosphines
Using (S)-4 and Catecholborane.
Scheme 2. Application of Phosphine Carbamate (S)-5 as a
Chiral Ligand for Pd-Catalyzed Enantioselective Allylation
of Benzylamine
a
−78 °C
% yield
23 °C
b
b
time
(h)
time
(h)
% yield
c
c
1
(% ee )
(% ee )
PhCOPPh2 (1a)
MeCOPPh2 (1b)
ClCH2COPPh2 (1c)
4
6
6
6
28 (8)
24
24
81 (85)
89 (90)
91 (90)
78 (71)
d
91 (95)
91 (98)
90 (87)
In conclusion, we have found that acylphosphines are
successfully reduced by borane reagents to produce α-chiral
hydroxyalkylphosphines as a borane complex. Me-CBS
oxazaborolidine was found to be effective for the enantiose-
lective reduction of the precomplexed acylphosphine−borane
prepared in situ. High enantioselectivities can be achieved
under suitable conditions, although the course of enantiose-
lection strongly depends on the electronic properties of the
substituent in the acylphosphines. The resulting chiral
phosphine can be modified and successfully applied to
asymmetric catalysis as a chiral ligand. Further applications of
the products and mechanistic elucidation of the unusual
chirality induction of the present reduction process are now
under investigation.
e
7
e
MeOCH2COPPh2
(1d)
7
EtCOPPh2 (1e)
iPrCOPPh2 (1f)
tBuCOPPh2 (1g)
7
7
f
59 (33)
17 (5)
f
6
4
91 (97)
68 (91)
g
23
24
0 (−)
p-MeC6H4COPPh2
(1h)
8
9 (25)
84 (84)
36 (73)
67 (53)
p-MeOC6H4COPPh2
(1i)
f
f
24
10
p-NO2C6H4COPPh2
(1j)
20
85 (97)
h
,
i
j
MeCOPCy2 (1k)
f
f
f
f
f
f
10
24
5
33 (68)
h,
MeCOP(tBu)2 (1l)
6 (35)
k
Ph2PCOCOPPh2 (1m)
25 (21)
a
The reaction was conducted by using 20 mol % of (S)-4 with
ASSOCIATED CONTENT
* Supporting Information
Determinaton of the absolute configuration of the product and
plausible mechanisms of chirality induction. Experimental
procedures and characterization data for all new compounds.
This material is available free of charge via the Internet at
http://pubs.acs.org.
■
catecholborane (1.8 equiv) after precomplexation with BH3·SMe2 (1
b
c
S
equiv) in CH2Cl2. Isolated yield. Enantiomeric excess was
d
determined by HPLC analysis and shown in parentheses. By using
e
(R)-4, (R)-3b was obtained in 85%, 98% ee. The reaction was
f
g
conducted at 0 °C. Not examined. No alcohol was formed.
Pivaloyldiphenylphosphine−borane was observed instead in the
h
mixture. Enantiomeric excess was determined after conversion to
i
phenoxyacetate. By using BH3·SMe2 as a reductant, 3k was obtained
j
in 64%, 83% ee (23 °C, 20 h). By using BH3 SMe2 as a reductant, 3l
AUTHOR INFORMATION
Corresponding Author
k
■
was obtained in 14%, 48% ee (23 °C, 24 h). Diastereomer mixture
(dl/meso = 88/12).
*E-mail: mhayashi@ehime-u.ac.jp.
temperature. These results clearly indicated that there are at
least two possible courses of selectivity-determining steps based
on the electronic situation in the carbonyl carbon. Dialkyl-
phosphino derivatives (1k, 1l) also gave the corresponding
hydroxyalkylphosphines, though low yield and selectivity were
observed under the standard conditions. Since the dialkylphos-
phino group in 1k and 1l has a more electron-donating nature
than diphenyl derivatives, they are less reactive than diphenyl
derivatives. Use of a more reactive reductant such as BMS is
sometimes effective in these cases. With appropriate choice of
the reaction conditions, the present borane reduction provides
a versatile route toward chiral α-hydroxyalkylphosphine
derivatives.
Application of the resulting hydroxyalkylphosphines was
investigated as a source of a chiral phosphine ligand for a
transition-metal-catalyzed asymmetric reaction. Several mod-
ifications17 of the hydroxy group in 1-hydroxyethylphosphine-
borane complex (S)-3b revealed that the carbamate (S)-5,
prepared by a treatment of (S)-3b with phenyl isocyanate, was
as an effective chiral ligand for Pd-catalyzed asymmetric
allylation. After deprotection of the borane complex (S)-5·
BH3 by treatment with morpholine,18 phosphine carbamate
(S)-5-Pd complex successfully catalyzed the allylation of
benzylamine in excellent enantioselectivity (Scheme 2).19
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
We thank the Applied Protein Research Laboratory of Ehime
University for NMR measurement.
■
REFERENCES
■
(1) (a) Jacobsen, E. N.; Pfaltz, A., Yamamoto, H., Eds. Comprehensive
Asymmetric Catalysis; Springer-Verlag, Berlin, 1999. (b) Ojima, I., Ed.
Catalytic Asymmetric Synthesis, 3rd ed.; Wiley: New York, USA, 2010.
(c) Borner, A., Ed.; Phosphorus Ligands in Asymmetric Catalysis; Wiley,
̈
New York, 2008.
(2) (a) Glueck, D. S. Chem.Eur. J. 2008, 14, 7108−7117.
(b) Sadow, A. D.; Haller, I.; Fadini, L.; Togni, A. J. Am. Chem. Soc.
2004, 126, 14704−14705. (c) Sadow, A. D.; Togni, A. J. Am. Chem.
Soc. 2005, 127, 17012−17024. (d) Butti, P.; Rochat, R.; Sadow, A. D.;
Togni, A. Angew. Chem., Int. Ed. 2008, 47, 4878−4881. (e) Carlone,
A.; Bartoli, G.; Bosco, M.; Sambri, L.; Melchiorre, P. Angew. Chem., Int.
Ed. 2007, 46, 4504−4506. (f) Ibrahem, I.; Rios, R.; Vesely, J.;
Hammar, P.; Eriksson, L.; Himo, F.; Cordova, A. Angew. Chem., Int. Ed.
2007, 46, 4507−4510. (g) Chan, V. S.; Bergman, R. G.; Toste, F. D. J.
Am. Chem. Soc. 2007, 129, 15122−15123.
(3) Recent representative examples of effective asymmetric catalysis
using α-chiral phosphines: (a) Sawano, T.; Ashouri, A.; Nishimura, T.;
C
dx.doi.org/10.1021/ol5024757 | Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
Hayashi, T. J. Am. Chem. Soc. 2012, 134, 18936−18939. (b) Chen, Y.-
R.; Duan, W.-L. Org. Lett. 2011, 13, 5824−5826. (c) Holz, J.; Zayas,
O.; Jiao, H.; Baumann, W.; Spannenberg, A.; Monsees, A.; Riermeier,
T. H.; Almena, J.; Kadyrov, R.; Borner, A. Chem.Eur. J. 2006, 12,
̈
5001−5013. (d) Gini, F.; Hessen, B.; Minnaard, A. J. Org. Lett. 2005,
7, 5309−5312.
(4) Holz, J.; Quirmbach, M.; Borner, A. Synthesis 1997, 983−1006.
(5) (a) Matsuura, Y.; Yamasaki, T.; Watanabe, Y.; Hayashi, M.
Tetrahedron: Asymmetry 2007, 18, 2129−2132. (b) Kolodiazhnyi, O.
I.; Guliaiko, I. V.; Kolodiazhna, A. O. Tetrahedron Lett. 2004, 45,
6955−6957.
́ ́
(6) (a) Fernandez-Perez, H.; Benet-Buchholz, J.; Vidal-Ferran, A.
Org. Lett. 2013, 15, 3634−3637. (b) Nesterov, V. V.; Kolodiazhnyi, O.
I. Tetrahedron 2007, 63, 6720−6731.
(7) (a) Kunzek, H.; Braun, M.; Nesener, E.; Ruhlmann, K. J.
̈
Organomet. Chem. 1973, 49, 149−156. (b) Lindner, E.; Frey, G. Chem.
Ber. 1980, 113, 3268−3274. (c) Lindner, E.; Hubner, D. Chem. Ber.
̈
1983, 116, 2574−2590.
(8) (a) Becker, G.; Mundt, O. Z. Anorg. Allg. Chem. 1980, 462, 130−
142. (b) Becker, G. Z. Anorg. Allg. Chem. 1977, 430, 66−76.
(9) (a) Jockusch, S.; Turro, N. J. J. Am. Chem. Soc. 1998, 120,
11773−11777. (b) Corrales, T.; Catalina, F.; Peinado, C.; Allen, N. S.
J. Photochem. Photobiol. A: Chemistry 2003, 159, 103−114. (c) Leca,
D.; Fensterbank, L.; Laco
858−865. (d) Zalibera, M.; Steb
̂
te, E.; Malacria, M. Chem. Soc. Rev. 2005, 34,
e, P.-N.; Dietliker, K.; Grutzmacher,
́
́
̈
H.; Spichty, M.; Gescheidt, G. Eur. J. Org. Chem. 2014, 331−337.
(10) (a) Imamoto, T.; Oshiki, T.; Onozawa, T.; Kusumoto, T.; Sato,
K. J. Am. Chem. Soc. 1990, 112, 5244−5252. (b) Consiglio, G. B.;
Queval, P.; Harrison-Marchand, A.; Mordini, A.; Lohier, J.-F.;
́
Delacroix, O.; Gaumont, A.-C.; Gerard, H.; Maddaluno, J.; Oulyadi,
H. J. Am. Chem. Soc. 2011, 133, 6472−6480.
(11) (a) Imamoto, T. Pure Appl. Chem. 1993, 65, 655. (b) Miura, T.;
Yamada, H.; Kikuchi, S. I.; Imamoto, T. J. Org. Chem. 2000, 65, 1877.
(12) Aliquot of the reaction mixture in CH2Cl2 was dissolved in C6D6
and the NMR spectrum was measured. Phosphine−borane com-
pounds usually show a couple of broad doublet signals in 31P{1H}
spectra due to B−P coupling.
(13) (a) Corey, E. J.; Shibata, S.; Bakshi, R. K. J. Org. Chem. 1988, 53,
2861−2863. (b) Cho, B. T. Chem. Soc. Rev. 2009, 38, 443−452.
(14) (a) Xu, J.; Wei, T.; Zhang, Q. J. Org. Chem. 2003, 68, 10146−
10151. (b) Xu, J.; Wei, T.; Lin, S.-S.; Zhang, Q. Helv. Chim. Acta 2005,
88, 180−186.
(15) Acylphosphine 1b was reacted with BH3·SMe2 (2.7 equiv) in
CH2Cl2 in the presence of (N,N-dimethylamino)ethanol (0.2 equiv) at
rt for 11 h. After the formation of alcohol was confirmed by TLC
analysis, (S)-4 (0.2 equiv) was added to the mixture, and then the
mixture was stirred for 6 h at rt. Enantiopurity of the product 3b was
determined by HPLC analysis (53% ee).
(16) Proposal of some plausible mechanisms for enantioselection
with some experimental evidence is described in the Supporting
Information.
(17) Several conversions of (S)-3 are described in the Supporting
Information.
(18) Conversion of the hydroxy group is necessary before
deprotection of phosphine−borane 3 with amine due to spontaneous
degradation; see: Moiseev, D. V.; Patrick, B. O.; James, B. R. Inorg.
Chem. 2007, 46, 11467−11474.
(19) (a) Hayashi, T.; Yamamoto, A.; Ito, Y.; Nishioka, E.; Miura, H.;
Yanagi, K. J. Am. Chem. Soc. 1989, 111, 6301−6311. (b) Trost, B. M.;
Bunt, R. C. J. Am. Chem. Soc. 1994, 116, 4089−4090. (c) Sudo, A.;
Saigo, K. J. Org. Chem. 1997, 62, 5508−5513. (d) Evans, D. A.;
́
Campos, K. R.; Tedrow, J. S.; Michael, F. E.; Gagne, M. R. J. Org.
Chem. 1999, 64, 2994−2995. (e) Uozumi, Y.; Tanaka, H.; Shibatomi,
K. Org. Lett. 2004, 6, 281−283. (f) Nemoto, T.; Masuda, T.; Akimoto,
Y.; Fukuyama, T.; Hamada, Y. Org. Lett. 2005, 7, 4447−4450.
NOTE ADDED AFTER ASAP PUBLICATION
Reference 6a was added in the version reposted September 25,
2014.
■
D
dx.doi.org/10.1021/ol5024757 | Org. Lett. XXXX, XXX, XXX−XXX