Unsymmetrical terdentate phosphorus-nitrogen-nitrogen (PNN) ligands: Effect of the M/L ratio and the pendant group on stereoselectivity

Article abstract of DOI:10.1016/S0957-4166(03)00361-6

Twelve novel, unsymmetrical terdentate pyrrolidine-based phosphorus-nitrogen ligands were prepared. These ligands promoted very high catalytic activity in the palladium-catalysed allylic substitution reactions of 1,3-diphenylpropenyl acetate with malonate

Full text of DOI:10.1016/S0957-4166(03)00361-6

TETRAHEDRON:
ASYMMETRY
Pergamon
Tetrahedron: Asymmetry 14 (2003) 2045–2052
Unsymmetrical terdentate phosphorus-nitrogen-nitrogen (PNN)
ligands: effect of the M/L ratio and the pendant group on
stereoselectivity
Xiaohui Cheng and King Kuok (Mimi) Hii*
Department of Chemistry, King’s College London, Strand WC2R 2LS, UK
Received 7 April 2003; accepted 29 April 2003
Abstract—Twelve novel, unsymmetrical terdentate pyrrolidine-based phosphorus-nitrogen ligands were prepared. These ligands
promoted very high catalytic activity in the palladium-catalysed allylic substitution reactions of 1,3-diphenylpropenyl acetate with
malonate esters (100% conversion at 0°C in <2 h). Enantioselectivities up to 94% were attained. An unusual effect of
metal-to-ligand ratio was observed. Various structural features affecting the enantioselectivity and stereoinduction were examined
and discussed. © 2003 Elsevier Ltd. All rights reserved.
1. Introduction
activity of a novel series of aminophosphine ligands 1–3
Fig. 1). In these examples, the pendant arm of the
(
The discovery of phosphinooxazolidine ligands that can
promote highly efficient and enantioselective catalysis
constituted an important breakthrough and has created
ligand (attached to the pyrrolidinyl nitrogen) was
modified, either by the introduction of different sub-
stituents on the nitrogen donor atom 2, or by the
introduction of an additional stereogenic centre 3.
Their reactivity is compared with a few structurally
related ligands.
1,2
significant impetus in the study of chiral non-C sym-
2
metrical mixed-donor ligands for asymmetric synthesis.
Previously, we reported a novel class of unsymmetrical,
proline-derived, terdentate phosphine ligands 1 and 2
(
R=Et) and their catalytic activity in the asymmetric
2. Results and discussion
2.1. Ligand syntheses
3
palladium-catalysed allylic alkylation reaction. Prior to
our study, there were very few reports of non-C sym-
2
metrical terdentate phosphine ligands in asymmetric
catalysis. To date, there are still only limited exam-
ples. Herein we report the preparation and catalytic
2
.1.1. N-Substitution. Ligands 2b and 2c were prepared
4
,5
by the direct N-alkylation of (S)-(−)-2-(diphenylphos-
6
phino)methyl-pyrroline 4 with the corresponding 2-
N,N-substituted-aminoethyl chloride or tosylate
(
Scheme 1, Method A). However, this procedure was
not always successful. An alternative procedure
Scheme 1, Method B) involves amide coupling of the
(
aminophosphine 4 with corresponding N,N-disubsti-
tuted glycines. The resultant amides 5 were reduced to
the triborane complexes 6, which were not isolated but
subjected to decomplexation in situ using a one-pot
3
procedure, to give the aminophosphines 2a and 2d.
This latter procedure also provided amidophosphine
Figure 1.
ligands 5a and 5d.
2
.1.2. Additional stereogenic centre. On the other hand,
enantiomerically pure 1-substituted-2-chloroethyl-N,N-
dimethylamine hydrochloride salts 7 may be derived
*
0
Corresponding author. Tel.: +44-20-7484-1183; fax: +44-20-
8482810; e-mail: mimi.hii@kcl.ac.uk
7
957-4166/$ - see front matter © 2003 Elsevier Ltd. All rights reserved.
doi:10.1016/S0957-4166(03)00361-6

2
046
X. Cheng, K. K. Hii / Tetrahedron: Asymmetry 14 (2003) 2045–2052
Scheme 3.
8
been reported independently by groups of Hiroi and
6
Tomioka.
2
.2. Catalytic reactions
All of the ligands prepared above were evaluated in the
palladium-catalysed allylic substitution reaction of 1,3-
diphenyl-2-propenyl acetate with dimethylmalonate
(
Scheme 4). Catalytic precursors were generated in situ
3
by mixing [Pd(h -C H )Cl] with the ligand in
3
5
2
dichloromethane prior to the introduction of the sub-
strates. The most popular reaction conditions were
initially adopted: 1.5 mol% of the p-allylpalladium pre-
cursor, with
3
equiv. of N,O-bis(trimethylsilyl)-
acetamide (BSA) and catalytic amount of potassium
acetate as bases.
Scheme 1. Synthesis of ligands 2.
from corresponding phenylalanine (both
D and L iso-
mers) and
L-valine through a series of functional group
7
transformations. These were used to prepare ligands
Scheme 4. Palladium-catalysed allylic alkylation reaction.
3
a–3c, incorporating a second stereogenic centre, via
Method A (Scheme 2).
2
.2.1. Metal-to-ligand ratio. We found that the stereose-
lectivity of the reaction was extraordinarily dependent
on the metal-to-ligand (M/L) ratio (Table 1). Adopting
the conventional ratio of 1:2, the enantioselectivity
achieved by ligand 2b was initially unpromising (entry
1
, 18%). However, a dramatic improvement was
observed when the amount of ligand used was reduced
entries 2–5). By using slightly less than 1 equiv. of the
(
ligand 2b (M/L=1:0.9) the optimal e.e. value of 82%
can be realised. This effect of the M/L ratio on the
enantioselectivity was also mirrored by ligands 2a
(
entries 8 and 9), 2c (entries 10 and 11) and 2d. In the
latter case, an enhancement as much as 60% was
observed by lowering the M/L ratio by just 0.1 (entries
1
4 and 15).
Scheme 2.
These observations led us to postulate that the coordi-
nation of these PNN ligands to palladium is very
sensitive to the metal-to-ligand ratio.
2.1.3. Amino and amido-phosphine (PN) ligands without
pendant arms. To assess the significance of the pendant
arm and donor group, ligands 8a and 8b were synthe-
sised from 1-naphthylmethylchloride and pivaloyl chlo-
ride via Methods A and B, respectively (Scheme 3).
Similar amido and amino phosphine ligands have also
Preliminary NMR experiments supported the above
conjecture. A solution containing the palladium com-
3
plex [(h -C H ) PdCl] and 2 equiv. (per palladium) of
3
5 3
2
ligand 2b gave rise to a complex mixture of resonances

X. Cheng, K. K. Hii / Tetrahedron: Asymmetry 14 (2003) 2045–2052
2047
Table 1. Asymmetric allylic substitution of 1,3-diphenyl-2-
2.2.2. Effect of the N-substituent and kinetic control.
The nature of the nitrogen pendant arm affects the
stereoselectivity of the reaction, and the following trend
was observed: 2a<2b<2c<2d (Table 1, entries 5, 9, 11
and 15), i.e. an increase in steric bulk of the N-sub-
stituent led to an enhancement in the enantioselectivity,
with the diphenyl-substituted ligand 2d affording the
highest e.e. (87%).
a
propenyl acetate by dimethyl malonate
Entry
Ligand
M:Lb
T (°C)
% e.e.c,d
1
2
3
4
5
6
7
8
9
2b
2b
2b
2b
2b
2b
2b
2a
2a
2c
2c
2c
2c
2d
2d
2d
2d
1:2
1:2
1:1
25
0
0
0
0
−20
−20
0
18
41
66
82
82
80
1:0.5
1:0.9
1:0.9
1:0.9
1:1
1:0.9
1:1
1:0.9
1:0.9
1:0.9
1:1
Optimal enantioselectivity was reached at 0°C. Lower-
ing the temperature further to −20°C did not appear to
affect the enantioselectivity (entries 5, 6, 11, 12, 15 and
e
89
54
70
74
86
86
0
0
0
1
6). We postulate that this is because the rate of the
10
11
12
13
14
15
16
17
formation of the active catalytic precursor has become
comparable with the rate of the stereo-defining step at
this temperature. To test this hypothesis, we performed
the reaction by adding BSA slowly via a syringe pump
over 8 h at −20°C (effectively decreasing the concentra-
tion of the nucleophile, and hence the rate of the
stereo-defining step). Indeed, subjecting the catalytic
reaction to such kinetic control led to e.e. enhance-
ments of roughly 10% (entries 7, 13 and 17)—with
ligand 2d, an excellent e.e. of 94% may be attained.
−20
−20
0
0
−20
−20
e
92
27
87
86
1:0.9
1:0.9
1:0.9
e
94
a
Reaction conditions are given in Section 4. All reactions were
complete (100% conversion, TLC analysis) in 16 h (unoptimised,
vide infra).
b
c
Metal-to-ligand ratio
Even though the slow addition has a beneficial effect on
the enantioselectivity, this was not applied in subse-
quent catalytic studies, due to the rather inconvenient
set-up.
Determined by chiral HPLC (Chiralpak AD column)
Absolute stereochemistry was found to be R, by comparison of the
specific rotation to literature values.
d
e
BSA added over a period of 8 h.
2
.2.3. The pendant donor group and the sense of stereoin-
3
1
in the P spectrum, including the resonance signal of
the uncoordinated phosphine ligand (lP −19.6 ppm).
When the experiment was repeated with equivalent
amounts of the palladium complex and 2b, only two
duction. If the pendant nitrogen donor group is absent,
as in ligands 8a and 8b, the enantioselectivity of the
allylic substitution reaction was found to be in favour
of the (R)-enantiomer, and is also dependent on the
nature of the pyrrolidinyl nitrogen substituent (Table 2,
entries 4 and 6), as was observed with analogous lig-
3
1
major P resonance peaks were observed (lP +34.3
1
and 32.6 ppm). H NMR spectrum exhibited a signifi-
cant downfield shift and broadening of the N-ethyl
resonances, suggesting the coordination of the pendant
nitrogen donor. In addition, we also observed a sharp
ABX spin system in the region corresponding to the
presence of an h -allyl fragment (l 5.5–6.5 ppm). Based
on these observations, we tentatively propose an equili-
brating mixture of h - and h -allyl complexes in the
stoichiometric mixture (Fig. 2), with the pyrrolidinyl
nitrogen donor acting as the hemilabile site. Similar
8
ands. However, what was not reported before was the
effect of the M/L ratio, which we found to be crucial in
maintaining high enantioselectivity (Table 2, entries
2
–4, 5 and 6). Once again, a slightly less than stoichio-
1
metric amount was essential for optimal e.e.s. With
ligand 8a, the sense of stereoinduction was dependent
on the M/L ratio (entries 2 and 4).
1
3
a
Table 2. Effect of the pendant donor
hemilabile behaviour has been previously proposed in
Entry
Ligand
M:L
T (°C)
% e.e.c,d
9
analogues C -symmetrical PNP ligands.
2
1
2
3
4
5
6
7
8
9
1
8a
8a
8a
8b
8b
9
5a
1:1b
1:2
1:1
1:0.9
1:1
1:0.9
1:0.9
1:0.9
1:0.9
1:0.9
1:0.9
1:0.9
0
0
0
0
0
0
0
0
0
0
0
0
80 (S)
11 (S)
20 (R)
74 (R)
13 (R)
46 (R)
18.5 (R)
36 (R)
22 (R)
47 (R)
72 (R)
42 (R)
Figure 2.
5d
A sub-stoichiometric amount of the ligand (with respect
to palladium) is therefore essential for the formation of
a single selective catalytic precursor. A slight excess will
generate highly competitive catalytic species that com-
promises the enantioselectivity of the process. This
effect has been previously noted for the Pd-catalysed
allylic substitution reaction in a system utilising phos-
10
11
12
3a (2S,2%S)
3b (2S,2%R)
3c (2S,2%S)
a
b
c
Reaction conditions as before.
M/L ratio has no effect on the e.e.
Determined by chiral HPLC (Chiralpak AD column)
By comparison of the specific rotation to literature values.
10
d
phine oxazoline ligands.

2
048
X. Cheng, K. K. Hii / Tetrahedron: Asymmetry 14 (2003) 2045–2052
The enantioselectivity attained by ligands 9, 5a and 5d
Table 2, entries 7–9) were much lower than their amine
(
analogues 8b, 2a and 2d (Table 2, entry 6; Table 1,
entries 9 and 15, respectively), thus providing further
evidence of the involvement of a centrally placed hemi-
labile site for the preservation of a well-defined coordi-
nation environment for stereo-discrimination.
Under the optimised M/L ratio (1:0.9), the sense of
stereoinduction remained the same in all these studies.
These observations were significant when compared to
the result reported with the PNP ligand 1 (Table 2,
entry 1), where the (S)-isomer was obtained in 80%
3
e.e. This is somewhat surprising, considering ligands 1
and 2d are identical except for the pendant donor group
(
N versus P).
All of the above observations suggest that the nature of
the pendant arm is at least as important as the stereo-
genic centre on the pyrrolidine ring, in determining the
stereochemical outcome and stereoinduction of these
allylic alkylation reactions.
Figure 4.
2.2.4. Additional stereogenic centre. Introduction of a
stereogenic centre next to the pendant dimethylamino
donor group led to interesting ‘match’ and ‘mismatch’
effects (Table 2, entries 10–12).
Comparison between the runs showed that the less
selective ligand 2b induced faster catalytic turnover
than ligand 2d. The optimal e.e. values were established
at the beginning and were maintained during the course
of the reactions.
The presence of a benzyl group with an (S)-configura-
tion resulted in lower enantioselectivity (entry 10, 47%
e.e.), but its presence in the (R)-configuration led to a
slight enhancement of the enantioselectivity (entry 11,
2
.2.6. Varying the malonate ester. Different malonate
esters were also evaluated as substrates (Table 3). The
reactivity and selectivity of the addition were essentially
unchanged by adopting the bulkier tert-butyl ester
7
2% e.e.). Changing the substituent from a benzyl 3a to
an isopropyl 3c group caused a drop in the enantiose-
lectivity (entry 12).
(
entries 1–4). However, enantioselectivity decreased
when methyl malonate ester was used (entries 5–8).
Ligand 2d achieved the highest enantioselectivities, 87
and 75%, respectively, in both cases (entries 4 and 8).
Interestingly, the sense of stereoinduction may be
reversed with ligand 10—the amide synthetic precursor
of ligand 3c. Containing amide functionality and a
second stereogenic centre (Fig. 3), this ligand induced a
small e.e. in favour of the S-isomer (compared to Table
2, entry 12).
a
Table 3. Reaction with different electrophiles
Figure 3.
Entry
Electrophile
Ligand
% e.e.
t
2
.2.5. Rate and stereochemical pathway. The rates and
1
2
3
4
5
6
7
8
CH(CO
2
Bu)
2
2a
2b
2c
2d
2a
2b
2c
2d
63 (R)
80 (R)
81 (R)
87 (R)
60 (S)
65 (S)
60 (S)
75 (S)
selectivities of the reactions catalysed by ligands 2b and
2
d were also compared. In this study, aliquots of reac-
tion mixtures were extracted periodically and analysed
CMe(CO
Et) b
2 2
1
for % conversion ( H NMR) and enantioselectivity
(
chiral HPLC). The result clearly showed that these
catalysts are not only highly selective, but are also
highly active (Fig. 4). Complete conversions were
achieved within 2 h at 0°C, placing these systems
among some of the most active catalysts reported to
date.
a
Reaction conditions are given in Section 4. All reactions were
complete (100% conversion, TLC analysis) in 16 h (unoptimised).
Determined by NMR using Eu(hfc) as chiral resolving agent.
b
3

X. Cheng, K. K. Hii / Tetrahedron: Asymmetry 14 (2003) 2045–2052
2049
3
. Conclusion
Al O , Et O:hexane:Et N, 20:78:2), furnishing the
2 3 2 3
product as a colourless liquid. Yield 0.30 g, 70%. Anal.
calcd for C H N P: C, 74.97; H, 9.03; N, 7.60%.
In summary, a series of novel PN(N) ligands have been
synthesised with systematic variations at the N-donor
group(s). These ligands effect high activity and enan-
tioselectivities in palladium-catalysed asymmetric allylic
alkylation reactions. The metal-to-ligand ratio was
found to affect the enantioselectivity of these ligands in
a profound way. In addition, there are clearly synergic
effects between the nature of the pendant donor group,
the central amine/amide functionality and the presence
of an additional stereogenic centre on the stereoselectiv-
ity of the catalysed reaction. This supports our belief
that unsymmetrical terdentate ligands provide new pos-
sibilities and opportunities in asymmetric catalysis by
offering several possible tunable sites for the optimal
control of reactivity and stereoselectivity.
2
3
33
2
31
Found C, 74.61; H, 9.00; N, 7.54%. P NMR (145
1
MHz, CDCl ) l: −19.7. H NMR (360 MHz, CDCl )
3
3
l: 0.93 (t, 6H, J=7.2 Hz), 1.50–1.79 (m, 3H), 1.83–1.96
(m, 2H), 2.01–2.12 (m, 2H), 2.19–2.29 (m, 1H), 2.41–
2.52 (m, 7H), 2.83–2.92 (m, 1H), 3.04–3.10 (m, 1H),
13
7.22–7.39 (m, 10H). C NMR (90.5 MHz, CDCl ) l:
3
1
1.9, 22.6, 23.9 (d, J=8 Hz), 34.2 (d, J=13 Hz), 47.8,
5
2.2, 52.8, 54.5, 63.0 (d, J=19 Hz), 128.1–140.2 (Ph).
+
HRMS (FAB): exact mass calcd for C H N P (MH )
23
34
2
−
1
3
69.2460, found 369.2468. IR (cm , thin film, NaCl
25
discs) w: 2965, 2798, 1585, 1433, 1069, 738, 696. [h] =
D
−121.4 (c 1.13, EtOH).
(
2S) - (−) - 1 - [2 - (N - Methyl - N - phenyl)aminoethyl] - 2-
[
(diphenylphosphino)methyl]-pyrrolidine, 2c was prepared
1
1
from 2-(N-methyl-N-phenyl)aminoethyl tosylate and
. Colourless liquid. Yield: 75%. Anal. calcd for
C H N P: C, 77.58; H, 7.76; N, 6.96%. Found C,
4
4
. Experimental
2
6
31
2
31
7
7.60; H, 7.63; N, 6.79%. P NMR (145 MHz, CDCl3)
1
4.1. General
l: −20.1. H NMR (360 MHz, CDCl ) l 1.45–1.79 (m,
3
3
H), 1.81–1.98 (m, 2H), 2.04–2.20 (m, 2H), 2.25–2.40
All reactions were carried out under an inert atmo-
sphere of nitrogen or argon using standard Schlenk line
techniques. Solvents were purified by standard proce-
dures. (R) and (S)-2-N,N-Dimethylamino-3-phenyl-1-
(
m, 1H), 2.38 (dt, 1H, J=3.4, 13.2 Hz), 2.82 (s, 3H),
2
3
1
.80–2.89 (m, 1H), 3.10–3.20 (m, 1H), 3.24 (m, 1H),
.38 (ddd, 1H, J=5.0, 9.9, 14.8 Hz), 6.39–7.62 (m,
13
5H). C NMR (90.5 MHz, CDCl ) l: 23.0, 32.0 (d,
3
propylchloride hydrochloride, (S)-(+)-2-N,N-dimethyl-
J=7 Hz), 34.6 (d, J=13 Hz), 39.0, 50.9, 52.3, 54.6, 62.9
d, J=19 Hz), 112.3–149.4 (Ph). HRMS (FAB): exact
7
amino-3-methyl-1-butylchloride hydrochloride
(
and
(
8
S)-(−)-2-[(diphenylphosphino)methyl]pyrroline 4 were
+
mass calcd for C H N P (MH ), 403.2303, found
2
6
32
2
prepared following reported procedures. Unless other-
wise stated, all other reagents were procured from
Avocado, Lancaster or Aldrich chemical companies
−
1
4
2
03.2322. IR (cm , thin film, NaCl discs) w: 3067,
25
959, 2795, 1599, 1505, 1103, 746, 694. [h] =−144.5 (c
D
31
1
13
1.05, EtOH).
and used as received. P, H and C NMR spectra
were recorded using Bruker AVANCE 360, 400 or 500
1
13
(2%S,2S) - 1 - [2 - (N,N - Dimethylamino) - 3 - phenyl)] - 2-
[(diphenylphosphino)methyl]pyrrolidine, 3a. From (S)-
MHz instruments with TMS ( H and C) and 85%
3
1
H PO ( P) as external standards. Mass spectrometry
3
4
(
+)-2-N,N-dimethylamino-3-phenyl-1-propylchloride
facilities were provided by the University of London
MS Service. Elemental analyses were performed at
London Metropolitan University. HPLC were per-
formed on a Gilson instrument.
hydrochloride and 4. Colourless oil. Yield: 65%. Anal.
calcd for C H N P: C, 78.11; H, 8.19; N, 6.51%.
Found C, 77.96; H, 8.31; N, 6.57%. P NMR (145
MHz, CDCl ) l: −20.0. H NMR (360 MHz, CDCl )
28
35
2
31
1
3
3
l: 150–1.78 (m, 3H), 1.88–1.96 (m, 2H), 2.08–2.14 (m,
4
4
.2. Ligand synthesis
1
2
2
3
H), 2.26 (dd, 1H, J=5.8, 12.4 Hz), 2.30 (s, 6H),
.29–2.68 (m, 1H), 2.38 (m, 1H), 2.61–2.68 (m, 2H),
.74 (dd, 1H, J=6.3, 13.9 Hz), 2.85–2.92 (m, 1H),
.2.1. Method A. Ligands 2b, 2c, 3a, 3b and 3c were
13
.12–3.18 (m, 1H), 7.32–7.43 (m, 15H). C NMR (90.5
prepared by the alkylation of (S)-(−)-2-[(diphenylphos-
phino)methyl]pyrroline 4 with the corresponding N,N-
disubstituted-aminoalkylchlorides or tosylate. A general
procedure is described below for the synthesis of (2S)-
MHz, CDCl ) l: 22.9, 31.9 (d, J=8 Hz), 33.9, 34.3 (d,
3
J=13 Hz), 41.2, 54.6, 54.9, 62.9 (d, J=19 Hz), 65.7,
1
25.9–142.1. HRMS (FAB): exact mass calcd for
+ −1
C H N P (MH ) 431.2616, found 431.2629. IR (cm ,
(
−) - 1 - [2 - (N,N - diethylamino)ethyl] - 2 - [(diphenylphosp-
28 36
2
thin film, NaCl discs) w: 3053, 2935, 2776, 1601, 1584,
hino)methyl]pyrroline, 2b: A mixture of (S)-(−)-2-
2
5
1
433, 738, 696. [h] =−83.1 (c 1.09, EtOH).
[
(diphenylphosphino)methyl]pyrroline 4 (0.40 g, 1.31
D
mmol), 2-(N,N-diethylamino)ethylchloride hydrochlo-
ride (0.20 g, 1.16 mmol) and K CO (2.0 g) were
(2%R,2S)-(−)-1-[2-(N,N-Dimethylamino)-3-phenyl)]-2-
[(diphenylphosphino)methyl]-pyrrolidine, 3b. From (R)-
(−)-2-N,N-dimethylamino-3-phenyl-1-propylchloride
hydrochloride and 4. Yield: 80%. Anal. calcd for
2
3
dissolved in degassed CH CN (30 ml). The mixture was
3
stirred for 30 min before the addition of KI (0.05 g,
0.30 mmol). Stirring was continued overnight at rt. The
reaction mixture was then filtered via cannula and the
residue was washed with degassed diethyl ether. The
combined filtrate was evaporated under reduced pres-
sure and subjected to column chromatography (neutral
C H N P: C, 78.11; H, 8.19; N, 6.51%. Found C,
28 35 2
31
78.27; H, 8.24; N, 6.48%. P NMR (145 MHz, CDCl3)
1
l: −19.1. H NMR (360 MHz, CDCl ) l: 1.37–1.53 (m,
3
2H), 1.56–1.85 (m, 4H), 1.88–1.99 (m, 1H), 2.03–2.25

2
050
X. Cheng, K. K. Hii / Tetrahedron: Asymmetry 14 (2003) 2045–2052
(
2
1
m, 2H), 2.28 (s, 6H), 2.39 (dd, 1H, J=3.2, 13.0 Hz),
C, 71.16; H, 7.68; N, 7.90%. Found C, 71.29; H, 8.06;
N, 8.09%. HRMS (FAB): exact mass calcd for
.72–2.89 (m, 3H), 2.91–2.99 (m, 1H), 7.03–7.37 (m,
13
+
5H). C NMR (90.5 MHz, CDCl ) l: 23.1, 32.8 (d,
C H N OP (MH ) 355.1939, found 355.1951. IR
3
21 28
−
2
1
J=7 Hz), 34.0, 34.6 (d, J=13 Hz), 41.7, 54.9, 55.6, 63.9
d, J=19 Hz), 65.6, 126.1–133.5. HRMS (FAB): exact
(cm , thin film, NaCl discs) w: 3051, 2944, 1654, 1428,
(
1170, 740, 700. A mixture of rotamers were observed in
+
mass calcd for C H N P (MH ) 431.2616, found
the NMR spectra of the compound in solution: Major
2
8
36
2
−
1
31
4
2
31.2603. IR (cm , thin film, NaCl discs) w: 3054,
934, 1602, 1584, 1433, 738, 697. [h] =−160.0 (c 1.18,
rotamer (78%): P NMR (145 MHz, CDCl ,) l: −21.1.
3
2
5
1
H NMR (360 MHz, CDCl ) l: 1.13–127 (m, 1H),
D
3
EtOH).
1.39–1.58 (m, 2H), 1.65–1.88 (m, 2H), 2.13 (s, 6H), 2.59
(
s, 2H), 2.98–3.23 (m, 1H), 4.11–4.29 (m, 1H), 7.18–7.64
13
(
2%S,2S)-(−)-1-[2-(N,N-Dimethylamino)-3-methyl]butyl-
-[(diphenylphosphino)methyl]-pyrrolidine, 3c. The com-
pound was synthesised from (S)-(+)-2-N,N-
(m, 10H). C NMR (90.5 MHz, CDCl ) l: 24.9, 30.7
(d, J=8 Hz), 33.5 (d, J=14 Hz), 45.8, 46.8, 56.1 (d,
J=20 Hz), 64.1, 125.2–141.0 (Ph), 168.3. Minor
3
2
31
dimethylamino-3-methyl-1-butylchloride hydrochloride
and 4. Colourless liquid. Yield: 54%. Anal. calcd for
C H N P: C, 75.36; H, 9.22; N, 7.32%. Found C,
rotamer (22%). P NMR (145 MHz, CDCl ) l: −20.2.
3
1
H NMR (360 MHz, CDCl ) l: 1.27–1.38 (m, 2H),
3
1.39–1.58 (m, 2H), 1.94 (s, 6H), 2.20–2.30 (m, 2H), 2.64
(d, 1H, J=13.9 Hz), 2.72 (d, 1H, J=13.9 Hz), 3.20–
3.50 (m, 2H), 4.00–4.10 (m, 1H), 7.18–7.64 (m, 10H).
2
4
35
2
3
1
7
5.33; H, 9.27; N, 7.32%. P NMR (145 MHz, CDCl3)
1
l: −19.1. H NMR (360 MHz, CDCl ) l: 0.84 (d, 3H,
J=6.9 Hz), 0.89 (d, 3H, J=6.8 Hz), 1.35–1.69 (m, 3H),
3
1
3
C NMR (90.5 MHz, CDCl ) l: 22.0, 31.8 (d, J=8
3
1
2
.70–1.90 (m, 3H), 1.93–2.10 (m, 2H), 2.18 (s, 6H),
.11–2.25 (m, 2H), 2.48 (dt, 1H, J=3.4, 13.1 Hz), 2.61
Hz), 35.3 (d, J=16 Hz), 45.7, 45.9, 55.7 (d, J=23 Hz),
25
61.4, 125.2–141.0, 167.9. [h] =−94.5 (c 1.24, EtOH).
D
(
(
dd, 1H, J=7.3, 12.1 Hz), 3.02–3.18 (m, 1H), 7.21–7.42
m, 10H). C NMR (90.5 MHz, CDCl ) l: 20.6, 21.4,
13
One-pot reduction and decomplexation to (2S)-(−)-
1-(2-N,N-dimethylamino)ethyl-2-[(diphenylphosphino)-
methyl]pyrrolidine, 2a. After the dropwise addition of
3
2
5
2.7, 29.4, 32.1 (d, J=7 Hz), 34.3 (d, J=13 Hz), 42.1,
2.2, 54.5, 63.2 (d, J=19 Hz), 67.4, 128.7–133.5.
+
HRMS (FAB): exact mass calcd for C H N P (MH )
BH (1.5 M in THF, 4.2 ml, 6.8 mmol) to a solution of
24
36
2
3
−
1
383.2616, found 383.2599. IR (cm , thin film, NaCl
5a (0.30 g, 0.85 mmol) in THF (10 ml) at rt, the
reaction mixture was refluxed gently for 8 h. The
solution was then cooled in ice-bath, and methanol (2
ml) was added carefully. When the effervescence ceased,
the solvents were removed in vacuo. Degassed pyrro-
lidine (6 ml), methanol (20 ml) and Raney Nickel (0.1
g) were then added and the reaction mixture was
allowed to reflux overnight. The solvent was then
discs) w: 3053, 2953, 2775, 1585, 1456, 1433, 1026, 737,
2
5
696. [h] =−132.6 (c 1.08, EtOH).
D
(
2S)-(−)-1-(1-Naphthylmethyl)-2-[(diphenylphosphino)-
methyl]-pyrrolidine, 8a. Prepared from 1-naphthyl-
methylchloride and 4. White solid. Yield: 60%. Mp
7
5–76°C. Anal. calcd for C H NP: C, 82.12; H, 6.89;
28 28
31
N, 3.42%. Found C, 82.17; H, 7.00; N, 3.32%.
P
removed in vacuo and the residue extracted with Et O
2
1
NMR (145 MHz, CDCl ) l: −19.6. H NMR (360
MHz, CDCl ) l: 1.61–1.80 (m, 3H), 2.02–2.28 (m, 3H),
(2×20 ml). The combined organic extracts were evapo-
rated and the residue was purified by column chro-
matography (neutral Al O , Et O:hexane:NEt , 40:58:2)
3
3
2
2
1
.50–2.64 (m, 1H), 2.76 (dt, 1H, J=3.5, 13.1 Hz),
.81–2.89 (m, 1H), 3.53 (d, 1H, J=12.9 Hz), 4.55 (d,
H, J=12.9 Hz), 7.23–8.50 (m, 17 H). C NMR (90.5
2
3
2
3
to afford a colourless oil. Yield: 0.18 g, 61%. Anal.
calcd for C H N P: C, 74.09; H, 8.59; N, 8.32%.
13
2
1
29
2
31
MHz, CDCl ) l: 22.5, 32.4 (d, J=7 Hz), 34.4 (d, J=13
Hz), 54.6, 57.0, 63.1 (d, J=19 Hz), 124.8–139.8.
HRMS: exact mass calcd for C H NP (MH )
Found C, 73.99; H, 8.68; N, 8.42%. P NMR (145
3
1
MHz, CDCl ) l: −19.7. H NMR (360 MHz, CDCl )
3
3
+
l: 1.44–1.78 (m, 3H), 1.84–2.08 (m, 4H), 2.13 (s, 6H),
2.17–2.32 (m, 3H), 2.46 (dt, J=3.3, 13.1 Hz), 2.83–2.91
(m, 1H), 3.06 (br t, 1H, J=8.4 Hz), 7.21–7.39 (m, 10H).
28
29
−
1
410.2038, found 410.2050. IR (cm , Nujol, NaCl discs)
2
5
w: 2927, 1459, 1376, 1139, 733. [h] =−154.8 (c 1.08,
EtOH).
D
13
C NMR (90.5 MHz, CDCl ) l: 22.6, 32.0 (d, J=7
3
Hz), 34.2 (d, J=13 Hz), 46.3, 52.8, 54.5, 58.9, 63.0 (d,
J=19 Hz), 128.7–139.8 (Ph). HRMS (FAB): exact mass
+
4
.2.2. Method B. General procedure is described for the
synthesis of (2S)-1-[2-(N,N-dimethylamino)]acetyl-2
diphenylphosphinomethyl)-pyrrolidine, 5a. A mixture of
-(N,N-dimethylamino)acetic acid hydrochloride (0.56
calcd for C H N P (MH ) 341.2147, found 341.2158.
2
1
30
2
−
1
IR (cm , thin film, NaCl discs) w: 3069, 2942, 2756,
1585, 1433, 738, 769. [h] =−118.2 (c 1.12, EtOH).
2
5
(
2
D
g, 4.0 mmol), 4 (1.10 g, 3.6 mmol), DMAP (1.46 g, 12
(2S)-(−)-1-[2-(N,N-Diphenylamino)acetyl]-2-(diphenyl-
phosphinomethyl)-pyrrolidine, 5d. The compound was
recrystallised from methanol. Yield: 59%. Mp 125–
126°C. Anal. calcd for C H N OP: C, 77.80; H, 6.53;
mmol) and DCC (1.24 g, 6.0 mmol) in CH Cl (20 ml)
2
2
was stirred at 30°C overnight. The reaction mixture was
then extracted with 1N aq. HCl (25 ml). The aqueous
layer was neutralised by the addition of 30% aq. NaOH
31
31
2
N, 5.85%. Found C, 77.72; H, 6.59; N, 5.75%. HRMS
+
(
4 ml) at 0°C, before extraction with n-hexane (30 ml).
(FAB): exact mass calcd for C H N OP (MH )
31
32
2
−
1
The combined organic layers were washed repeatedly
with water, then evaporated to dryness. The residue
479.2252, found 479.2266. IR (cm , Nujol, NaCl discs)
31
w: 2921, 1646, 1459. Major rotamer (73%): P NMR
1
was subjected to column chromatography (SiO , tolu-
(145 MHz, CDCl ) l: −21.3. H NMR (360 MHz,
CDCl ) l: 1.77–2.12 (m, 5H), 2.82 (dt, 1H, J=3.8, 13.0
2
3
ene:EtOAc:NEt , 90:8:2) to give the amide as a colour-
3
3
less oil. Yield: 0.65 g 51%. Anal. calcd for C H N OP:
Hz), 3.24–3.45 (m, 2H), 4.19 (d, 1H, J=16.9 Hz),
2
1
27
2

X. Cheng, K. K. Hii / Tetrahedron: Asymmetry 14 (2003) 2045–2052
2051
4
7
3
.17–4.26 (m, 1H), 4.29 (d, 1H, J=16.9 Hz), 6.80–
2.16 (m, 1H), 2.25–2.35 (m, 1H), 2.29 (d, 1H, J=13.3
Hz). 2.42 (dt, 1H, J=3.5, 13.2 Hz), 3.10–3.17 (m,
1H), 7.21–7.29 (m, 10H). C NMR (90.5 MHz,
1
3
.50 (m, 20H). C NMR (90.5 MHz, CDCl ) l: 25.0,
3
1
3
0.1 (d, J=9 Hz), 32.4 (d, J=14 Hz), 46.7, 55.7, 56.4
(
(
d, J=19 Hz), 121.1–148.3, 167.8. Minor rotamer
27%): P NMR (145 MHz, CDCl ,) l: −20.7.
CDCl ) l: 23.6, 29.0, 31.9 (d, J=7 Hz), 33.0, 35.1 (d,
3
3
1
1
H
J=13 Hz), 57.9, 64.4 (d, J=20 Hz), 68.1, 128.6–
3
NMR (360 MHz, CDCl ) l: 1.77–2.12 (m, 5H), 2.12–
140.1. HRMS (FAB): exact mass calcd for C H NP
3
22 31
+
−1
2.22 (m, 1H), 3.24–3.45 (m, 2H), 3.62–3.78 (1H, m),
(MH ) 340.2194, found 340.2207. IR (cm , thin film,
NaCl discs) w: 3053, 2950, 2805, 1585, 1433, 1360,
1
3
3
.90 (s, 2H), 6.80–7.50 (m, 20H). C NMR (90.5
2
5
MHz, CDCl ) l: 21.8, 32.1 (d, J=7 Hz), 35.6 (d,
J=16 Hz), 46.0, 54.2, 55.1 (d, J=21 Hz), 121.1–
1108, 737, 695. [h] =−152.1 (c 1.01, EtOH).
3
D
2
5
148.3, 167.8. [h] =−67.8 (c 1.06, EtOH).
D
(2%S,2S) - (−) - 1 - [2 - (N,N - Dimethylamino) - 3 - methyl]-
butyryl-2-(diphenylphosphinomethyl)-pyrrolidine,
10.
(
2S)-(−)-1-(2-N,N-Diphenylamino)ethyl-2-(diphenyl-
phosphinomethyl)-pyrrolidine, 2d. Colourless oil. Yield:
2%. Anal. calcd for C H N P: C, 80.14; H, 7.16;
Synthesised via the coupling between (S)-(+)-N,N-
dimethyl valine and 4, and was obtained as a colour-
less liquid. Yield: 71%. Anal. calcd for C H N OP:
8
31
33
2
24
33
2
N, 6.03%. Found C, 80.04; H, 7.05; N, 5.94%.
C, 70.17; H, 8.41; N, 3.90%. Found C, 70.26; H,
8.45; N, 3.83%+. HRMS: exact mass calcd for
C H N OP (M +1) 397.2409, found 397.2411. A
+
HRMS (FAB): exact mass calcd for C H N P (MH )
31
34
2
3
1
465.2460, found 465.2479.
P NMR (145 MHz,
24
34
2
1
CDCl ) l: −20.1 Hz. H NMR (360 MHz, CDCl ) l:
3
3
mixture of rotamers were observed in the NMR spec-
tra of the compound in solution: Major rotamer
1
.54–1.78 (m, 3H), 1.80–1.96 (m, 2H), 2.10 (dd, 1H,
J=8.6, 17.1 Hz), 2.22–2.48 (m, 3H), 2.93–3.01 (m,
H), 3.07–3.12 (m, 1H), 3.68–3.75 (m, 2H), 6.83–7.43
31
1
(
75%): P NMR (145 MHz, CDCl ) l: −20.3. H
3
1
NMR (360 MHz, CDCl ) l: 0.90 (d, 3H, J=6.6 Hz),
1
3
3
(
m, 20H). C NMR (90.5 MHz, CDCl ) l: 23.1, 32.0
3
1
.03 (d, 3H, J=6.6 Hz), 1.62–2.10 (m, 5H), 212–2.22
(d, J=7 Hz), 34.6 (d, J=13 Hz), 51.6, 51.8, 54.7,
(
m, 1H), 2.39 (s, 6H), 2.96 (d, 1H, J=10.1 Hz), 3.12–
−
1
62.8 (d, J=19 Hz), 121.1–148.1. IR (cm , thin film,
3
7
2
1
1
.18 (m, 1H), 3.42–3.65 (m, 2H), 4.32–4.45 (m, 1H),
NaCl discs) w: 3053, 1588, 1496, 1361, 1154, 747, 695.
13
.28–7.73 (m, 10H). C NMR (90.5 MHz, CDCl ) l:
2
5
3
[
h] =−102.1 (c 1.05, EtOH).
D
0.1, 20.2, 24.7, 28.8, 30.2 (d, J=9 Hz), 31.8 (d, J=
4 Hz), 41.8, 47.7, 55.6 (d, J=21 Hz), 71.1, 128.7–
(
S)-(−)-1-(2,2-Dimethylpropionyl)-2-(diphenylphosphino-
6
31
39.8, 170.8. Minor rotamer (25%): P NMR (145
methyl)-pyrrolidine, 9. Pivoyl chloride (0.41 ml, 3.3
mmol) was added dropwise to a solution of 4 (0.60
1
MHz, CDCl ) l: −19.8. H NMR (360 MHz, CDCl )
3
3
l: 0.61 (d, 3H, J=6.6 Hz), 0.75 (d, 3H, J=6.6 Hz),
ml, 2.2 mmol) and triethylamine (1.2 ml) in CH Cl2
2
1
3
1
2
4
.62–2.10 (m, 3H), 2.35 (s, 6H). 2.00–2.40 (m, 5H),
(
3
12 ml) at 0°C. The reaction mixture was stirred for
0 min, then diluted with CH Cl2 (20 ml), washed
.42–3.65 (m, 2H), 3.70–3.90 (m, 1H), 7.28–7.73 (m,
2
13
0H). C NMR (90.5 MHz, CDCl ) l: 19.4, 20.2,
3
with water (2×10 ml) and dried over Na SO . After
filtration and concentration, the residue was passed
through a silica column (hexane:EtOAc, 4:1) to afford
2
4
2.1, 29.1, 31.1 (d, J=9 Hz), 34.5 (d, J=14 Hz),
2.2, 45.3, 55.3 (d, J=21 Hz), 70.3, 128.7–139.8,
−
1
6
172.3. IR (cm , thin film, NaCl discs) w: 3051, 2954,
1633, 1449, 739, 695. [h] =−78.8 (c 1.24, EtOH).
0
9
.69 g of white solid. Yield: 88% (lit. 68%). Mp
6–97°C
2
5
6
D
(lit.
97–97.5°C).
Anal.
calcd
for
C H NOP: C, 74.76; H, 7.99; N, 3.96%. Found C,
22
28
31
4.3. General catalytic procedure
7
4.64; H, 8.06; N, 3.85%. P NMR (145 MHz,
1
CDCl ) l: −19.9. H NMR (360 MHz, CDCl ) l:
3
3
The appropriate ligand (27 mmol, 2.7 mol%) and
1
2
1
.11 (s, 9H), 1.68–1.73 (m, 1H), 1.80–2.00 (m, 4H),
3
12
[
PdCl (h -C H )]
(15 mmol, 1.5 mol%) were placed
2
3
5 2
.84–2.94 (m, 1H), 3.40–3.44 (m, 1H), 3.62–3.66 (m,
13
into a Young’s tube reaction tube, which was purged
and filled with argon successively. CH Cl (2 ml) was
H), 4.21–4.27 (m, 1H), 7.16–7.56 (m, 10H).
C
NMR (90.5 MHz, CDCl ) l: 25.9, 27.9, 29.9 (d, J=
2
2
3
then added and catalytic solution was refluxed for 2
h, then cooled down to 0°C. It was then treated with
a solution of (rac)-diphenylpropen-2-yl acetate (0.5
1
0 Hz), 33.0 (d, J=13 Hz), 39.4, 48.4, 57.5 (d, J=18
Hz), 128.7–137.7, 176.7. HRMS (FAB): exact mass
calcd for C H NOP (MH ) 354.1987, found
3
1
+
2
2
29
1
−
mmol), CH Cl2 (0.5 ml), dimethyl malonate (1.5
54.1983. IR (cm , Nujol, NaCl discs) w: 2928, 1615,
6
2
2
5
mmol), N,O-BSA (1.5 mmol) and KOAc (0.75 mg).
The reaction mixture was then stirred overnight (16
h) at 0°C. It was quenched by the addition of a
466, 1407, 746. [h] =−71.6° (c 1.70, EtOH). lit.
D
−67.3 (c 1.45, CHCl3).
saturated solution of NH Cl. After separation, the
organic layer was washed successively with an aq.
Compound 9 was then reduced with borane, as
described above to give (2S)-(−)-1-(2,2-dimethyl)-
propyl-2-(diphenylphosphinomethyl)-pyrrolidine, 8b as a
colourless liquid. Yield: 81%. Anal. calcd for
C H NP: C, 77.84; H, 8.91; N, 4.13%. Found C,
4
NaHCO , dried (MgSO ), filtered and evaporated
3 4
under reduced pressure. The residue was then passed
through a short silica column to afford the product
as a colourless oil. The e.e. and the conversion of the
reaction were determined by chiral HPLC using a
22
30
31
7
7.91; H, 9.03; N, 4.06%. P NMR (145 MHz,
1
CDCl ) l: −19.2. H NMR (360 MHz, CDCl ) l:
0
3
3
i
.80 (s, 9H), 1.32–1.43 (m, 1H), 1.50–1.70 (m, 2H),
.78–1.92 (m, 2H), 1.93 (d, 1H, J=13.3 Hz), 2.09–
Chiralpak AD column (Daicel Industries): 10% PrOH
1
in n-hexane, 1 ml/min, t =11.1 min, t =15.1 min.
R
S

2
052
X. Cheng, K. K. Hii / Tetrahedron: Asymmetry 14 (2003) 2045–2052
5. Barbaro, P.; Bianchini, C.; Giambastiani, G.; Togni, A.
Acknowledgements
Chem.Commun. 2002, 2672–2673.
6
. Kanai, M.; Nakagawa, Y.; Tomioka, K. Tetrahedron
1999, 55, 3843–3854.
. Hayashi, T.; Konishi, M.; Fukushima, M.; Kanehira, K.;
Hioki, T.; Kumada, M. J. Org. Chem. 1983, 48, 2195–
The authors thank the following organisations for
financial support: K. C. Wong Education Foundation,
British Foreign and Commonwealth Office (FCO),
Education Department of PR China and The Nuffield
Foundation. Palladium salts were provided courtesy of
Johnson Matthey plc, through a precious metal loan
agreement. K.K.H. thanks Professor Kevin Burgess
7
2
202.
. Hiroi, K.; Suzuki, Y.; Abe, I. Tetrahedron: Asymmetry
999, 10, 1173–1188.
8
9
1
. Hii, K. K.; Thornton-Pett, M.; Jutand, A.; Tooze, R. P.
Organometallics 1999, 18, 1887–1896.
(
Texas A&M) for helpful comments received during the
preparation of this manuscript.
10. Hou, D.; Reibenspies, J. H.; Burgess, K. J. Org. Chem.
2001, 66, 206–215. The metal-to-ligand ratio has been
previously reported to affect the coordination of
aminophosphinites to rhodium complexes: Kuznetsov, V.
F.; Facey, G. A.; Yap, G. P. A.; Alper, H. Organometal-
lics 1999, 18, 4706–4711.
References
1
. Helmchen, G.; Pfaltz, A. Acc. Chem. Res. 2000, 33,
3
36–345.
1
1. Wimalasena, K.; May, S. W. J. Am. Chem. Soc. 1987,
2
3
. Helmchen, G. J. Organomet. Chem. 1999, 576, 203–214.
. Lam, H.; Cheng, X.; Steed, J. W.; Aldous, D. J.; Hii, K.
K. Tetrahedron Lett. 2002, 43, 5875–5877.
109, 4036–4046.
1
2. Tatsuno, Y.; Yoshida, T.; Otsuka, S. Inorg. Synth. 1990,
28, 342–345.
4
. Fadini, L.; Togni, A. Chem.Commun. 2003, 30–31.