Amino acid derived amides and hydroxamic acids as ligands for asymmetric transfer hydrogenation in aqueous media

Article abstract of DOI:10.1016/j.catcom.2011.03.032

Amides and hydroxamic acids derived from α-amino acids were evaluated as ligands in combination with rhodium and iridium half-sandwich complexes in asymmetric transfer hydrogenation (ATH) of ketones. The reactions were performed in aqueous media using lithium formate as hydride source. The catalyst systems turned out to be highly efficient and ee's up to 90% were obtained.

Full text of DOI:10.1016/j.catcom.2011.03.032

Catalysis Communications 12 (2011) 1118–1121
Contents lists available at ScienceDirect
Catalysis Communications
journal homepage: www.elsevier.com/locate/catcom
Short Communication
Amino acid derived amides and hydroxamic acids as ligands for asymmetric transfer
hydrogenation in aqueous media
⁎
Katrin Ahlford, Hans Adolfsson
Department of Organic Chemistry, Stockholm University, The Arrhenius Laboratory, SE-106 91 Stockholm, Sweden
a r t i c l e i n f o
a b s t r a c t
Article history:
Amides and hydroxamic acids derived from α-amino acids were evaluated as ligands in combination with
rhodium and iridium half-sandwich complexes in asymmetric transfer hydrogenation (ATH) of ketones. The
reactions were performed in aqueous media using lithium formate as hydride source. The catalyst systems
turned out to be highly efficient and ee's up to 90% were obtained.
Received 28 February 2011
Accepted 21 March 2011
Available online 30 March 2011
© 2011 Elsevier B.V. All rights reserved.
Keywords:
Asymmetric catalysis
Reduction
Ketones
Iridium
Rhodium
1. Introduction
nature of these ligands, in combination with easily accessible chiral
starting material (i.e. amino acid building blocks), make them highly
Chiral secondary alcohols are key intermediates in the industrial
production of several important fine chemicals [1]. An efficient and
straightforward route towards the formation of enantiomerically
enriched alcohols is the reduction of the corresponding prochiral
ketones. Asymmetric transfer hydrogenation (ATH) has proven to
be an efficient, mild and highly selective method for this particular
transformation. A further advantage is that hazardous molecular
hydrogen or highly reactive hydride reagents can be avoided [2–4].
The ATH reaction is commonly conducted in alcoholic media, among
which 2-propanol is the classical example [5,6]. Conveniently, the
solvent also acts as the hydrogen donor. The reaction can furthermore
be performed in formic acid or variants thereof, where the most
common examples are the triethylamine/formic acid azeotrope
(TEAF) and alkali formate-salts in water. However, the number of
efficient catalyst systems that are compatible with these conditions
are rather limited [7]. The combination of [Ru(p-cymene)Cl2]2
together with the monotosylated diamine ligand 1,2-diphenyl-1,2-
diaminoethane (TsDPEN) developed by Noyori, is an exception, since
it is highly efficient in all reaction media mentioned above [8,9].
Previously, we have reported that N-Boc-protected amino acid
derived hydroxamic acids (1) are excellent chiral mediators in the
rhodium catalyzed ATH [10,11] in 2-propanol [12–15]. The modular
attractive for asymmetric catalysis. Ligands containing amino acid
moieties have been widely explored in recent years by us [16–19] as
well as by other research groups [20–25].
The use of formic acid or a formate salt as hydride donor in ATH is
most advantageous since the reductions thereby can be performed
under seemingly irreversible conditions. However, when the valine
based ligand 1 (R=isopropyl) was used together with [RhCp*Cl2]2 as
catalyst for the reduction of acetophenone using sodium formate as
hydride donor and water as solvent, poor conversion of the starting
material was observed (Table 1, entry 1). Most likely, the N-protecting
group on 1 in combination with the aqueous media restricts the ligand
from forming a catalytically active complex. We therefore decided to
evaluate the deprotected versions of our hydroxamic acid ligands
using the abovementioned conditions, and found to our delight that
they efficiently catalyze the ATH reaction also in aqueous media. A
⁎ Corresponding author. Tel.: +46 8 162484; fax: +46 8 154908.
E-mail addresses: katrin@organ.su.se (K. Ahlford), hansa@organ.su.se
(H. Adolfsson).
1566-7367/$ – see front matter © 2011 Elsevier B.V. All rights reserved.
doi:10.1016/j.catcom.2011.03.032

K. Ahlford, H. Adolfsson / Catalysis Communications 12 (2011) 1118–1121
1119
modular low molecular weight catalyst system of this kind is of course
highly attractive and the applicability is herein further extended. In
addition to hydroxamic acid based ligands, we decided to investigate
and compare the catalytic performance of the structurally even
simpler amino acid amides. Faller and Lavoie previously reported that
ruthenium-arene- and rhodium-arene complexes of L-prolineamide
catalyze the reduction of ketones under transfer hydrogenation
conditions using 2-propanol as solvent/hydride donor [26]. Further-
more, other ruthenium and rhodium amino acid amide complexes
have previously been evaluated as catalysts for ATH of ketones in the
formate/water system [27–31].
phase and passed through a pad of silica using EtOAc as eluent. The
resulting solutions were analyzed by GLC (CP Chirasil DEXCB).
2.3. General procedure for the asymmetric transfer hydrogenation of
ketones catalyzed by [IrCp*Cl2]2 and L-proline hydroxamic acid 6
[IrCp*Cl2]2 (4 mg, 0.005 mmol) and L-proline hydroxamic acid 6
(1.3 mg, 0.01 mmol) were evacuated together with SDS (29 mg,
0.1 mmol) and HCOOLi×H2O (350 mg, 5 mmol) in a Schlenk tube for
10 min. Degassed distilled water (2.5 mL) was added to the tube
under nitrogen atmosphere after which the ketone (1 mmol) was
added, and thereafter the reaction mixture was stirred at 24 °C for
14 h. The reaction was quenched by addition of EtOAc (2.5 mL) to the
reaction mixture after which aliquots were taken from the organic
phase and passed through a pad of silica using EtOAc as eluent. The
resulting solutions were analyzed by GLC (CP Chirasil DEXCB).
2. Experimental
2.1. General procedure for the asymmetric transfer hydrogenation of
ketones catalyzed by [RhCp*Cl2]2 and L-proline amide 4
3. Results and discussion
[RhCp*Cl2]2 (3.1 mg, 0.005 mmol) and L-proline amide 4 (1.14 mg,
0.01 mmol) were evacuated together with SDS (29 mg, 0.1 mmol)
and HCOOLi×H2O (350 mg, 5 mmol) in a Schlenk tube for 10 min.
Degassed distilled water (2.5 mL) was added to the tube under
nitrogen atmosphere after which the ketone (1 mmol) was added,
and thereafter the reaction mixture was stirred at 35 °C for 18 h. The
reaction was quenched by addition of EtOAc (2.5 mL) to the reaction
mixture after which aliquots were taken from the organic phase and
passed through a pad of silica using EtOAc as eluent. The resulting
solutions were analyzed by GLC (CP Chirasil DEXCB).
Amino acid amide ligands 2–4 and hydroxamic acid ligands 5–6
were readily prepared according to Scheme 1 [32]. Both the amide and
the hydroxamic acid can conveniently be obtained from the same
protected hydroxamic acid intermediate, which in turn was prepared
by coupling of the corresponding Cbz-protected amino acid with
hydroxylamine or O-benzylhydroxylamine. As illustrated in Scheme 1,
the choice of catalyst used in the hydrogenation/deprotection step
determined which product class was obtained. Hence, hydrogenolysis
using Pd/C resulted in the formation of amino acid amides, and for the
preparation of the corresponding hydroxamic acid, Pd(OH)2/C was
used [33]. It should be noted that attempts to prepare the hydroxamic
acid from valine using this strategy failed, and regardless of catalyst
used we obtained the primary amide.
2.2. General procedure for the asymmetric transfer hydrogenation of
ketones catalyzed by [RhCp*Cl2]2 and L-proline hydroxamic acid 6
[RhCp*Cl2]2 (3.1 mg, 0.005 mmol) and L-proline hydroxamic acid
6 (1.3 mg, 0.01 mmol) were evacuated together with SDS (29 mg,
0.1 mmol) and HCOOLi×H2O (350 mg, 5 mmol) in a Schlenk tube for
10 min. Degassed distilled water (2.5 mL) was added to the tube
under nitrogen atmosphere after which the ketone (1 mmol) was
added, and thereafter the reaction mixture was stirred at 24 °C for
14 h. The reaction was quenched by addition of EtOAc (2.5 mL) to the
reaction mixture after which aliquots were taken from the organic
With the set of ligands in hand we started to optimize the reaction
conditions using acetophenone as a substrate, and with the
isoelectronic complexes [RhCp*Cl2]2 and [IrCp*Cl2]2 as metal pre-
cursors (Table 1). The solubility of substrate and ligand/catalyst in
water is rather poor, and as a standard procedure we choose to
perform the reactions with SDS as additive.
Initially, the reaction was carried out using amide ligand 2 and
sodium formate as reducing agent (entry 2, Table 1), however; the use
Scheme 1. Preparation of amino acid amides and hydroxamic acid ligands (2–6).

1120
K. Ahlford, H. Adolfsson / Catalysis Communications 12 (2011) 1118–1121
Table 1
conversion of acetophenone is obtained after 17 h. In comparison to
Optimization of reaction conditions using ligands 2–6.a
reactions performed at lower temperature the selectivity of 1-
phenylethanol only decreased to 76% ee (entry 9, Table 1). Further
increase of the temperature to 40 °C led to lower enantioselectivity
(73% ee). We next turned our attention towards the hydroxamic acid
ligands 5 and 6. The selectivities obtained using the hydroxamic acid
ligands are in line with the corresponding amino acid amides and
thus, the proline-derived hydroxamic acid ligand 6 gave the best
results (entry 12, Table 1). The reaction went smoothly even at room
temperature and full conversions were obtained using ligand 6
together with either iridium or rhodium (entries 12 and 14, Table 1).
In contrast to the proline amide ligand 4, superior selectivity was
obtained with [IrCp*Cl2]2 as compared to [RhCp*Cl2]2. The use of the
isoelectronic [Ru(p-cymene)Cl2]2 complex as catalyst precursor led to
low conversion and moderate ee (entry 15, Table 1). Summarizing the
results presented in Table 1, it is obvious that the use of the proline-
derived ligands (4 and 6) results in the most productive catalysts. The
selectivity obtained in the reduction of acetophenone is similar
regardless of metal precursor, and therefore we decided to pursue a
comparative substrate screen using the rhodium complex of 4 and
both the rhodium and iridium complexes of ligand 6. Previous
investigations have shown that different substrates may give rise to
quite diverse results when screened with structurally similar catalysts
[35]. In fact, the modularity and simplicity of the current ligand/
Entry
Metal
Ligand
Conversion [%]b
ee [%]b
1c
[RhCp*Cl2]2
[RhCp*Cl2]2
[RhCp*Cl2]2
[IrCp*Cl2]2
[RhCp*Cl2]2
[IrCp*Cl2]2
[RhCp*Cl2]2
[RhCp*Cl2]2
[RhCp*Cl2]2
[IrCp*Cl2]2
[RhCp*Cl2]2
[RhCp*Cl2]2
[RhCp*Cl2]2
[IrCp*Cl2]2
1
2
2
2
3
3
4
4
4
4
5
6
6
6
6
2
99
96
99
99
99
55
80
99
99
99
99
99
99
6
–
2c
46 (R)
52 (R)
71 (R)
34 (R)
43 (R)
78 (R)
77 (R)
76 (R)
72 (R)
29 (R)
69 (R)
68 (R)
74 (R)
41 (R)
3
4d
5
6
7
8e
9f
10
11
12d
13c,d
14d
15d,g
[Ru(p-cymene)Cl2]2
a
Reduction of 1 mmol acetophenone (S/C 100/1) with 5 equiv. HCOOLi and 10 mol%
SDS in 2.5 mL distilled H2O at 28 °C for 17–22 h.
b
Conversions and enantioselectivities were determined by GLC analysis (CP Chirasil
DEX CB).
c
HCOONa was used as reducing agent.
24 °C.
32 °C.
35 °C.
d
e
f
g
40 h.
catalyst system makes it ideal for such
investigation.
a multidimensional
of lithium formate resulted in higher selectivity (entry 3, Table 1).
When employing [IrCp*Cl2]2 as metal precursor, the selectivity was
even further increased (entry 4, Table 1). The use of the sterically
more demanding ligand 3 derived from phenylalanine did not
improve the selectivity of the reaction (entries 5 and 6, Table 1), but
it is again observed that the use of the iridium precursor gives slightly
higher selectivity as compared to its rhodium counterpart. Ligand 4 on
the other hand is less flexible in comparison to the other amino acid
derived ligands, and should therefore result in a more strained chelate
when coordinated to the metal [34]. The best results are indeed
obtained when the rigid ligand 4 derived from proline is employed.
Here, the opposite trend is observed and the best selectivity was
measured with [RhCp*Cl2]2 as the metal source, whereas the
conversion was considerably lower in comparison to the use of
the iridium-complex (entries 7 and 10, Table 1). Interestingly, the
rhodium-catalyzed reaction did reach full conversion when
the reaction mixture was left for 70 h, and the selectivity remained
at 78% ee. In order to increase the rate of the rhodium-catalyzed
reaction, a temperature variation investigation was performed to
optimize the reaction conditions further. As expected, the selectivity
decreases with increasing reaction temperature, however, most
gratifyingly the decrease is only minor. The best balance between
conversion, time and selectivity is obtained at 35 °C, where full
A substrate screen was consequently performed under the
optimized conditions, where differently functionalized and substitut-
ed acetophenones were reduced (Scheme 2 and Table 2). Substrates
containing both electron rich and electron poor substituents were
evaluated.
Regarding the proline amide derived catalyst, conversions are
generally high or excellent, whereas the selectivity differs among the
substrates evaluated. The best result is obtained with (3′-methoxy)
acetophenone (19), which gives a conversion of 98% into the
corresponding secondary alcohol and an ee of 88% (R). High
selectivities are also obtained in the reduction of 2-acetonaphthone
(9), (4′-trifluoromethyl)acetophenone (13) and (3′,5′-dimethoxy)
acetophenone (17), where the ee's range from 80% to 85% in favor of
the R-isomer. The use of the hydroxamic acid based catalysts resulted
in excellent conversions for most substrates, even at shorter reaction
time as compared to reactions performed with the amide catalyst (14
Table 2
Asymmetric transfer hydrogenation of ketones 7–21.a
b
c
c
Ketone
4/[RhCp*Cl2]2
6/[RhCp*Cl2]2
6/[IrCp*Cl2]2
Conv.
[%]
ee
[%]
Conv.
[%]
ee
[%]
Conv.
[%]
ee
[%]
7
8
9
99
57
82
97
95
92
56
46
59
94
83
68
98
73
65
76 (R)
48 (R)
85 (R)
69 (R)
54 (R)
71 (R)
81 (R)
58 (R)
78 (R)
76 (R)
80 (R)
65 (R)
88 (R)
73 (R)
46 (R)
98
73
70 (R)
83 (R)
99
99
73 (R)
90 (R)
10
11
12
13
14
15
16
17
18
19
20
21
97
95
96
40 (R)
31 (R)
69 (R)
97
99
99
52 (R)
27 (R)
76 (R)
76
92
90 (R)
63 (R)
86
98
80 (R)
76 (R)
95
81
62 (R)
rac
96
99
76 (R)
30 (R)
a
Reduction of 1 mmol substrate using [MCp*Cl2]2 and ligand 4/6 (S/C 100/1)
together with 5 equiv. HCOOLi and 10 mol% SDS in 2.5 mL distilled H2O. Conversions
and enantioselectivities were determined by GLC analysis (CP Chirasil DEX CB).
b
Reactions performed at 35 °C for 18 h.
Reactions performed at 24 °C for 14 h.
c
Scheme 2. Optimized reaction conditions, and ketones evaluated in the ATH.

K. Ahlford, H. Adolfsson / Catalysis Communications 12 (2011) 1118–1121
1121
versus 18 h). Evidently, the reactions catalyzed by the hydroxamic
Appendix A. Supplementary data
acid derived complexes are more sensitive to substitution in the
ortho-position of the substrate aryl-ring, as compared to the
corresponding amide complexes. The selectivities are in general
higher using the iridium precursor than with its rhodium counterpart,
albeit for certain substrates higher selectivity was obtained using the
latter catalyst. The best selectivity obtained using the iridium catalyst
was in the reduction of 2-acetonaphthone (9) where the product was
formed in 90% ee (R). Using the rhodium-based catalyst we obtained
the same level of enantioselectivity (90%) in the reduction of (4′-
methyl)acetophenone (15).
From the presented results it is apparent that the amino acid
derived amides as well as hydroxamic acids are efficient ligands in the
ATH reaction in aqueous media. Interestingly, there is a strong
correlation between the amino acid side chain and the degree of
selectivity obtained with catalysts from the two different ligand
classes. Moreover, the selectivity-correlation is in agreement with the
results obtained employing the parent amino acids as ligands together
with [RhCp*Cl2]2 under similar reaction conditions. Hence, the use of
L-proline gives the best selectivity in the formation of 1-phenyletha-
nol (76% ee (R)), with L-valine a 50% ee (R) is obtained, and using L-
phenylalanine results in a merely 33% ee (R). It should be pointed out
that the conversion was substantially lower (around 20%, reactions
performed at ambient temperature) using the amino acid based
catalysts. The correlation is interesting from a fundamental perspec-
tive, since an initial ligand screen can be performed with non-
functionalized amino acids, rather than using more complex ligand
structures. Similar levels of selectivity should be expected from the
simpler catalyst systems, and these results can therefore act as a
guideline in further ligand developments.
Supplementary data to this article can be found online at
doi:10.1016/j.catcom.2011.03.032.
References
[1] V. Farina, J.T. Reeves, C.H. Senanayake, J.J. Song, Chem. Rev. 106 (2006) 2734.
[2] C. Wang, X. Wu, J. Xiao, Chem. Asian J. 3 (2008) 1750.
[3] S. Gladiali, E. Alberico, Chem. Soc. Rev. 35 (2006) 226.
[4] K. Everaere, A. Mortreux, J.-F. Carpentier, Adv. Synth. Catal. 345 (2003) 67.
[5] H.C. Maytum, J. Francos, D.J. Whatrup, J.M.J. Williams, Chem. Asian J. 5 (2010) 538.
[6] T. Zweifel, J.-V. Naubron, T. Buttner, T. Ott, H. Gruetzmacher, Angew. Chem. Int. Ed.
Engl. 47 (2008) 3245.
[7] S. Gladiali, R. Taras, in: P.G. Andersson, I.J. Munslow (Eds.), Modern Reduction
Methods, Wiley-VCH, Weinheim, 2008, pp. 135–157.
[8] X. Wu, C. Wang, J. Xiao, Platinum Metals Rev. 54 (2010) 3.
[9] R. Noyori, S. Hashiguchi, Acc. Chem. Res. 30 (1997) 97.
[10] K. Ahlford, J. Ekström, A.B. Zaitsev, P. Ryberg, L. Eriksson, H. Adolfsson, Chem. Eur.
J. 15 (2009) 11197.
[11] K. Ahlford, A.B. Zaitsev, J. Ekström, H. Adolfsson, Synlett (2007) 2541.
[12] W. Baratta, F. Benedetti, A. Del Zotto, L. Fanfoni, F. Felluga, S. Magnolia, E.
Putignano, P. Rigo, Organometallics 29 (2010) 3563.
[13] M.T. Reetz, X. Li, J. Am. Chem. Soc. 128 (2006) 1044.
[14] S.J. Nordin, P. Roth, T. Tarnai, D.A. Alonso, P. Brandt, P.G. Andersson, Chem. Eur. J. 7
(2001) 1431.
[15] D.S. Matharu, D.J. Morris, A.M. Kawamoto, G.J. Clarkson, M. Wills, Org. Lett. 7
(2005) 5489.
[16] L. Zani, L. Eriksson, H. Adolfsson, Eur. J. Org. Chem. (2008) 4655.
[17] A.B. Zaitsev, H. Adolfsson, Org. Lett. 8 (2006) 5129.
[18] A. Bøgevig, I.M. Pastor, H. Adolfsson, Chem. Eur. J. 10 (2004) 294.
[19] I.M. Pastor, P. Västilä, H. Adolfsson, Chem. Commun. (2002) 2046.
[20] J. Mao, J. Guo, Chirality 22 (2010) 173.
[21] B. Liégault, X. Tang, C. Bruneau, J.-L. Renaud, Eur. J. Org. Chem. (2008) 934.
[22] P. Pelagatti, A. Bacchi, F. Calbiani, M. Carcelli, L. Elviri, C. Pelizzi, D. Rogolino, J.
Organomet. Chem. 690 (2005) 4602.
[23] D. Carmona, F.J. Lahoz, R. Atencio, L.A. Oro, M.P. Lamata, F. Viguri, E. San José, C.
Vega, J. Reyes, F. Joó, Á. Kathó, Chem. Eur. J. 5 (1999) 1544.
[24] D. Carmona, M.P. Lamata, F. Viguri, I. Dobrinovich, F.J. Lahoz, L.A. Oro, Adv. Synth.
Catal. 344 (2002) 499.
4. Conclusion
[25] T. Ohta, S.-I. Nakahara, Y. Shigemura, K. Hattori, I. Furukawa, Appl. Organomet.
Chem. 15 (2001) 699.
[26] J.W. Faller, A.R. Lavoie, Organometallics 20 (2001) 5245.
[27] Z. Zhou, L. Wu, Catal. Commun. 9 (2008) 2539.
[28] S. Zeror, J. Collin, J.-C. Fiaud, L.A. Zouioueche, J. Mol. Catal. A Chem. 256 (2006) 85.
[29] H.Y. Rhyoo, H.-J. Park, W.H. Suh, Y.K. Chung, Tetrahedron Lett. 43 (2002) 269.
[30] H.Y. Rhyoo, H.-J. Park, Y.K. Chung, Chem. Commun. (2001) 2064.
[31] H.Y. Rhyoo, Y.-A. Yoon, H.-J. Park, Y.K. Chung, Tetrahedron Lett. 42 (2001) 5045.
[32] G. Giacomelli, A. Porcheddu, M. Salaris, Org. Lett. 5 (2003) 2715.
[33] M. Rowley, P.D. Leeson, B.J. Williams, K.W. Moore, R. Baker, Tetrahedron 48
(2002) 3557.
In summary, we have reported two classes of simple, low
molecular weight ligands, namely amino acid derived primary amides
and hydroxamic acids, that together with [RhCp*Cl2]2 or [IrCp*Cl2]2
are efficient and compatible catalysts for transfer hydrogenation not
only in 2-propanol as previously published, but also in aqueous media.
When using these ligands in the water/lithium formate/SDS system,
conversions and selectivities are high for most of the substrates
evaluated, proving the usefulness of these catalyst systems.
[34] A. Bacchi, P. Pelagatti, C. Pelizzi, D. Rogolino, J. Organomet. Chem. 694 (2009)
3200.
[35] K. Ahlford, M. Livendahl, H. Adolfsson, Tetrahedron Lett. 50 (2009) 6321.
Acknowledgements
The Swedish Research Council, the Knut and Alice Wallenberg
foundation and the Carl Trygger Foundation are gratefully acknowl-
edged for financial support.