Modular syntheses of oxazolinylamine ligands and characterization of group 10 metal complexes

Article abstract of DOI:10.1139/V06-188

The syntheses of aminoalkyloxazoline and pyrrolidinyloxazoline ligands, each of which bear a pair of chiral centres, by both known and new routes are reported. Variable temperature NMR studies show that the known stepwise syntheses of the pyrrolidinyl compounds are not complicated by epimerization; however, coordination of one of the aminoalkyl derivatives to Pt(II) under conditions of prolonged heating to 80°C does give mixtures of diastereomeric N,N′-chelated complexes that result from inversion of the chiral centre associated with the aminoalkyl fragment. A new synthesis of pyrrolidinyloxazoline ligands that involves the Zn-catalyzed cyclization of Cbz-protected 2-cyanopyrrolidine and β-amino alcohols is also reported. This procedure offers the advantages of economy, shorter time, and fewer purification steps over the previously reported synthesis. In addition, the crystal structure of an enantiopure Pd(II) complex of an N.N′-chelated pyrrolidinyloxazoline is disclosed. This compound has a pseudo-C₂ axis of symmetry, which may make it suitable for asymmetric catalytic applications.

Full text of DOI:10.1139/V06-188

85
Modular syntheses of oxazolinylamine ligands and
characterization of group 10 metal complexes
Christine A. Caputo, Florentino d.S. Carneiro, Michael C. Jennings, and Nathan
D. Jones
Abstract: The syntheses of aminoalkyloxazoline and pyrrolidinyloxazoline ligands, each of which bear a pair of chiral
centres, by both known and new routes are reported. Variable temperature NMR studies show that the known stepwise
syntheses of the pyrrolidinyl compounds are not complicated by epimerization; however, coordination of one of the
aminoalkyl derivatives to Pt(II) under conditions of prolonged heating to 80 °C does give mixtures of diastereomeric
N,N′-chelated complexes that result from inversion of the chiral centre associated with the aminoalkyl fragment. A new
synthesis of pyrrolidinyloxazoline ligands that involves the Zn-catalyzed cyclization of Cbz-protected 2-
cyanopyrrolidine and β-amino alcohols is also reported. This procedure offers the advantages of economy, shorter time,
and fewer purification steps over the previously reported synthesis. In addition, the crystal structure of an enantiopure
Pd(II) complex of an N,N′-chelated pyrrolidinyloxazoline is disclosed. This compound has a pseudo-C2 axis of symme-
try, which may make it suitable for asymmetric catalytic applications.
Key words: chiral ligands, ligand design, oxazolines, variable temperature NMR spectroscopy, asymmetric catalysis,
coordination compounds, palladium, platinum
Résumé : Faisant appel à des méthodes connues et d’autres nouvelles, on a réalisé la synthèse de ligands aminoalky-
loxazolines et pyrrolidinyloxazolines qui portent une paire de centres chiraux. Des études de RMN à température va-
riable ont permis de montrer que la méthode synthèse connue par étape pour les composés pyrrolidinyles ne sont pas
compliquées par l’épimérisation; toutefois, la coordination d’un des dérivés aminoalkyles du Pt(II), dans des conditions
de chauffage prolongé à 80 °C, ne conduit pas à des mélanges de complexes N,N′-chélatés diastéréomères résultant de
l’inversion du centre chiral associé au fragment aminoalkyle. On a aussi réalisé une nouvelle synthèse des ligands pyr-
rolidinyloxazolines qui implique une cyclisation catalysée par le zinc de β-amino-alcools sur la 2-cyanopyrrolidine por-
tant un groupe protecteur Cbz. Par rapport aux méthodes antérieures, cette méthode présente l’avantage d’être plus
économique, d’impliquer des temps de réaction plus courts et moins d’étapes de purification. De plus, on a déterminé
la structure cristalline du complexe énantiopur du Pd(II) avec une pyrrolidinyloxazoline N,N′-chélatée. Ce produit pré-
sente un axe de symétrie pseudo-C2 qui pourrait le rendre susceptible d’être utilisé dans des applications de catalyse
asymétrique.
Mots clés : ligands chiraux, développement de ligands, oxazolines, spectroscopie RMN à température variable, catalyse
asymétrique, composés de coordination, palladium, platine.
[Traduit par la Rédaction] Caputo et al. 95
Introduction
that have been elaborated with aminoalkyl functional groups
predominantly in the 2-position (e.g., Chart 1, I) (4) have
come to the fore as an important subclass of ligands in their
own right and as building blocks for more elaborate ligand
systems (Chart 1, II–IV) (5–7). Less common variations
bearing aminomethyl groups in the 4-position have also ap-
peared (Chart 1, XIV) (8).
The advantages of the ligand family I have been docu-
mented in a recent report by Rajaram and Sigman (4): the
compounds are conveniently made in a modular manner
from readily available, optically pure starting materials, bear
multiple independently variable stereocentres, and are ame-
nable to further elaboration through manipulation of the
amine (e.g., to give II–IV) (5–7, 9). We also point out that
the presence of NH protons may have important ramifica-
tions for the use of these ligands in the asymmetric hydroge-
nation of C=O bonds. Variations on this theme have also
included oxazolines incorporating pyrrolidine (V) (10),
quinoline (VI) (11), aniline (VII, XIII) (12–14),
Beginning in the late 1980s and continuing to the present
day, the oxazoline (4,5-dihydrooxazole) ring has featured
prominently as the chiral component of a broad range of lig-
ands that have been used to great effect in asymmetric cata-
lytic transformations (1, 2). In particular, the C2-symmetric
bis(oxazoline) class has seen widespread and very successful
application, e.g., in cycloaddition, aldol, Michael, carbonyl-
ene, and related Lewis acid-catalyzed reactions (3). More
recently, C1-symmetric compounds bearing oxazoline rings
Received 4 October 2006. Published on the NRC Research
Press Web site at http://canjchem.nrc.ca on 20 February 2007.
C.A. Caputo, F.d.S. Carneiro, M.C. Jennings, and
N.D. Jones.1 Department of Chemistry, The University of
Western Ontario, London, ON N6A 5B7, Canada.
1Corresponding author (e-mail: njones26@uwo.ca).
Can. J. Chem. 85: 85–95 (2007)
doi:10.1139/V06-188
© 2007 NRC Canada

86
Can. J. Chem. Vol. 85, 2007
Chart 1. Examples of oxazolinylamine ligands.
Scheme 1. General synthesis of pyrrolidinyloxazolines by the Zn-catalyzed cyclization of N-protected 2-cyanopyrrolidine and β-amino
alcohols.
azabicyclo[2.2.1]heptane (IX) (15), pyrrole (XI) (16–18),
and pyridine (not shown) ring systems (21).
from the reaction of the appropriate ligand and PtCl2(SMe2)2
in boiling dichloroethane.
To date, access to the pyrrolidinyloxazoline ligands V has
been limited to a stepwise synthetic route outlined by
McManus et al. (10). Herein, we propose an alternate ap-
proach that involves the Zn-catalyzed cyclization of Cbz-
protected 2-cyanopyrrolidine and β-amino alcohols
(Scheme 1). This new approach is quicker, more economi-
cal, and requires fewer purifications. It also puts the diversi-
fication step later in the synthesis, which allows expansion
of ligand libraries more easily.
Ligands in classes I and V bear two stereocentres and are,
in principle, susceptible to epimerization. Through careful
stepwise synthesis, we show in this report that there is no
epimerization of V either during ligand synthesis by the
McManus route or during metal complexation. However,
epimerization does occur during the coordination of a ligand
of type I to Pt(II); two diastereomers of 1 (Chart 2) result
The compounds shown in Chart 1 have been used as lig-
ands in asymmetric allylation (II, XIII) (5,13), alkylation
(IV) (7,9), and crotylation (XIII) of aldehydes (13), Diels-
Alder (XI) (16), hetero-Diels-Alder (III, XI) (6), and trans-
fer hydrogenation (V, VI VII, VIII and IX) reactions (10–
12,14,15); phosphine derivatives (X, XII) (19,20) have also
been made and applied successfully in asymmetric hydro-
genations.
Catalysts for these reactions have typically been generated
in situ. Therefore, well-characterized examples of N,N′-chelated
oxazolinylamine metal complexes are rare. Those N,N′-
chelated complexes that have been structurally characterized
are for the most part confined to the pyridinyloxazolines
(21) and to the adducts of VII (12) and XI (16). As part of
our ongoing exploration of modularly-made chiral N,N′-che-
lating ligands, we also present in this work the synthesis and
© 2007 NRC Canada

Caputo et al.
87
Chart 2. Group 10 metal complexes of the oxazolinylamines re-
ported in this work.
High resolution mass spectrometry data were recorded us-
ing a Finnigan MAT 8200 instrument. IR spectra were re-
corded using a Bruker Vector 33 FT-IR spectrometer;
samples were run either neat between KCl plates or prepared
as KBr pellets. Data are reported in cm–1; peak intensities
are given as follows: strong (s), medium (m), weak (w), very
(v).
Syntheses of oxazolinylamine ligands
1-{(4S)-4-Benzyl-4,5-dihydro-oxazol-2-yl}-(2S)-2-methyl-
propylamine-carbamic acid ester 9H-fluoren-9-ylmethyl (4)
The title compound was made according to the method of
Rajaram and Sigman (4). H NMR (DMSO-d6, 353 K) δ:
0.89 (pt, 6H, CH(CH3)2), 2.00 (m, 1H, CH(CH3)2), 2.70 (dd,
characterization of group 10 metal complexes of ligands of
type I and V, including an X-ray structural determination of
a Pd(II) complex of V (Chart 2, 3).
1
2
3
1H, CHH′Ph, JHH = 13.8, JHH = 6.8), 2.85 (dd, 1H,
2
3
CHH′Ph, JHH = 13.6, JHH = 6.4), 3.93 (pt, 1H, ox
OCHH′), 4.03 (pt, 1H, ox OCHH′), 4.23–4.36 (m, 5H,
Fmoc CH, Fmoc CH2, NHCH, ox NCH), 7.18 (m, 1H, Ph),
7.22–7.26 (m, 4H, Ph), 7.32 (m, 2H, Fmoc Ar), 7.39 (t, 2H,
Experimental section
General considerations
Reagents were obtained from commercial sources and
used as supplied unless otherwise indicated. These were of
American Chemical Society (ACS) grade or finer. Solvents
were dried and deoxygenated either by N2 purge followed by
passage through alumina columns (Innovative Technology or
MBraun solvent purification systems) or by distillation un-
der N2 from the appropriate drying agent. All reactions were
carried out under N2 atmosphere using standard Schlenk
techniques unless stated otherwise. Amino alcohols were
prepared by the reduction of corresponding amino acids us-
ing a standard procedure (22). The metal complexes, trans-
PdCl2(PhCN)2 and PtCl2(SMe2)2, were prepared according
to literature procedures (23, 24).
3
Fmoc Ar, JHH = 7.6), 7.72 (pt, 2H, Fmoc Ar), 7.87 (d, 2H,
Fmoc Ar, JHH = 7.6). 13C{1H} NMR δ: 18.40, 19.15, 30.17,
3
41.15, 46.66, 54.81, 65.66, 66.53, 71.31, 120.07, 125.32,
125.36*, 126.09, 127.01, 127.61, 128.15, 129.25, 138.23,
140.69*, 143.76*, 143.85, 156.05, 165.55*. HRMS calcd.
for C29H30N2O3: 454.2256; found: 454.2268.
1-{(4S)-4-Benzyl-4,5-dihydro-oxazol-2-yl}-(2S)-2-methyl-
propylamine (5)
The synthesis by Rajaram and Sigman (4) was modified
to use Et2NH instead of piperidine. A 50 mL round-
bottomed flask was charged with 4 (0.303 g, 0.662 mol) and
MeOH (6 mL) was added to give a white suspension. Et2NH
(5 mL) was added over a period of 10 min at RT. The result-
ing pale beige solution was stirred for 30 min. The solution
was concentrated in vacuo and the residue was purified by
silica gel chromatography (MeOH–CH2Cl2 10%–20%). The
title compound was obtained as a beige oil. Yield 0.154 g
(97%). 1H NMR δ: 0.91 (d, 3H, CH(CH3)2, 3JHH = 6.8), 0.93
Thin-layer chromatography was performed using 250 µm
silica gel glass plates with fluorescent indicator (254 nm,
Rose Scientific, Edmonton, Alberta) and viewed by expo-
sure to UV light and (or) immersion in a staining solution
(ninhydrin, KMnO4, or phosphomolybdic acid) followed by
heating. Flash chromatography was performed using ACP
ultra pure silica gel, 60 Å, 230–400 mesh (Silicycle, Québec,
Quebec), using the eluent(s) specified. All solvent mixtures
are reported as volume percentages.
3
(d, 3H, CH(CH3)2, JHH = 6.8), 1.77 (br s, 2H, NH2), 1.91
3
(m, 2H, CH(CH3)2, JHH = 6.8), 2.62 (dd, 1H, CHH′Ph,
3
2
2JHH = 13.6, JHH = 8.4), 3.06 (dd, 1H, CHH′Ph, JHH
=
1H, 13C{1H}, 19F{1H}, correlated spectroscopy (COSY),
and heteronuclear multiple quantum coherence (HMQC)
NMR data were recorded on a 400 MHz Varian Mercury
3
3
13.6, JHH = 5.2), 3.28 (d, 1H, NH2CH, JHH = 5.2), 3.97
2
3
(dd, 1H, OCHH′, JHH = 8.8, JHH = 7.2), 4.16 (pt, 1H,
OCHH′), 4.36 (m, 1H, NCH), 7.16–7.29 (m, 5H, Ph).
13C{1H} NMR δ: 17.79, 19.58, 32.38, 42.02, 55.77, 67.17,
72.10, 126.74, 128.75, 129.10, 129.17, 129.30, 129.49.
HRMS calcd. for C14H21N2O: 233.1660; found: 233.1648.
1
(400.085 MHz for H, 100.602 MHz for 13C, 376.458 MHz
for 19F) or
a 400 MHz Varian Inova spectrometer
1
(399.762 MHz for H, 100.520 MHz for 13C). Unless other-
wise indicated, spectra were recorded at room temperature
(RT) in a CDCl3 solution using residual solvent proton (rela-
tive to external SiMe4, δ 0.00) or solvent carbon (relative to
external SiMe4, δ 0.00) as an internal reference. Downfield
Syntheses of Fmoc- and Cbz-protected
pyrrolidinyloxazolines
1
shifts were taken as positive. Data for the H NMR spectra
Compounds 7 and 9 were made according to modifica-
tions to the methods reported by McManus et al. (10), who
developed them for the analogous Cbz-protected pyrrolidine
derivatives 8 and 10. Room temperature spectroscopic data
for 8 and 10 matched those given in the literature. However,
high-temperature NMR spectroscopic data were acquired by
us to ensure that the reported splitting and broadening of
peaks was in fact due to E and Z isomers arising from the
amide and (or) carbamate (Cbz) bonds and not instead to the
presence of diastereomers resulting from epimerization. This
are reported as follows: chemical shift (δ), multiplicity, inte-
gration, assignment, and coupling constant(s). All coupling
constants are reported in Hertz (Hz) and the spin multiplici-
ties are indicated as follows: singlet (s), triplet (t), quartet
(q), multiplet (m), pseudo (p), and broad (br). Peaks due to
the minor geometrical (E or Z isomer) when present in 1H or
1
13C spectra are denoted by *. Fractional values for the H
NMR integrations indicate the relative concentrations of E
and Z isomers in solution.
© 2007 NRC Canada

88
Can. J. Chem. Vol. 85, 2007
determination was made based on the coalescence at high
temperature of the isopropyl methyl peaks from a multiplet
into two doublets of equal intensity and of the benzyl CH2
multiplet into a sharp singlet. Both sets of data (low and
high temperatures) are given for comparison.
Compounds 11 and 12 were prepared by deprotection of
the Cbz group using Pd–C and H2 gas as the hydrogen
source. This differed from the procedure of McManus et al.
(10), which used cyclohexene as the hydrogen source.
2.28 (br s, 1H, pyr NCHCHH′), 3.49–3.68 (m, 5H, CH2OH,
NHCH, pyr NCH2), 4.35 (m, 1H, pyr NCH), 5.15 (s, 2H,
Cbz OCH2), 7.28–7.36 (m, 5H, Cbz Ar).
(2S)-2-{(4S)-4-Isopropyl-4,5-dihydro-oxazol-2-yl}-
pyrrolidine-1-carboxylic acid 9H-fluoren-9-ylmethyl ester (9)
In a 100 mL Schlenk tube, a solution of 7 (0.384 g,
95.0 mmol) in dry CH2Cl2 (50 mL) was cooled to –78 °C.
Diethylaminosulfur trifluoride (DAST) (0.137 mL,
1.05 mmol) was added dropwise over 5 min and the resul-
tant pale yellow solution was stirred at –78 °C for 1 h,
warmed to RT over ~3 h, and stirred overnight. The mixture
was poured into satd. aq. NaHCO3 (70 mL). The aqueous
layer was discarded and the organic layer was washed with
H2O (2 × 50 mL). The aqueous layer was extracted with
CH2Cl2 (2 × 50 mL). Combined organic layers were washed
with brine (25 mL), dried over MgSO4, and concentrated in
vacuo. The residue was purified by silica gel chromatogra-
phy (EtOAc–hexanes 30%–50%) to afford the title com-
pound as an off-white foam. Yield 0.31 g (85%). 1H NMR δ:
(2S)-Chlorocarbamoyl-pyrrolidine-1-carboxylic acid 9H-
fluoren-9-ylmethyl ester (6)
Thionyl chloride (1.29 mL, 17.8 mmol) was added
dropwise over a period of 2 min to a 100 mL round-
bottomed flask charged with Fmoc-Pro-OH (3.38 g,
10.0 mmol) and dry toluene (30 mL). The mixture was
heated to dissolve all solids and held at reflux for 2 h. The
yellow solution was cooled to RT, then concentrated in
vacuo to afford a yellow oil, which crystallized when stored
at 5 °C. The product was used without further purification.
IR (neat) ν: 1793 (s, C=O COCl), 1707 (s, C=O Fmoc).
3
0.80 (d, 3H, CH(CH3)2, JHH = 6.8), 0.84 (d, 3H, CH(CH3)2,
3JHH = 6.8), 1.61 (m, 1H, CH(CH3)2), 1.91 (m, 3H, pyr
NCH2CH2, pyr NCHCHH′), 2.17 (m, 1H, pyr NCHCHH′),
3.83 (pq, 1H, Fmoc CH), 3.93 (pt, 1H, ox OCHH′), 4.17 (pt,
1H, ox OCHH′), 4.25 (m, 2H, pyr NCH2), 4.35 (m, 1H, ox
(2S)-{(1S)-Hydroxylmethyl-2-methyl-propylcarbamoyl}-
pyrrolidine-1-carboxylic acid 9H-fluoren-9-ylmethyl ester (7)
A solution of L-valinol (0.290 g, 2.40 mmol) and NEt3
(0.390 mL, 2.80 mmol) in dry CH2Cl2 (5 mL) was added
dropwise to a 100 mL Schlenk tube containing 6 (1.00 g,
2.80 mmol) and dry CH2Cl2 (30 mL) that had been cooled to
0 °C and flushed with N2. The mixture was stirred at 0 °C
for 20 min, warmed to RT, and stirred overnight. The mix-
ture was concentrated to ~15 mL and washed with satd. aq.
NH4Cl (3 × 50 mL), H2O (2 × 50 mL), and brine (50 mL).
The organic layer was dried over MgSO4 and concentrated
in vacuo. The residue was purified by silica gel chromatog-
raphy (MeOH–CH2Cl2 2%) to afford a nearly colourless oil.
3
NCH), 4.45 (br d, 1H, Fmoc OCH2, JHH = 8.8), 7.33 (pt,
2H, Fmoc Ar), 7.41 (pt, 2H, Fmoc Ar), 7.65 (d, 2H, Fmoc
3
3
Ar, JHH = 7.6), 7.86 (d, 2H, Fmoc Ar, JHH = 7.2). 13C{1H}
NMR δ: 18.12, 19.60*, 22.93*, 22.95, 28.14, 28.28*, 30.13,
31.76*, 46.49*, 46.60, 46.61, 47.25*, 55.61*, 55.62, 59.51,
59.76*, 61.30, 61.40*, 66.44, 66.99*, 120.11, 125.16,
125.31*, 125.69, 127.14*, 127.67, 127.68*, 128.21*,
128.90, 140.61*, 140.74, 143.39, 143.78*, 143.93, 143.94*
153.96, 171.65, 172.02*. HRMS calcd. for C25H28N2O3:
404.2100; found: 404.2107.
1
Yield 0.91 g (91%). H NMR δ: 0.87 (d, 3H, CH(CH3)2,
3
3JHH = 6.8), 0.89 (d, 3H, CH(CH3)2, JHH = 6.8), 1.81–2.02
(2S)-2-{(4S)-4-Isopropyl-4,5-dihydro-oxazol-2-yl}-
pyrrolidine-1-carboxylic acid benzyl ester (10)
The title compound was made according to the procedure
(m, 4H, CH(CH3)2, pyr NCH2CH2, pyr NCHCHH′), 2.16–
2.30 (m, 1H, pyr NCHCHH′), 3.41–3.66 (m, 6H, pyr NCH2,
CONHCH, CH2OH, OH), 4.21 (pt, 1H, Fmoc CH), 4.32–
4.45 (m, 4H, Fmoc OCH2, pyr NCH), 6.37* (br s, 0.3H,
1
reported by McManus et al. (10). H NMR (DMSO-d6, 298
3
K) δ: 0.73 (d, 3H, CH(CH3)2, JHH = 6.8), 0.80 (m, 3H,
3
CONH), 6.83 (d, 0.7H, CONH, JHH = 6.8), 7.30 (pt, 2H,
CH(CH3)2), 1.53–1.58 (m, 0.6H, CH(CH3)2), 1.59–1.66 (m,
0.4H, CH(CH3)2), 1.81–1.90 (m, 3H, pyr NCH2CH2, pyr
NCHCHH′), 2.10–2.20 (m, 1H, pyr NCHCHH′), 3.73–3.78
(m, 1.2H, pyr NCH2), 3.84–3.87 (m, 0.8H, pyr NCH2), 3.90
(pt, 0.6H, ox OCHH′), 3.98 (pt, 0.4H, ox OCHH′), 4.07 (pt,
0.6H, ox OCHH′), 4.16 (pt, 0.4H, ox OCHH′), 4.41 (dd,
Fmoc Ar), 7.39 (pt, 2H, Fmoc Ar), 7.57 (d, 2H, Fmoc Ar,
3
3JHH = 7.6), 7.75 (d, 2H, Fmoc Ar, JHH = 7.2). 13C{1H}
NMR δ: 18.79, 19.78, 24.86, 28.78*, 29.04, 47.33, 57.57,
61.01, 63.68, 68.00, 120.24, 125.21, 127.30, 128.00, 141.49,
143.91, 156.43, 172.56. HRMS calcd. for C25H31N2O4:
423.2292; found: 423.2278.
3
0.4H, pyr NCH, JHH = 3.2, 8.8), 4.46 (dd, 0.6H, pyr NCH,
3JHH = 2.8, 8.0), 4.98–5.12 (m, 2H, Cbz OCH2), 7.29–7.36
1
(2S)-{(1S)-Hydroxylmethyl-2-methyl-propylcarbamoyl}-
pyrrolidine-1-carboxylic acid benzyl ester (8)
(m, 5H, Cbz Ar). H NMR (DMSO-d6, 393 K) δ: 0.79 (d,
3
3
3H, CH(CH3)2, JHH = 6.8), 0.83 (d, 3H, CH(CH3)2, JHH
=
The title compound was made according to the procedure
6.8), 1.61 (m, 1H, CH(CH3)2), 1.89 (m, 3H, pyr NCH2CH2,
pyr NCHCHH′), 2.17 (m, 1H, pyr NCHCHH′), 3.83 (pq,
1H, pyr NCH2), 3.81 (pq, 1H, ox NCH), 3.92 (pt, 1H, ox
OCHH′), 4.14 (m, 2H, ox OCHH′), 4.48 (pd, 1H, pyr
NCH), 5.08 (s, 2H, Cbz OCH2), 7.29–7.35 (m, 5H, Cbz Ar).
1
reported by McManus et al. (10). H NMR (298 K) δ: 0.84
3
(pd, 3H, CH(CH3)2), 0.87 (d, 3H, CH(CH3)2, JHH = 7.2),
1.80 (m, 1H, CH(CH3)2), 1.91 (m, 2H, pyr NCH2CH2), 2.17
(br s, 1H, pyr NCHCHH′), 2.32 (br s, 1H, pyr NCHCHH′),
3.46–3.66 (m, 5H, CH2OH, NHCH, pyr NCH2), 4.35 (m,
1H, pyr NCH), 5.06–5.20 (m, 2H, Cbz OCH2), 6.19 (s,
0.4H, CONH), 6.75 (s, 0.6H, CONH), 7.13–7.38 (m, 5H,
Syntheses of (4S)-4-isopropyl-2-pyrrolidin-(2S)-2-yl-4,5-
dihydrooxazole (11)
Cbz Ar). 1H NMR (353 K) δ: 0.84 (d, 3H, CH(CH3)2, 3JHH
=
3
6.8), 0.88 (d, 3H, CH(CH3)2, JHH = 6.8), 1.81 (m, 1H,
CH(CH3)2), 1.91 (m, 3H, pyr NCH2CH2, pyr NCHCHH′),
By deprotection of Fmoc
A 100 mL round-bottomed flask charged with 9 (0.310 g,
© 2007 NRC Canada

Caputo et al.
89
0.810 mmol) and MeOH (10 mL) was cooled to 0 °C before
Et2NH (5 mL) was added dropwise. The mixture was
warmed to RT with stirring over 1.5 h. Volatiles were re-
moved in vacuo. The residue was purified by silica gel chro-
matography (MeOH–CH2Cl2 4%–10%) to afford 9 as a
nearly colourless oil. Yield 0.30 g (55%). Spectroscopic data
were identical to those reported by McManus et al. (10).
(2S)-2-Carbamoyl-pyrrolidine-1-carboxylic acid tert-butyl
ester (14)
The title compound was made by the method of Wallén et
1
al. (25). Yield >99%. H NMR (DMSO-d6, 353 K) δ: 1.39
(m, 9H, O(CH3)3), 1.72–1.87 (m, 3H, NCH2CH2CH2,
NCH2CH2CHH′), 2.07 (m, 1H, NCH2CH2CHH′), 3.28–3.40
(m, 2H, NCH2) 4.03 (pdd, 1H, NCHCN). 13C{1H} NMR
(DMSO-d6) δ: 23.85, 24.54*, 28.68, 28.80*, 30.66*, 31.66,
46.06, 47.01, 47.22*, 59.99*, 60.20, 78.99, 79.08*, 154.00,
154.28*, 174.94*, 175.32. IR (neat) ν: 3406 (m, N-H), 1683
(s, CO amide), 1403 (m, C-N). HRMS calcd. for
C10H18O3N2: 214.1317; found: 214.1312.
By deprotection of Cbz
A solution of 10 (1.0 g, 3.16 mmol) in MeOH (30 mL)
was added slowly to a 100 mL Schlenk tube containing Pd–
C 10% w/w (0.30 g, 30 wt%.). The flask was purged with H2
and was stirred under H2 atmosphere at RT for 24 h. The
suspension was filtered through a pad of Celite that was sub-
sequently rinsed with MeOH (30 mL). The filtrate was con-
centrated in vacuo to give a pale yellow oil that was purified
by silica gel chromatography (MeOH–CH2Cl2 5%) to afford
11 as a colourless oil. Yield 0.35 g (61%). Spectroscopic
data were identical to those reported by McManus et al.
(10).
(2S)-2-Cyano-pyrrolidine-1-carboxylic acid benzyl ester (15)
A 250 mL round-bottomed flask was charged with 13
(4.37 g, 17.6 mmol) and dry THF (60 mL). The solution was
cooled to 0 °C and NEt3 (5.88 mL, 42.3 mmol) was added
dropwise. Against a flow of N2, trifluoroacetic anhydride
(TFAA) (2.94 mL, 21.1 mmol) was added dropwise and the
mixture was stirred for 1 h at 0 °C. The reaction was
quenched by the addition of H2O (25 mL) and all volatiles
were removed in vacuo. The residue was taken up in CH2Cl2
(50 mL) and H2O (20 mL) was added. The organic layer was
washed with satd. aq. NaHCO3 (20 mL) and brine (20 mL),
dried over MgSO4, and concentrated in vacuo. The resultant
yellow oil was used without further purification. Yield
Synthesis of (4S)-4-Benzyl-2-pyrrolidin-(2S)-2-yl-4,5-
dihydrooxazole (12)
A solution of 18 (0.990 g, 2.70 mmol) in MeOH (40 mL)
was deprotected according to the method described in the
immediately preceding section. Yield 0.19 g (30%). Spectro-
scopic data matched those reported by McManus et al. (10).
1
2.44 g (83%). H NMR δ: 2.04–2.31 (m, 4H, NCH2CH2,
NCH2CH2CH2), 3.38–3.44 (m, 1H, NCH2), 3.55–3.62 (m,
1H, NCH2), 4.54–4.63 (m, 1H, NCHCN), 5.14–5.20 (m, 2H,
1
OCH2), 7.26–7.43 (m, 5H, Cbz Ar). H NMR (333 K) δ:
Synthesis of cyanopyrrolidine derivatives
2.01–2.30 (m, 4H, NCH2CH2), 3.42–3.48 (m, 1H, NCH2),
3.56–3.61 (m, 1H, NCH2), 4.58 (br s, 1H, NCHCN), 5.20 (s,
2H, OCH2), 7.32–7.40 (m, 5H, Cbz Ar). 13C{1H} NMR δ:
23.68, 24.58*, 30.72*, 31.69, 45.86*, 86.26, 46.92, 47.45*,
67.57*, 67.75, 118.63*, 118.81, 128.08, 128.22, 128.50,
135.80*, 135.99, 153.54*, 154.24. HRMS calcd. for
C13H14N2O2: 230.1055; found: 230.1061.
Our syntheses of 13 and 15 are modifications of those re-
ported by Wallén et al. (25) for the analogous Boc-protected
pyrrolidines 14 and 16. The compounds were previously
made by a different route by Cobb et al. (26, 27); they are
now commercially available but are very expensive. To the
1
best of our knowledge, only RT H NMR characterization
data have been reported in the open literature.
(2S)-2-Cyano-pyrrolidine-1-carboxylic acid tert-butyl ester
(16)
(2S)-2-Carbamoyl-pyrrolidine-1-carboxylic acid benzyl
ester (13)
The title compound was made using the method of Wallén
et al. (25). Yield 82%. 1H NMR δ: 1.46–1.50 (m, 9H,
C(CH3)3), 2.01–2.24 (m, 4H, NCH2CH2, NCH2CH2CH2),
3.30–3.39 (m, 1H, NCHH′), 3.42–3.55 (m, 1H, NCHH′),
Ethyl chloroformate (3.99 mL, 41.7 mmol) was added
dropwise to a cooled solution (–5 °C) of Cbz-Pro-OH
(4.60 g, 18.5 mmol), NEt3 (2.57 mL, 18.5 mmol), and THF
(60 mL) in a 100 mL round-bottomed flask. The resulting
slurry was stirred for 20 min at –5 °C and satd. aq. NH4OH
(5 mL) was then added dropwise. The solution was stirred at
RT for 18 h. Volatiles were removed in vacuo and the resi-
due was taken up in EtOAc (60 mL). The undissolved white
precipitate was filtered off and rinsed with EtOAc (30 mL).
The organic layer was dried over MgSO4 and concentrated
in vacuo to give an off-white amorphous solid, which was
1
4.42 (pdd, 0.5H, NCH), 4.55 (pd, 0.5H, NCH). H NMR
(DMSO-d6, 353 K) δ: 1.46 (s, 9H, C(CH3)3), 1.94 (m, 2H,
NCH2CH2), 2.12–2.25 (m, 2H, NCH2CH2CH2), 3.34 (m, 2H,
3
NCH2), 4.60 (dd, 1H, NCH, JHH = 3.6, 8.0). 13C{1H} NMR
δ: 24.13, 25.01*, 28.62, 30.88*, 31.71, 46.37, 46.67*, 47.41*,
47.55, 80.65, 120.53, 153.16, 153.83*. IR (neat, cm–1) ν:
2980 (m, C-H), 2240 (vw, CϵN), 1701 (vs, C=O). HRMS
calcd. for C10H16O2N2: 196.1212; found: 196.1213.
1
used without further purification. Yield 4.59 g (>99%). H
NMR δ: 1.87–1.99 (m, 2H, NCH2CH2), 2.16 (br s, 1H,
NCHCHH′), 2.31 (br s, 1H, NCHCHH′), 3.44–3.54 (m, 2H,
NCH2), 4.28–4.35 (m, 1H, NCH), 5.10–5.18 (m, 2H,
OCH2Ar), 5.74* (br s, 0.5H, NH2), 5.84 (br s, 0.5H, NH2),
6.06* (br s, 0.5H, NH2), 6.72 (br s, 0.5H, NH2), 7.34 (m,
5H, Cbz Ar). 13C{1H} NMR δ: 23.39*, 24.28, 28.73, 30.97*,
46.78, 47.23*, 60.00, 60.38*, 67.04, 127.62, 127.87, 128.30,
136.20, 154.67*, 155.65, 174.63, 175.50*. HRMS calcd. for
C13H16N2O3: 248.1161; found: 248.1169.
(2S)-2-Cyano-pyrrolidine·TFA salt (17)
Via an addition funnel, TFA (34 mL) was added dropwise
over a period of 10 min to a 100 mL round-bottomed flask
containing a cooled (0 °C) CH2Cl2 (124 mL) solution of 16
(2.44 g, 12.4 mmol). The solution was stirred for 2 h at
0 °C, then concentrated in vacuo to give an off-white oil.
Et2O (30 mL) was added slowly to produce a white crystal-
line solid. The slurry was filtered, and the solid rinsed with
Et2O (10 mL) before being dried under vacuum to obtain the
© 2007 NRC Canada

90
Can. J. Chem. Vol. 85, 2007
title compound as a white crystalline solid. Yield 1.10 g
(56%), mp 92–94 °C. H NMR (CD3OD) δ: 2.07–2.33 (m,
X-ray diffraction analysis were grown by slow diffusion of
iPr2O into a concentrated solution of 1 in CH2Cl2.
1
3H, NHCH2CH2, NHCHCHH′), 2.43–2.52 (m, 1H,
NHCHCHH′), 3.35–3.47 (m, 2H, NHCH2), 4.68 (pt, 1H,
NHCH), 4.92 (s, 1H, NH). 13C{1H} NMR δ: 23.15, 29.65,
45.14, 45.97, 116.53. 19F{1H} NMR δ: –77.37 (s). IR (neat)
ν: 2261 (vw, CϵN), 1681 (vs, C=O). HRMS calcd. for
Dichloro{(4S)-4-isopropyl-2-pyrrolidin-(2S)-2-yl-4,5-
dihydro-oxazole}platinum(II) (2)
A 50 mL round-bottomed flask was charged with PtCl2(SMe2)2
(0.227 g, 0.580 mmol), 11 (0.107 g, 0.580 mmol), and 1,2-
dichloroethane (20 mL). The mixture was heated to reflux
and held for 4 h. The suspension was filtered and rinsed with
Et2O (10 mL). The yellow solid was dried under vacuum.
–
C5H9N2: 97.0766; found 97.0760; calcd. for C2F3O2 :
112.9850; found: 112.9683.
1
Yield 0.11 g (42%). H NMR (DMSO-d6) δ: 0.74 (d, 3H,
Syntheses of pyrrolidinyloxazolines via 2-
cyanopyrrolidines
3
3
CH(CH3)2, JHH = 6.8), 0.79 (d, 3H, CH(CH3)2, JHH = 6.8),
1.68 (m, 1H, NHCH2CHH′), 1.91 (m, 1H, NHCHCHH′),
2.00 (m, 1H, NHCH2CHH′), 2.17 (m, 1H, NHCHCHH′),
2.47 (m, 1H, CH(CH3)2), 2.99 (m, 1H, NHCHH′), 3.35 (m,
1H, NHCHH′), 4.01 (m, 1H, NHCH), 4.27 (pq, 1H, NCH),
(2S)-2-{(4S)-4-Isopropyl-4,5-dihydro-oxazol-2-yl}-
pyrrolidine-1-carboxylic acid benzyl ester (10)
A solution of 15 (1.55 g, 6.70 mmol) and L-valinol
(1.08 g, 10.5 mmol) in chlorobenzene (120 mL) was added
to a 250 mL round-bottomed flask containing ZnCl2 (1.43 g,
10.5 mmol). The mixture was heated to reflux for 24 h. Sol-
vent was removed in vacuo and the yellow residue was puri-
fied by silica gel chromatography (EtOAc–Hexanes 30%).
The title compound was obtained as a yellow oil. Yield
1.82 g (85%). Spectroscopic data for this compound
matched those reported by McManus et al. (10).
2
4.63 (t, 1H, OCHH′, JHH = 9.2), 4.77 (m, 1H, OCHH′,
2JHH = 9.2, JHH = 4.4), 7.29 (q, 1H, NH, JHH = 4.8). Anal.
calcd. for C10H18Cl2N2OPt: C 26.79, H 4.05, N 6.25; found:
C 26.62, H 4.04, N 6.12.
3
3
Dichloro{(4S)-4-isopropyl-2-pyrrolidin-(2S)-2-yl-4,5-
dihydro-oxazole}palladium(II) (3)
A solution of 11 (87.0 mg, 0.500 mmol) in CH2Cl2
(20 mL) was added to a 25 mL round-bottomed flask con-
taining trans-PdCl2(PhCN)2 (19.2 mg, 0.500 mmol). The re-
sultant orange solution was stirred for 3 h at RT. An orange
precipitate began to form within the first 5 min of stirring.
The solid was collected by filtration and rinsed with Et2O
(20 mL), then dried under vacuum. Yield 0.12 g (66%).
1H NMR (DMSO-d6) δ: 0.71 (pt, 6H, CH(CH3)2), 1.64 (m,
1H, NHCH2CHH′), 1.82–1.92 (m, 2H, NHCHCHH′,
NHCH2CHH′), 2.07 (m, 1H, NHCHCHH′), 2.39 (m, 1H,
CH(CH3)2), 2.84 (m, 1H, NHCHH′), 3.23 (m, 1H, NHCHH′),
3.86 (m, 1H, NCH), 4.16 (m, 1H, NHCH), 4.47 (pt, 1H,
OCHH′), 4.60 (pq, 1H, OCHH′), 6.58 (m, 1H, NH). Anal.
calcd. for C10H18Cl2N2OPd: C 33.40, H 5.05, N 7.79; found:
C 33.66, H 5.25, N 7.48. Crystals of the title compound suit-
able for X-ray diffraction analysis were grown by slow dif-
fusion of iPr2O into a concentrated solution of 3 in CH2Cl2.
(2S)-2-{(4S)-Benzyl-4,5-dihydro-oxazol-2-yl}-pyrrolidine-1-
carboxylic acid benzyl ester (18)
A similar procedure to that outlined earlier was carried
out using 15 (0.605 g, 2.63 mmol), L-phenylalaninol
(0.665 g, 4.40 mmol), and ZnCl2 (0.599 g, 4.40 mmol) to
obtain 18. Yield 0.99 g (quant.). Spectroscopic data matched
those reported by McManus et al. (10).
Syntheses of oxazolinylamine metal complexes
{1-((4S)-4-Benzyl-4,5-dihydro-oxazol-2-yl)-2-
methylpropylamine}dichloroplatinum(II) (1)
A 100 mL round-bottomed flask was charged with
PtCl2(SMe2)2 (0.084 g, 0.215 mmol), 5 (0.050 g, 0.215 mmol),
and 1,2-dichloroethane (20 mL). The mixture was heated to
reflux and held for 7 h. After cooling to RT, stirring was
continued overnight (18 h). Volatiles were removed in vacuo
and Et2O (10 mL) was added. The suspension was sonicated
for 5 min to produce a dark yellow powder, which was fil-
tered and rinsed with Et2O (10 mL). The yellow solid (iso-
lated as a mixture of diastereomeric metal complexes) was
Crystallography
Crystal data for 3 were collected at low temperature
(150 K) using a Nonius Kappa CCD area detector
diffractometer running COLLECT software (Nonius BV,
Delft, The Netherlands, 1997–2002). The unit cell parame-
ters were calculated and refined from the full data set. Crys-
tal cell refinement and data reduction were carried out using
HKL2000 DENZO-SMN (Otwinowski and Minor 1997),
and absorption correction was applied using HKL2000
DENZO-SMN (SCALEPACK). The SHELXTL PC V6.14
for Windows NT (Bruker AXS Inc., Madison, Wisconsin)
suite of programs was used to solve the structure by direct
methods. Subsequent Fourier difference syntheses allowed
the remaining atoms to be located. All non-hydrogen atoms
were refined with anisotropic thermal parameters. Hydrogen
atom positions were calculated geometrically and were in-
cluded as riding on their respective carbon atoms. The abso-
lute structure parameter for 3 was 0.00 (5), which indicated
that the correct diastereomer had been refined and that there
1
dried under vacuum. Yield 0.10 g (95%). H NMR δ: 0.83
3
3
(d, 3H, CH(CH3)2, JHH = 6.8), 0.95 (d, 3H, CH(CH3)2, JHH
3
= 6.8), 1.07* (d, 6H, CH(CH3)2, JHH = 6.8), 1.13* (d, 6H,
CH(CH3)2, JHH = 6.8), 2.19 (m, 1H, CH(CH3)2), 2.34* (m,
1H, CH(CH3)2), 2.68* (dd, 1H, JHH = 13.2, JHH = 10.4,
3
2
3
2
3
CHH′Ph), 3.01 (dd, 2H, CHH′Ph, JHH = 13.6, JHH = 8.4),
2
3
3.38 (dd, 1H, CHH′Ph, JHH = 13.6, JHH = 2.6), 3.73* (dd,
2
3
1H, CHH′Ph, JHH = 13.2, JHH = 2.4), 4.03 (dd, 1H,
3
3
NH2CH, JHH = 10.4, JHH = 5.2), 4.36–4.63 (m, 6H,
NH2CH*, NCH, NCH*, OCHH′, OCHH′*, OCHH′*), 4.78
(pt, 1H, OCHH), 5.84 (t, 2H, NH2), 5.98* (t, 2H, NH2),
7.13–7.39 (m, 10H, Ph, Ph*). Anal. calcd. for
C14H20Cl2N2OPt: C 33.74, H 4.05, N 5.62; found: C 33.99,
H 4.18, N, 5.32. Crystals of the title compound suitable for
© 2007 NRC Canada

Caputo et al.
91
Table 1. Crystal data and refinement parameters for 3.
Empirical formula
C10H18Cl2N2OPd
Formula weight (g mol–1)
Temperature (K)
Wavelength (Å)
Crystal system
359.56
150(2)
0.710 73
Monoclinic
P21
Space group
a (Å)
b (Å)
c (Å)
6.8508(3)
10.0099(5)
9.7236(4)
90
100.676(2)
655.26(5)
2
1.822
1.805
360
α, γ (°)
β (°)
Volume
Z
Density (calcd., g cm–3)
Absorption coefficient (mm–1)
F(000)
Crystal size (mm)
Range for data collection (°)
Index ranges
0.47 × 0.20 × 0.17
2.95 to 27.47
–8 ≤ h ≤ 8,
–11 ≤ k ≤ 12,
–12 ≤ l ≤ 12
8168
Reflections collected
Independent reflections (Rint)
2781 (0.0460)
Absorption correction
Maximum and minimum transmission
Refinement method
Data, restraints, parameters
Goodness-of-fit on F2
Semi-empirical from equivalents
0.7430 and 0.4810
Full-matrix least-squares on F2
2781, 1, 146
1.119
Final R indices [I > 2σ(I)]
R indices (all data)
Absolute structure parameter
Largest difference in peak and hole (e Å–3)
R1 = 0.0293, wR2 = 0.0723
R1 = 0.0317, wR2 = 0.0733
0.00(5)
1.125 and –0.703
was no sign of racemic twinning. The crystal data and re-
but insoluble in nonpolar solvents such as diethyl ether and
hexanes.
The H NMR spectrum of the free ligand 5 was consis-
finement parameters for 3 are listed in Table 1.2
1
tent, as expected from previous work (4), with a single
diastereomer (e.g., the isopropyl methyl peaks appeared as
the expected two doublets at δ: 0.91, 0.93). However, the H
Results and discussion
1
Aminomethyl-substituted ligands and complexes
The oxazolinylamine ligand 5 was made according to
a minor modification of the method by Rajaram and Sigman
(4) (Scheme 2). The reaction between Fmoc-protected L-
valine and L-phenylalaninol using PPh3–CCl4 as the activator
and diisopropylethylamine (DIPEA) as the base for both the
initial amide bond formation and the subsequent ring closing
to install the oxazoline gave the protected precursor 4.
Deprotection by diethylamine gave the desired ligand 5 in
54% overall yield.
The Pt(II) complex 1 was made in 95% yield by the
equimolar reaction of 5 and PtCl2(SMe2)2 in a refluxing
dichloroethane solution (Scheme 3). This yellow compound
was soluble in moderately polar chlorinated organic solvents
NMR spectrum of the corresponding Pt(II) complex 1 was
dependent on the time allowed for the reaction between
PtCl2(SMe2)2 and 5. It showed a “doubling” of all of the
peaks (e.g., the isopropyl methyl peaks appeared as two
pairs of doublets of almost equal intensity at δ: 0.83, 0.95
and 1.07, 1.13) when the reaction time was longer than 18 h;
shorter reaction times gave simpler spectra in which dou-
bling was greatly diminished. This phenomenon was consis-
tent with the presence of a pair of diastereomeric metal
complexes, which was confirmed by X-ray crystallography.
Although the structure of 1 could not be refined sufficiently
well for publication, the basic atom connectivity could be
determined unequivocally. The unit cell contained equal
2 Supplementary data for this article are available on the journal Web site (http://canjchem.nrc.ca) or may be purchased from the Depository
of Unpublished Data, Document Delivery, CISTI, National Research Council Canada, Ottawa, ON K1A 0R6, Canada. DUD 5124. For more
information on obtaining material refer to http://cisti-icist.nrc-cnrc.gc.ca/irm/unpub_e.shtml. CCDC 633044 contains the crystallographic
data for this manuscript. These data can be obtained, free of charge, via http://www.ccdc.cam.ac.uk/conts/retrieving.html (Or from the Cam-
bridge Crystallographic Data Centre, 12 Union Road, Cambridge CB2 1EZ, UK; fax +44 1223 336033; or deposit@ccdc.cam.ac.uk).
© 2007 NRC Canada

92
Can. J. Chem. Vol. 85, 2007
Scheme 2. Synthesis of the oxazolinylamine ligand 5 by the method of Rajaram and Sigman. (i) PPh3–CCl4–DIPEA; (ii) Et2NH.
Scheme 3. Synthesis of 1.
Scheme 4. Synthesis of pyrrolidinyloxazolines by the method of McManus et al. (i) SOCl2, toluene, 5 h; (ii) NEt3, CH2Cl2; (iii) NEt3,
DAST, CH2Cl2, –78 °C; (iv) PG = Fmoc: Et2NH; PG = Cbz: H2, 10% Pd–C, MeOH.
numbers of both the expected S,S-diastereomer and the un-
expected S,R-diastereomer of 1. This showed that the
stereocentre associated with the aminoalkyl chain was sus-
ceptible to epimerization under the relatively harsh condi-
tions used for the coordination reaction. 1H NMR data
indicated that the epimerization could be minimized (<10%)
by simply reducing the reaction time to -8 h. This result is
important because catalysts incorporating ligands of this
type have typically been generated in situ and have therefore
generally not been well characterized; therefore, relation-
ships drawn between the absolute configuration of the active
catalyst and the observed ee in the product(s) of an asym-
metric reaction are necessarily of limited validity. A solution
of free ligand in dichloroethane was heated to reflux in the
absence of metal for 24 h and, despite some decomposition,
It is well-known that the E and Z isomers of amide bonds
can be seen by NMR spectroscopy at RT. These are ususally
evidenced by a coalescence of some (if not all) peaks at high
temperature because of rapid chemical exchange of E and Z
isomers. Given our experience with 1 (vide supra), we were
concerned that the doubling observed for all of the com-
pounds 7–11 may have resulted not only from the presence
of E and Z isomers but also from diastereomers arising at
some stage during ligand synthesis. Presumably, epimeri-
zation of the pyrrolidine stereocentre could occur via the in-
termediate acid chloride 6 or, less likely, during the DAST-
induced oxazoline ring-closing step to form 9 and 10. It has
been shown by Wipf and co-workers (28) that DAST is able
to induce this cyclization with less than 1.5% racemization
of sensitive substrates.
1
no epimerization was seen by H NMR spectroscopy.
McManus et al. did not report high-temperature NMR
data for their compounds. Therefore, we repeated the syn-
theses of the Cbz-protected compounds and examined both
them and our Fmoc-protected species closely by variable
temperature NMR spectroscopy. For the open chain com-
pounds 7 and 8, we concentrated on peaks due to the amide
NH proton, and for the oxazolines 9 and 10, on those due to
the isopropyl methyl protons. Figure 1a shows the amide re-
Pyrrolidine-substituted ligands
The pyrrolidinyloxazoline ligand 11 was made according
to Scheme 4 (PG = Fmoc). This general method has been
applied previously to the synthesis of 11 (and 12, not shown)
by McManus et al. (10), who instead used the Cbz-protecting
group. These workers noted that the NMR spectra of their
intermediates 8 and 10 were complicated by a doubling of
many of the peaks but did not provide a rationale for this
phenomenon.
1
gion of the H NMR spectrum of 8 at RT and at 155 °C;
1
Fig. 1b shows the isopropyl methyl region of the H NMR
spectrum of 10 at RT, and at 70 °C in DMSO-d6 solution.
© 2007 NRC Canada

Caputo et al.
93
Scheme 5. Synthesis of pyrrolidinyloxazolines by Zn-catalyzed cyclization of 2-cyanopyrrolidine and β-amino alcohols. (i) EtOC(O)Cl,
NEt3; (ii) NH4OH; (iii) TFAA, NEt3; (iv) TFA; (v) amino alcohol, ZnCl2, PhCl; (vi) H2, Pd–C.
1
Cbz (Scheme 4), the deprotection was by hydrogenolysis;
for PG = Fmoc, on the other hand, deprotection was
achieved by treatment with base. These two routes gave iso-
lated yields of 86% and 54%, respectively. The Fmoc
deprotection was complicated by the decomposition of some
Fig. 1. (a) The δ 6.1–6.9 range of the H NMR spectra of 8 at
155 °C (top) and at RT (bottom), and (b) the δ 0.6–0.9 range of
1
the H NMR spectra of 10 at 70 °C (top) and at RT (bottom) in
DMSO-d6 solution.
1
of the intended product into an open chain peptide. The H
NMR spectra of the final products arising from the Cbz and
Fmoc manifolds were identical and unequivocally consistent
with the presence of only a single diastereomer.
We investigated an alternate approach to the
pyrrolidinyloxazolines starting from protected (2S)-2-
cyanopyrrolidines (Scheme 5, 15 and 16), which were made
using a modification to the syntheses presented by the
groups of Wállen (25) and Ley (26, 27). Judicious choice of
protecting group was essential. When PG = Boc, there was
no reaction between the cyanopyrrolidine 16 and the amino
alcohol; moreover, given the acidic conditions, a putative
Boc-protected pyrrolidinyloxazoline would be difficult to
deprotect without destruction of the oxazoline ring. Cleav-
age of Boc prior to oxazoline formation gave the TFA salt
17, which in our hands could not be induced to react with
the amino alcohol either in the presence or absence of ZnCl2
and (or) base. The Fmoc protecting group was also problem-
atic because its corresponding cyanopyrrolidine (not shown)
decomposed under the reaction conditions. Finally, we found
that treatment of Cbz-protected (2S)-2-cyanopyrrolidine 15
with amino alcohols in the presence of ZnCl2 at elevated
temperature cleanly gave the intended Cbz-protected
pyrrolidinyloxazolines 10 and 18, which were easily
deprotected by hydrogenolysis to give the intended products
11 and 12.
This route had distinct advantages over the route devel-
oped by McManus et al. First, the Cbz-protected 2-
cyanopyrrolidine 15 could be made quickly and in high yield
(a few hours in ~83% overall yield); second, the diversity-
introducing step was placed later in the synthesis; third, the
overall cost was reduced by elimination of the expensive
DAST reagent; and finally, the number of required purifica-
tion steps was reduced (there was no need for purification
until after ZnCl2-mediated oxazoline formation).
Coalescence of the peaks due to NH proton in 8 (from
two broad singlets at δ: 6.18, 6.76 in the RT spectrum into a
single broad peak at δ: 6.58 at 155 °C) and of those due to
the isopropyl methyl protons in 10 (four doublets at δ: 0.73,
0.78, 0.79, 0.82 in the RT spectrum to two doublets at δ:
0.84, 0.88 at 70 °C) indicated that neither the acid chloride
nor the DAST cyclization step were implicated in
epimerization. The overall number, relative integrations, and
splittings of the peaks in the high-temperature spectra were
consistent with only a single diastereomer of the compound
under study. These experiments implied that the likely cause
of the doubling was not the presence of diastereomers but
rather of E and Z isomers resulting from restricted rotation
about the carbamate bond in 10 and the amide and (or)
carbamate bond in 8. The NMR spectra of our new Fmoc-
protected analogues 7 and 9 were also complicated by the
presence of E and Z isomers. These likewise simplified at
high temperature.
Metal complexes
Yields for the amide bond forming and cyclization steps
in the Cbz- and Fmoc-protected pathways were not signifi-
cantly different, being between 85% and 91%. However, dif-
ferences arose in the final deprotection step. When PG =
The Pd(II) complex 3 was made in 66% yield by the
equimolar reaction of trans-PdCl2(PhCN)2 and 11 in CH2Cl2
solution at RT. The analogous Pt(II) complex 2 was also
© 2007 NRC Canada

94
Can. J. Chem. Vol. 85, 2007
Table 2. Selected bond distances, bond an-
gles, and torsion angles for 3 with standard
deviations in parentheses.
Fig. 2. ORTEP representation of the molecular structure of 3 (el-
lipsoids drawn at 50% probability). Except for those associated
with chiral centres, H-atoms have been omitted for clarity. All of
the chiral centres have the S-configuration.
Bond distances (Å)
Pd–N(10)
2.008(4)
Pd–Cl(2)
Pd–N(1)
2.3000(12)
2.054(4)
Pd–Cl(1)
2.3040(13)
Bond angles (°)
N(10)–Pd–N(1)
N(1)–Pd–Cl(2)
N(1)–Pd–Cl(1)
N(10)–Pd–Cl(2)
N(10)–Pd–Cl(1)
Cl(2)–Pd–Cl(1)
Torsion angles (°)
Cl(2)–Pd–N(1)–C(2)
Pd–N(10)–C(9)–C(11)
81.84(17)
90.50(11)
176.39(12)
172.16(13)
94.57(13)
93.08(5)
58.7(3)
69.0(6)
made in 42% yield by equimolar reaction of PtCl2(SMe2)2
and 11 in dichloroethane at 80 °C.
The molecular structure of 3 is represented as a thermal
ellipsoid plot in Fig. 2; selected bond distances and angles
appear in Table 2. As expected, the coordination geometry
about the metal was essentially square planar. The N(1)–Pd–
N(10) and Cl(1)–Pd–Cl(2) angles were approximately 81.8°
and 93.1°, respectively. Although the Pd–N(10) (Pd–
oxazoline) distance (2.008(4) Å) was shorter than the Pd–
N(1) (Pd–pyrrolidine) distance (2.054(4) Å), the difference
was not reflected in a corresponding trans influence for the
chloride ligands: the Pd–Cl bond lengths were nearly identi-
cal and approximately 2.30 Å. The Pd–N(1) distance was in
the range of those typically observed for Pd(II) complexes of
pyrrolidines (2.05–2.11 Å). The mean plane of the oxazoline
ring was very nearly coincident with the mean coordination
plane containing the metal centre and its four ligated atoms;
these two planes intersected at an angle of 7.98°. The
pyrrolidine ring adopted an envelope configuration in which
C(3) rested 0.639 Å from the mean plane defined by C(2),
N(1), C(5), and C(4). This plane made an angle of 53.68°
with the mean plane of the oxazoline ring.
The absolute configurations of C(9) and C(5) were both S,
which was consistent with the expected configurations based
on the L-proline and L-valinol starting materials; no
epimerization occurred either during the ligand synthesis or
on coordination to the metal. These configurations allowed
the quadrant analysis typically applied to chiral C2-symmet-
ric complexes to be superimposed on this formally C1-sym-
metric compound. In the orientation shown in Fig. 2, the
steric bulk of the ligand occupied the top left and bottom
right quadrants; the near 180° C(11)–N(10)–N(1)–C(2) tor-
sion angle of 172.33°, together with the similar C(11)–C(9)–
N(10) and C(2)–N(1)–N(10) angles (111.67° and 127.76°,
respectively), further strengthened the pseudo-C2-symmetric
approach to thinking about this ligand.
heterocycles) are noninnocent and may participate in cata-
lytic cycles (30).
Conclusion
We have successfully characterized the first simple N,N′-
chelated Pt and Pd complexes of oxazolinylamine ligands of
type I and V. Our finding that type I ligands may epimerize
upon metal complexation illustrates the importance of
proper characterization of complexes prior to their use as
catalysts. We also highlight the importance of including
high-temperature NMR data in the characterization of the N-
protected intermediates to these ligands as proof of
stereochemical purity. We have applied a shorter and more
cost effective route to the synthesis of pyrrolidinyloxazoline
ligands via the ZnCl2-mediated cyclization of Cbz-protected
(2S)-2-cyanopyrrolidine with amino alcohols. This allows
for quicker and easier access to this important class of chiral
ligands. The application of ligands of type I and V in the
asymmetric allylic oxidation of cyclic alkenes is currently
under investigation.
Acknowledgments
We thank the Natural Sciences and Engineering Research
Council of Canada (NSERC) and the University of Western
Ontario for financial support of this work, and Professor
Robert Hudson for useful discussions.
References
1. H.A. McManus and P.J. Guiry. Chem. Rev. 104, 4151 (2004).
2. G. Helmchen and A. Pfaltz. Acc. Chem. Res. 3, 336 (2000).
3. J.S. Johnson and D.A. Evans. Acc. Chem. Res. 33, 325 (2000).
4. S. Rajaram and M.S. Sigman. Org. Lett. 4, 3399 (2002).
5. J.-Y. Lee, J.J. Miller, S.S. Hamilton, and M.S. Sigman. Org.
Lett. 7, 1837 (2005).
6. S. Rajaram and M.S. Sigman. Org. Lett. 7, 5473 (2005).
7. I.M. Pastor and H. Adolfsson. Tetrahedron Lett. 43, 1743 (2002).
8. G.J. Rowlands and J.A. Griffith. Synthesis, 19, 3446 (2005).
The solid-state structure of 3 featured weak intermole-
cular N–HÿCl interactions of 3.33 Å that were only very
slightly longer than the sum of the van der Waals radii of N
and Cl (3.30 Å); the N–H–Cl angle was 156.15°. These in-
teractions in the solid state point to the possibility that the
NH groups of coordinated pyrrolidine (as well as of
oxazolidine (29) and, presumably, related nitrogen
© 2007 NRC Canada

Caputo et al.
95
9. P. Västilä, I.M. Pastor, and H. Adolfsson. J. Org. Chem. 70,
2921 (2005).
20. P.G. Cozzi, N. Zimmermann, R. Hilgraf, S. Schaffner, and A.
Pfaltz. Adv. Synth. Catal. 343, 450 (2001).
10. H.A. McManus, S.M. Barry, P.G. Andersson, and P.J. Guiry.
21. D.W. Dodd, H.E. Toews, F.d.S. Carneiro, M.C. Jennings, and
N.D. Jones. Inorg. Chim. Acta, 359, 2850 (2006).
22. M.J. McKennon and A.I. Meyers. J. Org. Chem. 58, 3568
(1993).
23. F.R. Hartley. Organomet. Chem. Rev. A, 119 (1976).
24. J.D. Scott and R.J. Puddephatt. Organometallics, 2, 1643 (1983).
25. E.A.A. Wallén, J.A.M. Christiaans, M.M. Forsberg, J.I.
Venäläinen, P.T. Männistö, and J. Gynther. J. Med. Chem. 45,
4581 (2002).
Tetrahedron, 60, 3405 (2004).
11. Y.-B. Zhou, F.-Y. Tang, H.-D. Xu, X.-Y. Wu, J.-A. Ma, and Q.-
L. Zhou. Tetrahedron: Asymmetry, 13, 469 (2002).
12. J.A. Cabeza, I. da Silva, I. del Rio, R.A. Gossage, D. Miguel,
and M. Suárez. Dalton Trans. 2450 (2006).
13. H.A. McManus, P.G. Cozzi, and P.J. Guiry. Adv. Synth. Catal.
348, 551 (2006).
14. Y. Jiang, Q. Jiang, and X. Zhang. J. Am. Chem. Soc. 120,
3817 (1998).
15. A. Trifonova, K.E. Källström, and P.G. Andersson. Tetrahe-
dron, 60, 3393 (2004).
26. A.J.A. Cobb, D.M. Shaw, and S.V. Ley. Synlett, 558 (2004).
27. A.J.A. Cobb, D.M. Shaw, and S.V. Ley. Org. Biomol. Chem. 3,
84 (2005).
16. H. Brunner, F. Henning, and M. Zabel. Z. Anorg. Allg. Chem.
28. A.J. Phillips, Y. Uto, P. Wipf, M.J. Reno, and D.R. Williams.
630, 91 (2004).
Org. Lett. 2, 1165 (2000).
17. H. Brunner, B. Nuber, and T. Tracht. Tetrahedron: Asymmetry,
9, 3763 (1998).
29. S.A. Cardile, M.C. Jennings, and N.D. Jones. Dalton Trans.
4672 (2006).
18. H. Brunner and T. Tracht. Tetrahedron: Asymmetry, 9, 3773
(1998).
30. C.A. Sandoval, T. Ohkuma, K. Muniz, and R. Noyori. J. Am.
Chem. Soc. 125, 13490 (2003).
19. A. Trifonova, J.S. Diesen, and P. G. Andersson. Chem Eur. J.
12, 2318 (2006).
© 2007 NRC Canada