P-chiral β-aminophosphine oxides vs. β-aminophosphines as auxiliaries for ruthenium catalysed enantioselective transfer hydrogenation of arylketones

Article abstract of DOI:10.1016/S0022-328X(01)00678-7

Enantiopure P-chiral β-aminophosphine oxides and the corresponding β-aminophosphines have been synthesised and used as chiral auxiliaries in ruthenium catalysed asymmetric transfer hydrogenation of arylketones producing optically active alcohols up to 80%

Full text of DOI:10.1016/S0022-328X(01)00678-7

Journal of Organometallic Chemistry 626 (2001) 157–160
www.elsevier.nl/locate/jorganchem
P-chiral b-aminophosphine oxides vs. b-aminophosphines as
auxiliaries for ruthenium catalysed enantioselective transfer
hydrogenation of arylketones
Anna M. Maj a, K. Michal Pietrusiewicz a,b,*, Isabelle Suisse c, Francine Agbossou c,
Andre´ Mortreux c,*
a Department of Organic Chemistry, Maria Curie-Sklodowska Uni6ersity, ul. Gliniana 31, 20031 Lublin, Poland
b Institute of Organic Chemistry, Polish Academy of Sciences, ul. Kasprzaka 44, 01-224 Warsaw, Poland
c Laboratoire de Catalyse de Lille, UPRESA 8010, ENSCL, BP 108, 59652 Villeneu6e d’Ascq Cedex, France
Received 27 November 2000; accepted 5 January 2001
Abstract
Enantiopure P-chiral b-aminophosphine oxides and the corresponding b-aminophosphines have been synthesised and used as
chiral auxiliaries in ruthenium catalysed asymmetric transfer hydrogenation of arylketones producing optically active alcohols up
to 80% ee. Both types of auxiliaries provide comparable induction levels but the b-aminophosphine oxide ligands induce higher
catalytic activities generally. In some experiments, when changing the achiral arene ligand in the catalyst precursor, a peculiar
reversal of the product configuration was observed. © 2001 Elsevier Science B.V. All rights reserved.
Keywords: Ruthenium; Transfer hydrogenation; Chirality; Ketone; Alcohol; Phosphine oxide
ety in their structure proved to be the most efficient.
We sought to undertake a systematic study to compare
the properties of the above chiral b-aminophosphine
oxides and the corresponding P-chiral b-aminophos-
phines obtained by the reduction of the P-chiral alkyl-
arylphosphinoyl moiety. However, the above-
mentioned ligands could not be converted into the
corresponding P-chiral aminophosphines. Thus, new
ligands bearing a reducible chiral phosphinoyl group
were sought.
Here we report the synthesis and use in asymmetric
transfer hydrogenation of two new P-chiral aminophos-
phine oxide ligands and their corresponding P-chiral
aminophosphines. This time the former has borne a
readily reducible methylphenylphosphinoyl group,
which was meant to secure their easy conversion into
the corresponding phosphines. These new ligands also
bear a secondary NH functionality, which appears to
be essential to obtain good activity [2]. The selected
ligands enable us to compare the efficiency of P-chiral
aminophosphine oxide vs. P-chiral aminophosphine in
the asymmetric transfer hydrogenation of the ketones
studied.
1. Introduction
Asymmetric hydrogen transfer using alcohols as the
hydrogen source appears to be an especially valuable
method for the reduction of prochiral substrates into
optically active compounds because of the easy set up
of the experiments [1]. Ketones have been reduced with
high levels of efficiency and enantioselectivity in the
presence of ruthenium [2], iridium [3], and rhodium [4]
catalysts. Very efficient catalytic systems have been
developed for ruthenium associated to aminophosphi-
nes [5], aminoalcohols [6], monotosylated-1,2-diamines
[7], and diamines [8].
Recently, we have demonstrated that asymmetric
transfer hydrogenation of alkylarylketones could be
achieved efficiently in the presence of chiral b-
aminophosphine oxides [9]. We have found that the
ligands containing a P-chiral alkylarylphosphinoyl moi-
* Corresponding author.
E-mail addresses: michal@hermes.umcs.lublin.pl (K. M. Piet-
rusiewicz), isabelle.suisse@ensc-lille.fr (I. Suisse), andre.mortreux@
ensc-lille.fr (A. Mortreux).
0022-328X/01/$ - see front matter © 2001 Elsevier Science B.V. All rights reserved.
PII: S0022-328X(01)00678-7

158
A.M. Maj et al. / Journal of Organometallic Chemistry 626 (2001) 157–160
Scheme 1.
2. Results and discussion
The obtained ligands were then examined in the
ruthenium catalysed transfer hydrogenation of arylke-
tones 3–5 to the corresponding chiral alcohols (Scheme
2). The Ru catalysts were prepared by heating the
precursor [RuCl2(arene)]2 (A: arene=p-cymene, B: hexa-
methylbenzene) with two equivalents of the ligand in
i-PrOH at 80°C during 20 min. The catalytic reactions
were carried out following the procedure described
previously [9]. The results are summarised in Table 1.
The examination of the results indicates clearly that
with each of the tested ligands, the best enantioselectiv-
ity was achieved in the reduction of i-butyrophenone
(5) and when [RuCl2(hexamethylbenzene)]2 (B) was
used as the catalyst precursor.3 The highest ees (80–
78%) were reached with aminophosphine ligands 1b
and 2b although it should be noted that with the
corresponding aminophosphine oxides 1a and 2a only a
slightly lower level of induction was observed (65–
76%). In the context of the four best results (entries 6,
12, 18, and 24), it could be reasonably argued that the
absolute configuration of the product is governed by
the carbon centred chirality of the amine residue re-
gardless of the configuration and the oxidation level of
the phosphine part of the ligand used. Intriguingly,
however, when the arene ligand in the catalyst precur-
sor was changed from hexamethylbenzene to p-cymene
in those i-butyrophenone reductions, a dramatic de-
crease of catalyst activity as well as of the resulting ee
was observed (entries 5, 11, 17, and 23). In addition, in
the two cases involving phosphine oxide ligands, this
deteriorating effect was also accompanied by the rever-
sal of the configuration of the major alcohol produced
(entries 5 and 17). It thus appears that the nature of the
achiral ligand can also play a crucial role in the asym-
metric transfer hydrogenations, as it was already re-
ported for other substrates [14].
The optically pure diastereomeric b-aminophosphine
oxides SP,SC-1a and SP,RC-2a were prepared by the
reaction of (SP)-vinylmethylphenylphosphine oxide with
(SC)- and (RC)-1-phenylethylamine, respectively, follow-
ing our previously reported protocol (Scheme 1) [9,10].
The oxides 1a and 2a were then reduced stereoselec-
tively under standard conditions using HSiCl3 in the
presence of triethylamine [11]. Under these conditions,
the inversion of configuration at phosphorus is rou-
tinely observed [12].1 This step provided the corre-
sponding P-chiral aminophosphines SP,SC-1b and
SP,RC-2b in quantitative yields (Scheme 1). Impor-
tantly, no loss of the optical purity was observed during
this process. The oxidation of 1b and 2b by means of
H2O2 in chloroform, known to occur with a clean
retention of configuration [13], provided ent-2a(RP,SC)
and ent-1a(RP,RC), respectively, which accordingly ex-
hibited specific rotations equal in magnitude and oppo-
site in sign to those recorded for 2a and 1a. This
correlation confirmed unequivocally the optical purity
and the assigned absolute configuration of phosphines
1b and 2b.2
Scheme 2.
For acetophenone (3) and propiophenone (4) reduc-
tions, the p-cymene catalyst precursor A proved more
suitable than B and for both substrates similar ees were
obtained. Again, the above-mentioned reversal of
configuration was observed here (entries 10 vs. 9 and 22
vs. 21) and it is difficult to trace a clear-cut correlation
between the ligand and the product chiralities. How-
1 The inversion of the spatial arrangement of the groups around
phosphorus in the reduction step does not require a change of the
descriptor of the absolute configuration. As a matter of fact, in the
process, the oxygen ligand of the highest priority is replaced by a lone
pair of the lowest priority, according to the Cahn–Ingold–Prelog
system.
2 Selected physical properties: 1a: 31P-NMR (CDCl3) l=36.1 ppm,
[h]D= +27.3 (c 5.3, CHCl3)10; 1b: 31P-NMR (CDCl3) l= −40
ppm, [h]D= −5.6 (c 3.2, CHCl3); 2a: 31P-NMR (CDCl3) l=36.1
ppm, [h]D= −56.1 (c 3.2, CHCl3)11; 2b: 31P-NMR (CDCl3) l=
−40 ppm, [h]D = −24.3 (c 3.4, CHCl3).
3 This favourable substrate–catalyst matching was already ob-
served in our previous study involving analogous aminophosphine
oxide ligands. See Ref. [9].

A.M. Maj et al. / Journal of Organometallic Chemistry 626 (2001) 157–160
159
Table 1
Transfer hydrogenation of arylketones in the presence of ruthenium catalysts a
Entry
Chiral ligand
Substrate
Ru precursor
Time (h)
Conversion (%) b
ee (%) b
Configuration c
1
2
3
4
5
6
1a
3
3
4
4
5
5
A
B
A
B
A
B
1
0.1
3
0.25
24
3
95
93
94
95
22
93
55
50
53
0
8
76
R
R
R
–
S
R
7
8
9
10
11
12
1b
2a
2b
3
3
4
4
5
5
A
B
A
B
A
B
17
1.5
24
2.5
24
17
95
93
85
65
74
92
57
49
57
9
13
80
S
S
S
R
R
R
13
14
15
16
17
18
3
3
4
4
5
5
A
B
A
B
A
B
3
0.25
3
1
17
3
91
91
92
97
48
93
39
22
34
16
5
S
S
S
S
R
S
65
19
20
21
22
23
24
3
3
4
4
5
5
A
B
A
B
A
B
4
4
17
17
3
93
71
71
57
44
83
23
19
18
43
6
S
S
S
R
S
S
43
78
a Reactions were carried out by using 2 mmol of the substrate in 20 ml i-PrOH at 20°C in the presence of either complex A [RuCl2(p-cymene)]2
or B [RuCl2(hexamethylbenzene)]2, (substrate/Ru=100), the ligand (ligand/Ru: 2) and KOH (0.1 mmol).
b The conversions of the substrate and the ees were monitored by GC analysis using a Chirasildex capillary column.
c Absolute configurations were determined by comparing the sign of the optical rotations with those of the literature.
ever, the carbon centred chirality seems to dominate in
the phosphine oxide ligands (entries 1–6 vs. 13–18)
whereas for the phosphine ones, this tendency is less
pronounced (entries 7–12 vs. 19–24).
latter is much more governed by the carbon configura-
tion within the ligand frame.
The catalyst activities in the studied hydrogen trans-
fer reactions were generally much higher for the phos-
phine oxide ligands 1a and 2a than for the phosphine
ligands 1b and 2b. For example, under identical condi-
tions, transfer hydrogenation of acetophenone 3 with
the system [RuCl2(p-cymene)]2–1a led to 95% conver-
sion and 55% ee within 1 h whereas with 1b as the
chiral auxiliary, the same 95% conversion and 57% ee
was achieved only after a 17 h period (entries 1 vs. 7).
This higher catalyst activity can be explained by the
inherent hemilabile character [15] of the phosphine
oxide ligands which can generate an open coordination
site at ruthenium more easily, thus allowing a faster
substrate complexation. The tendency observed above
when considering the outcome of the chirality at phos-
phorous in the sense of enantioselection can also illus-
trate the hemilabile properties of the phosphine oxides.
Indeed, the chirality at phosphorous presents a greater
influence and dictates in most cases the configuration of
the product for the most coordinating ligand, e.g. the
phosphines, whereas using the phosphine oxides, the
3. Conclusions
In summary, we have shown that the P-chiral
aminophosphine oxides constitute an interesting source
of precursors for the synthesis of P-chiral aminophos-
phines, which are valuable chiral ligands for the asym-
metric catalysis. For the ruthenium catalysed transfer
hydrogenation of ketones, the latter are, however, less
efficient than their parent P-chiral aminophosphine ox-
ides. We have indeed demonstrated that the use of
readily available and air-stable P-chiral b-aminophos-
phine oxides as chiral ligands in ruthenium catalysed
asymmetric transfer hydrogenation of ketones results in
the formation of efficient catalytic systems which yield
products of high enantiomeric purity. We are still un-
able to definitely prove that the phosphine oxide moiety
of those ligands really coordinate to the ruthenium
centre but the recent results published by Faller and
coworkers [15] suggest strongly that such coordination
can be effective. Studies in this direction are in
progress.

160
A.M. Maj et al. / Journal of Organometallic Chemistry 626 (2001) 157–160
J.-X. Gao, T. Ikariya, R. Noyori, Organometallics 15 (1996)
Acknowledgements
1087.
[6] (a) M. Palmer, T. Walsgrove, M. Wills, J. Org. Chem. 62 (1997)
5226. (b) J. Takehara, S. Hashiguchi, A. Fujii, S. Inoue, T.
Ikariya, R. Noyori, Chem. Commun. (1996) 233. (c) D.A.
Alonso, D. Guijarro, P. Pinho, O. Temme, P.G. Andersson, J.
Org. Chem. 63 (1998) 2749. (d) J.A. Kenny, K. Versluis, A.J.R.
Heck, T. Walsgrove, M. Wills, Chem. Commun. (2000) 99.
[7] (a) S. Hashiguchi, A. Fujii, T. Ikariya, R. Noyori, J. Am. Chem.
Soc. 117 (1995) 7562. (b) K. Pu¨ntener, L. Schwink, P. Knochel,
Tetrahedron Lett. 37 (1996) 8165. (c) L. Cai, Y. Han, H.
Mahmoud, B.M. Segal, J. Organomet. Chem. 568 (1998) 77.
[8] (a) Y. Jiang, Q. Jiang, X. Zhang, J. Am. Chem. Soc. 120 (1998)
3817. (b) L. Schwink, T. Ireland, K. Pu¨ntener, P. Knochel,
Tetrahedron: Asymmetry 9 (1998) 1143.
This work was supported by a grant (A.M.M.) from
the ‘Ministe`re des Affaires Etrange`res’. The authors
gratefully thank the ‘Ministe`re de la Recherche et de la
Technologie’ and the ‘Centre National de la Recherche
Scientifique’ for financial support.
References
[1] (a) M.J. Palmer, M. Wills, Tetrahedron: Asymmetry 10 (1999)
2045. (b) G. Zassinovich, G. Mestroni, S.Gladiali, Chem. Rev.
92 (1992) 1051.
[2] R. Noyori, S. Hashigushi, Acc. Chem. Res. 30 (1997) 97.
[3] (a) K. Mashima, T. Abe, K. Tani, Chem. Lett. (1998) 1199. (b)
K. Murata, T. Ikariya, R. Noyori, J. Org. Chem. 64 (1999) 2186.
(c) C. Bianchini, L. Glendenning, F. Zanobini, E. Farnetti, M.
Graziani, E. Nagy, J. Mol. Catal. 132 (1998) 13.
[4] (a) see Refs. [3a,b]. (b) F. Touchard, M. Bernard, F. Fache, F.
Delbecq, V. Guiral, P. Sautet, M. Lemaire, J. Organomet. Chem.
567 (1998) 133.
[5] (a) J.-X. Gao, X.-D. Yi, P.-P. Xu, C.-L. Tang, H.-L. Wan, T.
Ikariya, J. Organomet. Chem. 592 (1999) 290. (b) Y. Jiang, Q.
Jiang, G. Zhu, X. Zhang, Tetrahedron Lett. 38 (1997) 6565. (c)
[9] A.M. Maj, K.M. Pietrusiewicz, I. Suisse, F. Agbossou, A.
Mortreux, Tetrahedron: Asymmetry 10 (1999) 831.
[10] K.M. Pietrusiewicz, M. Zablocka, Tetrahedron Lett. 29 (1988)
1991.
[11] B.E. Smart, Org. Synth. 67 (1988) 20.
[12] (a) G.M. Kosolapoff, L. Maier, Organic Phosphorous Com-
pounds, vol. 3, Wiley, New York, 1972, p. 408. (b) K.M.
Pietrusiewicz, M. Zablocka, Chem. Rev. 94 (1994) 1375.
[13] L. Horner, Pure Appl. Chem. 9 (1964) 225.
[14] K. Everaere, J.-F. Carpentier, A. Mortreux, M. Bulliard, Tetra-
hedron: Asymmetry 9 (1999) 2971.
[15] J.W. Faller, B.P. Patel, M.A. Albrizzio, M. Curtis, Organometal-
lics 18 (1999) 3096.
.