A new, modular approach towards 2-(1-hydroxyalkyl)oxazolines, effective bidentate chiral ligands

Article abstract of DOI:10.1016/j.jorganchem.2005.10.048

Addition of β-hydroxyisonitriles towards aldehydes gives rise to hydroxyalkyl oxazolines, an interesting class of bidentate chiral ligands. This concept allows a very easy ligand tuning and optimization of the ligand structure as illustrated in a first ex

Full text of DOI:10.1016/j.jorganchem.2005.10.048

Journal of Organometallic Chemistry 691 (2006) 2155–2158
www.elsevier.com/locate/jorganchem
Communication
A new, modular approach towards
2-(1-hydroxyalkyl)oxazolines, effective bidentate chiral ligands
*
Michael Bauer, Uli Kazmaier
Institut fu¨r Organische Chemie, Universita¨t des Saarlandes, Im Stadtwald, 66041 Saarbru¨cken, Germany
Received 30 August 2005; received in revised form 20 October 2005; accepted 20 October 2005
Available online 15 December 2005
Abstract
Addition of b-hydroxyisonitriles towards aldehydes gives rise to hydroxyalkyl oxazolines, an interesting class of bidentate chiral
ligands. This concept allows a very easy ligand tuning and optimization of the ligand structure as illustrated in a first example.
Ó 2005 Elsevier B.V. All rights reserved.
Keywords: Asymmetric catalysis; Addition; Isonitriles; Ligands; Ligand acceleration
Oxazolines are an important class of heterocyclic com-
pounds, not only because of their widespread incidence in
(mainly marine) natural products [1], but also as chiral
building blocks in organic synthesis. Especially Meyers
et al. [2] developed efficient chiral auxiliaries based on oxaz-
olines in the 1970s. But also in the Ôpost auxiliary eraÕ the
oxazolines enjoy great popularity. Taking into account that
in biological systems oxazolines are decisively involved in
the complexation of metal ions such as Cu2+ and Zn2+
[3], it is not surprising that they find also application as
building blocks for chiral ligands [4]. In most cases the
oxazolines were combined with a second coordinating
group giving rise to bidentate ligands. Besides bisoxazo-
lines [5], especially phosphine oxazolines [6] and phosphite
oxazolines [7] are applied in a wide range of catalytical pro-
cesses, which were discussed in detail in the reviews cited.
Phosphite oxazolines such as A and B (Fig. 1) give nice
results in Cu-catalyzed ZnEt2-additions towards enones
[8], in Pd-catalyzed allylic alkylations [9] as well as in cata-
lytic hydrogenations [10]. The same reactions were also
investigated with the amino derivative C [11]. All these
ligands can easily be obtained from hydroxy- or amin-
oalkyl oxazolines. In principle, these functionalized oxazo-
lines should be themselves good bidentate ligands. But
surprisingly, for this class of ligands only a few applications
are reported in the literature so far. While few reports
describe Et2Zn additions to aromatic aldehydes or imines
in the presence of ligands from type D [12], we are aware
of only one application of ligands of type E for the pheny-
lation of 4-chlorobenzaldehyde [13]. This caused our inter-
est in a straightforward synthetic approach towards these
hydroxyalkyl oxazolines (E), to use them either for the syn-
thesis of phosphite oxazolines (A) or directly as bidentate
ligands.
Most oxazoline syntheses are based either on a cycload-
dition of a metallated isonitrile with an aldehyde [14] or on
a condensation of a suitable carboxylic acid derivative with
a b-amino alcohol [2]. Via this last approach hydroxyalkyl
oxazolines E with two stereogenic centers can be obtained
from optically active a-hydroxy acids.
Unfortunately, this synthetic pathway requires relatively
drastic reaction conditions and gives moderate yields in
many cases [15]. This protocol is also limited with respect
to the substitution pattern because of the availability of
the chiral a-hydroxy acids. While methyl- and phenyl-
substituted derivatives can easily be obtained from lactic
or mandelic acid, other derivatives require more synthetic
efforts.
*
Corresponding author. Tel.: +49 681 302 3409; fax: +49 681 302 2409.
E-mail address: u.kazmaier@mx.uni-saarland.de (U. Kazmaier).
0022-328X/$ - see front matter Ó 2005 Elsevier B.V. All rights reserved.
doi:10.1016/j.jorganchem.2005.10.048

2156
M. Bauer, U. Kazmaier / Journal of Organometallic Chemistry 691 (2006) 2155–2158
HCOOEt
Δ, 2h
> 95%
1.1 eq. HMDS
H2N
OHCHN
OHCHN
TMSCl, CH2Cl2
92%
OH
OH
OSiMe3
3
1
2
1.0 eq. POCl3
2.5 eq. NEt3
.
0.1 eq. BF3 OEt2
CN
CN
MeOH, rt, 15 min
> 95%
CH2Cl2, 0°C, rt
84%
OSiMe3
OH
4
5a
Ph
O
CN
CN
CN
CN
OH
Fig. 1. Hydroxy- and phosphite oxazolines.
OH
OH
OH
5b
(62%)
5d
(67%)
5e
(55%)
5c
(65%)
To have more flexibility with respect to variations of the
substitution pattern, what should be indispensable for an
efficient ligand screening, we decided to develop a new syn-
thetic approach towards this class of ligands. In combina-
torial chemistry, multicomponent reactions are extremely
popular, and the isonitrile-based Passerini and Ugi reac-
tions play a dominant role [16]. This becomes understand-
able by having a close look onto the mechanism of these
reactions (Scheme 1). The first step is an activation of the
carbonyl group via protonation followed by the addition
of the isonitrile giving rise to acylimidoyl cation F. Addi-
tion of the carboxylate provides the intermediate G, which
then undergoes a Mumm rearrangement to yield the
expected product. By using a non-nucleophilic acid and
an isonitrile with an additional b-hydroxy functionality,
one might expect a cyclization of the corresponding inter-
mediate F0 directly to hydroxyalkyl oxazoline E.
Scheme 2. Synthesis of hydroxyisonitriles.
Starting from valinol 1 the formamide 2 was obtained by
refluxing 1 in ethyl formiate for 2 h in nearly quantitative
yield as a crystalline solid [17]. Attempts to get 5a directly
from the formamide provided the isonitrile in yields as low
as 20%. This problem could be solved by converting this
functionality into the corresponding silyl ether 3 employing
hexamethyl disilazane (HMDS), which then underwent a
clean isonitrile formation. These two steps could be per-
formed as a one-pot procedure as well. The silylated isoni-
trile 4 was deprotected in almost quantitative yield. In
analogy, a range of other isonitriles 5b–5e was obtained.
Their overall yields from the corresponding amino alcohols
are given in parentheses. Interestingly, protection of the
hydroxy group was not necessary with derivative 5d.
With isonitrile 5a in hand, we optimized the subsequent
oxazoline formation using various proton and Lewis acids.
The best results in the reaction with pivalaldehyde were
obtained in the presence of the relatively mild pyridinium
p-toluenesulfonate (PPTS), which gave rise to oxazoline
6a in 66% yield. By using isonitrile 5b the yield could be
increased to 81% (Scheme 3), even with the sterically
demanding pivalaldehyde. The results obtained with the
other isonitriles investigated are given in Fig. 2. It is worth
mentioning that there is no loss of enantiomeric purity
throughout the synthesis with respect to the stereocenter
resulting from the amino alcohol.
Therefore, we focused our synthetic efforts on the syn-
thesis of such hydroxyisonitriles which should be accessible
easily from amino alcohols. The synthesis is illustrated
exemplary for the isopropyl derivative 5a (Scheme 2).
R1
R2
R3
NC
O
R2
CHO
3
R
NH
O
R1
COOH
O
R2
Fortunately, we were able to separate the hydroxyalkyl
oxazolines, which were formed as 1:1 diastereomeric mix-
tures, easily via chromatography into the diastereomers.
The relative configurations for all compounds could be
assigned by comparing their chromatographic data to that
of 6a, for which an X-ray structure of the (S,S)-diastereo-
R2
+
N
R1
HO
O
R1
HO
N
O
R3
COO
R3
F
G
R2
R2
R2
CHO
R1
R1
H+
H+
O
HO
HO
10% PPTS
+
N
O
+
HO
OH
Et2O, rt, 16h
81%
CN
N
O
H
CN
OH
N
R1
OH
5b
6b
F'
E
Scheme 1. Passerini reaction and its modification for the synthesis of
hydroxyalkyl oxazolines.
Scheme 3. Synthesis of 2-(1-hydroxyalkyl)oxazoline 6b.

M. Bauer, U. Kazmaier / Journal of Organometallic Chemistry 691 (2006) 2155–2158
2157
generally high, except for the already mentioned valine
derivative 6a, what can be explained by the high volatility
of the ligand. Next, we varied the aldehyde component.
Isobutyraldehyde derived oxazoline 6f also appeared to
be quite volatile and could only be isolated in a poor yield,
whereas for 6g in contrast, the yield was high. All isomeric
mixtures could be separated by column chromatography,
giving rise to stereochemically pure samples.
O
N
O
N
O
N
Ph
HO
HO
HO
6d
6a
(66%)
6c
(85%)
(85%)
Trt
O
N
O
N
O
N
The matched ligands were investigated in the ZnEt2
addition (Table 2). As expected, the ligand with the steri-
cally demanding t-butyl group (6b) gave the best eeÕs, but
the influence of the oxazoline substituent was not dramatic
(entries 1–4). All other ligands gave eeÕs between 80% and
90%, and they all showed a comparable reactivity and com-
plete consumption of the aldehyde under the standard con-
ditions. A much stronger influence on the stereochemical
outcome of the reaction was observed for the exocyclic sub-
stituent. For example, replacing the t-butyl group by an
isopropyl group (6f) resulted in a selectivity drop of 66%
(Table 1, entry 2 vs. Table 2, entry 6). Interestingly, the
ligand with the lowest selectivity also showed the lowest
reactivity leading to an incomplete consumption of the
starting material after 18 h. A quarternary center at the
exocylic substituent thus seems to be crucial for both good
reactivity and good selectivity, probably for steric reasons.
The good result for 6g with the bulky trityl group supports
that thesis (entry 6).
In conclusion, we could show that a modification of the
Passerini reaction allows a straightforward and efficient
approach towards new chiral ligands. This concept allows
a very easy ligand tuning and optimization of the ligand
structure as illustrated in a first example. In the reaction
investigated a strong ligand acceleration was observed for
the matched ligand, but not the mismatched one, what
allows the application of the chiral ligand as diastereomeric
mixture. Kinetic studies as well as further applications of
these hydroxyalkyl oxazolines as chiral ligands are cur-
rently under investigation.
O
HO
HO
HO
6ga
6e
6f
(81%)
(98%)
(39%)
Fig. 2. Synthesis of hydroxyalkyl oxazolines 6. a Abbreviation: Trt =
triphenylmethyl.
mer was obtained. The stereoisomers were then subjected
to our test reaction. As a standard reaction for ligand eval-
uation we chose the addition of ZnEt2 towards benzalde-
hyde. In the reaction with ligand 6b we made an
interesting observation (Table 1). While the (R,S)-ligand
(entry 1) gave rise to the (R)-alcohol with moderate ee (mis-
matched case), the corresponding (S,S)-ligand (entry 2)
provided the (S)-configured alcohol with excellent ee
(matched case). Kinetic studies (GC) indicated that with
(S,S)-6b the reaction was almost complete after 30 min,
while with (R,S)-6b nearly no reaction was observed. This
caused us to use the diastereomeric mixture (obtained in
the oxazoline synthesis) directly as a ligand. And indeed,
the results obtained were comparable to those obtained
with pure (S,S)-6b, making this ligand concept highly
attractive.
To prove its generality we first reacted the prepared iso-
nitriles (Scheme 2) with pivalaldehyde to vary the substitu-
ent in the oxazoline moiety (Fig. 2). The yields were
Table 1
Et2Zn-Addition in the presence of ligand 6b
OH
CHO
General procedure for hydroxyalkyl oxazoline formation.
PPTS (20 mg) was added to a solution of the hydroxyisonit-
rile (1.0 mmol) and the corresponding aldehyde (1.5 mmol)
in ether (1 ml). The mixture was allowed to stir at room
temperature overnight, before it was subjected directly to
column chromatography (silica, eluent: hexanes/ether).
2 mol% 6b
Et2Zn
+
benzene/hexane
rt, 18 h
Entry
1
Liganda
Conversion
[%]
ee
Conf.
[%]b
88
100
100
23
94
92
(R)
O
N
HO
Table 2
(R,S)
Evaluation of hydroxyalkyl oxazolines 6 in the ZnEt2 addition
2
3
(S)
(S)
Entry
Liganda
Conv. (%)
ee (%)b
Configuration
O
N
1
2
3
4
5
6
a
6a
6c
6d
6e
6f
100
100
100
100
80
91
84
88
85
28
93
(S)
(S)
(S)
(R)
(S)
(S)
HO
HO
(S,S)
O
N
6g
100
(R/S,S)
Stereochemically pure samples employed. Relative configurations
a
Relative configurations assigned by analogy to the X-ray structure of
assigned by analogy to the X-ray structure of 6a.
b
6a.
Determined by GC (column: CP-Cyclodextrin-b).

2158
M. Bauer, U. Kazmaier / Journal of Organometallic Chemistry 691 (2006) 2155–2158
[7] Reviews: (a) J. Ansell, M. Wills, Chem. Soc. Rev. 31 (2002) 259–268;
(b) F. Agbossou-Niedercorn, I. Suisse, Coordin, Chem. Rev. 242
(2003) 145–158.
The separated diasteromeres were obtained as colorless
crystals or oils and should be stored under argon in the
refrigerator.
[8] I.H. Escher, A. Pfaltz, Tetrahedron 56 (2000) 2879–2888.
´ ˆ
[9] (a) R. Pretot, A. Pfaltz, Angew. Chem. 110 (1998) 337–339;
Acknowledgements
Angew. Chem. Int. End. 37 (1998) 323–325;
(b) G. Jones, C.J. Richards, Tetrahedron Lett. 42 (2001) 5553–5555.
[10] (a) J. Blankenstein, A. Pfaltz, Angew. Chem. 113 (2001) 4577–4579;
Angew. Chem. Int. Ed. 40 (2001) 4445–4447;
(b) F. Menges, A. Pfaltz, Adv. Synth. Catal. 344 (2002) 40–44.
[11] (a) G.P. Xu, S.R. Gilbertson, Tetrahedron Lett. 44 (2003) 953–955;
(b) C. Blanc, F. Agbossou-Niedercorn, G. Nowogrocki, Tetrahedron:
Asymmetr. 15 (2004) 2159–2163.
Financial support by the Deutsche Forschungsgemeins-
chaft as well as the Fonds der Chemischen Industrie is
gratefully acknowledged. We also thank Degussa AG for
gifts of amino alcohols.
References
[12] (a) J.V. Allen, J.M.J. Williams, Tetrahedron: Asymmetr. 5 (1994)
277–282;
(b) A.L. Braga, R.M. Rubim, H.S. Schrekker, L.A. Wessjohann,
M.W.G. de Bolster, G. Zeni, J.A. Sehnem, Tetrahedron: Asymmetr.
14 (2003) 3291–3295;
(c) G. Jones, C.J. Richards, Tetrahedron: Asymmetr. 15 (2004) 653–
664;
[1] Review: D.J. Faulkner, Nat. Prod. Rep. 18 (2001) 1–49, and earlier
reviews of this series.
[2] Review: A.I. Meyers, A.C. Mihelich, Angew. Chem. 88 (1976) 321–
332;
Angew. Chem., Int. Ed. Engl. 15 (1976) 270–281, and references cited
therein.
[3] D.J. Freeman, G. Pattenden, A.F. Drake, G. Siligardi, J. Chem.
Soc., Perkin Trans. 2 (1998) 129–135.
[4] Review: (a) Y. Langlois, C. Pouilhes, C. Kouklovsky, J.-F. Morelli,
A. Haudrechy, M. Kobayakawa, C. Andre-Barres, T. Berranger, O.
Dirat, Bull. Soc. Chim. Belg. 105 (1996) 639–657;
(b) C. Jonsson, K. Hallman, H. Andersson, G. Stemme, M. Malkoch,
E. Malmstrom, A. Hult, C. Moberg, Bioorg. Med. Chem. Lett. 12
(2002) 1857–1861.
[5] Reviews: (a) A. Pfaltz, Acta Chem. Scand. 50 (1996) 189–194;
(b) O.B. Sutcliffe, M.R. Bryce, Tetrahedron: Asymmetr. 14 (2003)
2297–2325;
(d) X.-M. Zhang, H.-L. Zhang, W.-Q. Lin, L.-Z. Gong, A.-Q. Mi,
X. Cui, Y.-Z. Jiang, K.-B. Yu, J. Org. Chem. 68 (2003) 4322–4329.
[13] C. Bolm, L. Zani, J. Rudolph, I. Schiffers, Synthesis (2004) 2173–
2180.
[14] (a) F. Gerhart, U. Scho¨llkopf, Tetrahedron Lett. 9 (1968) 6231–
6234;
(b) Y. Ito, M. Sawamura, E. Shirakawa, K. Hayashizaki, T.
Hayashi, Tetrahedron 44 (1988) 5253–5262;
(c) Y. Ito, M. Sawamura, M. Kobayashi, T. Hayashi, Tetrahedron
Lett. 29 (1988) 6321–6324.
[15] (a) D.K. Heldmann, D. Seebach, Helv. Chim. Acta 82 (1999) 1096–
1110;
(b) P. Muller, P. Nury, Helv. Chim. Acta 84 (2001) 662–677.
¨
(c) F. Glorius, M. Neuburger, A. Pfaltz, Helv. Chim. Acta (2001)
3178–3196.
[6] Reviews: (a) G. Helmchen, J. Organomet. Chem. 576 (1999) 203–214;
(b) G. Helmchen, A. Pfaltz, Acc. Chem. Res. 33 (2000) 336–345;
(c) A. Pfaltz, Chimia 55 (2001) 708–714.
[16] Review: A. Do¨mling, I. Ugi, Angew. Chem. 112 (2000) 3300–3344;
Angew. Chem. Int. Ed. 39 (2000) 3168–3210, and references cited
therein.
[17] M. Casey, M.P. Smyth, Synlett (2003) 102–106.