Novel chiral P,O-ligands for homogeneous Pd(0) catalysed asymmetric allylic alkylation reactions

Article abstract of DOI:10.1016/j.tetasy.2008.01.024

Two novel P,O-coordinating chiral phoshine-amide ligands with hydroxymethyl side chains 5a and 5b were prepared and tested in some benchmark Pd(0) catalysed asymmetric allylic alkylations. The conversions and enantioselectivities were generally good, with

Full text of DOI:10.1016/j.tetasy.2008.01.024

Available online at www.sciencedirect.com
Tetrahedron: Asymmetry 19 (2008) 454–458
Novel chiral P,O-ligands for homogeneous Pd(0) catalysed
asymmetric allylic alkylation reactions
Vanda Raquel Marinho,a Ana Isabel Rodriguesb and Anthony J. Burkea,*
a
´
´
´
Departamento de Quı´mica and Centro de Quı´mica de Evora, Universidade de Evora, Rua Roma˜o Romalho 59, 7000 Evora, Portugal
bDepartamento de Tecnologia de Indu´strias Quı´micas, Instituto Nacional de Engenharia, Tecnologia e Inovaßca˜o,
Estrada do Paßco do Lumiar, Edifı´cio F, 1649-038 Lisboa, Portugal
Received 30 November 2007; accepted 17 January 2008
Abstract—Two novel P,O-coordinating chiral phoshine–amide ligands with hydroxymethyl side chains 5a and 5b were prepared and
tested in some benchmark Pd(0) catalysed asymmetric allylic alkylations. The conversions and enantioselectivities were generally good,
with a highest conversion of 92% and a highest ee of 62% obtained using 1,3-diphenylpropenyl acetate as substrate. A number of pal-
ladium pre-catalysts were used, but it was [Pd(allyl)Cl]2 which gave the best results. Ligand 5b was seen to give better enantioselectivities.
Ó 2008 Elsevier Ltd. All rights reserved.
1. Introduction
An interesting study of this ligand in Pd(0) catalysed allylic
alkylations by Lloyd-Jones and co-workers4 demonstrated
that under certain circumstances, an alternative P,O mode
of coordination of this ligand with Pd was possible giving a
catalytically active complex (Fig. 1). Amatore et al.5 have
unambiguously shown this type of coordination in a
Pd–allyl complex formed with Trost’s diaminocyclohex-
ane-2-diphenylphosphinonapthoyl ligand. Clayden et al.6
developed an atropisomeric phosphinamide that showed
excellent enantioselectivities in the benchmark allylic alkyl-
ation of 1,3-diphenylpropenyl acetate. Also worthy of note
are the phosphine–amide ligands 3 and 4 (Fig. 2) derived
from O-methoxy (S)-phenylalaninol developed by Mino
et al.7 that were successfully used in palladium catalysed
asymmetric allylic alkylations giving a highest ee of 85%.
Sinou and co-workers8 have also developed a series of
sugar-derived amides that have also been successful for this
Catalytic asymmetric synthesis has currently considerable
application in providing useful enantiomerically pure
compounds. One of the outstanding endeavours in the area
is the creation of new C–C and C–X bonds. Since 1977, the
asymmetric allylic alkylation (AAA) reaction has become a
standard approach for achieving this objective.1,2 Many
chiral ligands have been screened for this particular
reaction; these generally contain P or N, possess C2 or C1
symmetry and can be of a homobidentate type (e.g.,
diphosphines and bisoxazolines) or of a heterobidentate
nature (e.g., phosphinooxazolines).1
Trost’s ligand 1 (Fig. 1) has shown considerable applicabil-
ity for this particular reaction,2,3 and has elicited much
interest from the chemical community.
6-membered
chelate
Ph
Ph
PPh2
NH HN
Ph2P
P
NH HN
P
Pd
O
O
Ph
Ph
O
O Pd
(R,R)-1
2
Figure 1.
*
Corresponding author. Tel.: +351 266745310; fax: +351 266745303; e-mail: ajb@dquim.uevora.pt
0957-4166/$ - see front matter Ó 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetasy.2008.01.024

V. R. Marinho et al. / Tetrahedron: Asymmetry 19 (2008) 454–458
455
Ph
OMe
O
R
Ph
OMe
O
O
R
O
O
OH
i
OH
N
OH
PPh2
N
N
N
H
H
H
H
PPh2
5a = Ph (78%)
5b = Bn (87%)
PPh2
5a = Ph (78%)
5b = Bn (87%)
PPh2
PPh2
6
3
4
Scheme 1. Reagents and conditions: (i) (S)-phenylglycinol or (S)-phen-
Figure 2.
ylalaninol, DCC, DMAP, CH2Cl2, rt, 24 h.
reaction. Jiang and Huang9 prepared a series of P,O-type
ligands based on (S)-prolinol that gave ees of up to
90% in the asymmetric allylic alkylation of dimethyl
malonate with 1,3-diphenyl propenyl acetate. Gilbertson
and coworkers10 have shown that certain b-turn derived
palladium phosphine complexes give ees up to 95% in
the alkylation of 3-acetoxycyclopentene with dimethyl
malonate.
Once in hand, these ligands were first tested in the asym-
metric allylic alkylation reaction of dimethyl malonate with
rac-1,3-diphenylpropenyl acetate (Table 1). Overall, the
results were quite encouraging with a number of good to
excellent conversions and with generally good enantioselec-
tivites (a highest of 62% was obtained, Table 1, entry 4). It
was ligand 5b that gave the best ees, which is attributed to
the greater stereochemical hindrance associated with the
benzyl group over a phenyl group. Ligand 5a afforded
the (R)-enantiomer as the major enantiomer for the malon-
ate product when [Pd(allyl)Cl]2 was used, but afforded the
(S)-enantiomer when the other two Pd pre-catalysts were
used. Ligand 5b gave the (S)-enantiomer as the major
enantiomer, except when DMF was used as the solvent
where it was found that the (R)-enantiomer was the major
enantiomer (19% ee and a conversion of 12%). We find it
difficult to explain this ‘chiral switching’ phenomenon,
but are aware that a similar phenomenon was shown by
Sinou and co-workers8c when the reaction solvent was
changed. Elevated temperatures were needed to achieve
the highest conversions in all the systems studied, and in
some cases there was an increase in the enantioselectivity.
For example, when [Pd(allyl)Cl]2 was used as the pre-cata-
lyst at room temperature with ligand 5a in DCE, a conver-
sion of 44% was registered and an ee of 16%, but, when the
reaction was conducted at reflux (Table 1, entry 2), a con-
version of 91% was registered with a rise in the ee to 37%.
This trend was also observed when ligand 5b was used un-
der the same conditions. When DCM was used with 5b
(Table 1, entry 5), there was also a substantial rise in the
conversion (38% was obtained at rt), although the ee
remained about the same (59% and 55%, respectively).
However, this was not a general effect, for instance, when
Due to our interest in the catalytic asymmetric allylic alkyl-
ation,11 we became interested in preparing the free OH
derivative of Mino’s ligand7 5b (Fig. 2) and its phenyl gly-
cinol analogue 5a. We postulated that the free OH group
might act as a third coordination site with the metal leading
to more rigid transition states that promote greater enan-
tiocontrol in the reaction.
Herein, we report our preliminary results on the synthesis
of ligands 5a and 5b, and their application in some Pd(0)
catalysed asymmetric allylic alkylations.
2. Results and discussion
Ligands 5a and 5b were prepared in good yields after much
experimentation from o-(diphenylphosphine)benzoic acid 6
using (S)-phenylglycinol and (S)-phenylalanine according
to the conditions shown in Scheme 1.
Purification of these ligands was hampered by the tendency
of the dicyclohexylurea by-product to adhere to these
ligands.
Table 1. Asymmetric allylic alkylation of rac-1,3-diphenylpropenyl acetate using ligands 5a and 5ba
Pd (0), (MeO)2CH2,
Base (3 equivs), L*
MeO2C
CO2Me
Ph
OAc
Ph
Ph
Ph
Ligand
Entry
Pre-catalyst
Solvent
Temp (°C)
Base(s)
Conver.b (%)
eec (%)
1
2
3
4d
5
[Pd(allyl)Cl]2
[Pd(allyl)Cl]2
[Pd(allyl)Cl]2
[Pd(allyl)Cl]2
[Pd(allyl)Cl]2
[Pd(allyl)Cl]2
Pd2(dba)3ꢀCHCl3
5a
5a
5a
5b
5b
5b
5a
DCM
DCE
THF
DCM
DCM
THF
DCE
Reflux
Reflux
Reflux
rt
Reflux
Reflux
Reflux
BSA/KOAc
BSA/KOAc
BSA/KOAc
BSA/KOAc
BSA/KOAc
BSA/KOAc
BSA/KOAc
48
91
92
63
90
90
86
23 (R)
37 (R)
20 (R)
62 (S)
55 (S)
17 (S)
46 (S)
6
7
a Reaction conditions: 2.5 mol % of ligand and 1 mol % pre-catalyst, complexation conducted at reflux temperature for 2 h and then rt or at reflux for 20 h
(when Base = BSA/KOAc, 1 mol % KOAc was used).
b Pertains to the ratio of substrate to product as determined by HPLC.
c Determined by HPLC using a Chiralcel OD-H column.
d 5 mol % ligand to 2.5 mol % pre-catalyst.

456
V. R. Marinho et al. / Tetrahedron: Asymmetry 19 (2008) 454–458
[Pd(allyl)Cl]2 was used as the pre-catalyst at rt in DCM
with ligand 5a, a conversion of only 5% was registered
and an ee of 43%, but when carried out at reflux (Table
1, entry 1), a conversion of 48% was registered at the ex-
pense of a decrease in the ee to 23%. When one compares
the various pre-catalysts used, it can be seen that it was
[Pd(allyl)Cl]2 which gave the best conversions in all cases.
For example, using ligand 5a with DCE and THF, conver-
sions of 91% and 92% (Table 1, entries 2 and 3) could be
obtained. Ligand 5b also afforded a conversion of 90% in
THF (Table 1, entry 6). Using Pd(dba)2 it was found that
both ligands 5a and 5b gave low conversions, but 5b
gave the highest ee (53%). When 5 mol % ligand 5b (with
2.5 mol % [Pd(allyl)Cl]2) was used under the same condi-
tions (Table 1, entry 4), there was an increase in the
substrate conversion from 38% to 63%, while the enantio-
selectivity remained constant.
that Mino’s ligand gives slightly better enantiocontrol (76%
ee in DCM) than with our ligands 5a (37% ee in DCE,
Table 1, entry 2) and 5b (62% in DCM, Table 1, entry 4)
contrary to what had been predicted at the outset. A very
recent publication by Pons and Ros et al.13 has shown that
´
coordinating solvents lead to an enhanced fluxional behav-
iour of the allyl Pd complex (g3-g1-g3 isomerisation or
allyl rotation) with enhanced endo/exo (M/W) isomerisa-
tion,1 which we expect should reduce the reaction enanti-
oselectivity. To test their hypothesis, these workers
prepared a pyridylpyrazole probe bearing an hydroxyethyl
1
appendage and H NMR spectroscopy implied coordina-
tion of the OH group with the palladium atom.
1
It was not possible to confirm this experimentally by H
NMR in our case, due to the poor resolution of the spectra,
but the following mechanism based on the observations of
Pons and Ros is presented in Scheme 3.
The results were poorer when rac-3-acetoxycyclohexene
was used as substrate under standard conditions (Scheme
2). The reactions only proceeded when DCM was used as
the solvent and with [Pd(allyl)Cl]2 as the palladium source.
The conversions were poor (6–12%). Reactions using other
solvents (DCE and THF) and pre-catalysts (Pd(dba)2,
Pd2(dba)3ꢀCHCl3) failed to give a product. It was ligand
5b that furnished the best enantioselectivity (36% ee).
In an attempt to garner additional support for this hypoth-
esis, we prepared the following Pd complexes 7a and 7b
(Fig. 3) using the procedure previously reported11 in the
hope of obtaining crystals of these complexes appropriate
for X-ray analysis. Unfortunately, these complexes could
not be characterised by this technique due to their instabil-
ity (neither was it possible to do so by NMR due to the
poor spectral resolution) and were thus characterised by
mass spectrometry.
MeO2C
CO2Me
OAc
[Pd (C3H5)Cl]2, (MeO)2CH2,
BSA/KOAc (3 equivs), 5a or 5b
Pd
Pd
BF4
Ph
OH
BF4
Ph2P
O
Ph
Ph2P
O
Scheme 2.
OH
N
N
H
H
It is presumed that the generally lower conversions
obtained in the reactions performed at room temperature
were possibly a consequence of coordination of the free
ligand hydroxyl group with Pd, with concomitant deactiva-
tion of the metal. Sinou et al. and Desimoni et al. have pre-
viously observed that certain chiral pybox ligands give low
conversions for this particular reaction, no matter what
conditions were used,12 which might imply that coordina-
tion of an additional binding group (such as an hydroxyl)
to the Pd has a deactivating effect.
7b
7a
Figure 3.
Another possibility might be the displacement of the amide
carbonyl by the free hydroxyl group giving a competing
chiral catalyst.14
3. Conclusions
For the malonate alkylation reaction using 1,3-diphenyl-
propenyl acetate with allyl palladium dimer, it is apparent
The simple phosphine–amides 5a and 5b were obtained in
very good yields from o-(diphenylphosphine)benzoic acid
Pd
H
O
H
O
Pd
Pd
Ph2P
O
R
Ph2P
O
Ph2P
O
.
OH
N
H
N
H
R
N
H
R
Pd
Ph2P
O
R
OH
N
H
Scheme 3.

V. R. Marinho et al. / Tetrahedron: Asymmetry 19 (2008) 454–458
457
6 and the readily available amino-alcohols (S)-phenylgly-
cinol and (S)-phenylalaninol. Compounds 5a and 5b com-
plexed with various palladium pre-catalysts in situ to form
chiral catalysts that catalyse the asymmetric alkylation
reactions of 1,3-diphenylpropenyl acetate and 3-acetoxycy-
clohexene with moderate to good enantioselectivities. In
the case of the former substrate, very good substrate con-
versions were obtained when the reaction was carried out
at an elevated temperature. Further work is currently in
progress at immobilising these ligands onto solid supports
and applying these ligands in other catalytic asymmetric
reactions.
phine)benzoic acid 6 (0.500 g, 1.6 mmol) in CH2Cl2
(5 mL) was stirred under nitrogen at rt for 24 h. A white
precipitate was formed. The reaction was monitored by
TLC. The precipitate was filtered off and the solvent was
evaporated under reduced pressure. The crude product
was purified by chromatography on silica using hexane/
AcOEt (3:2) as eluent giving the title compound as a white
solid as product (78%). Mp 113.6–116.3 °C. IR [m (cmꢂ1)]:
1032.12, 1055.82, 1456.56, 1548.00, 1622.60, 2882.76,
1
2930.84, 3054.97, 3257.98; H NMR (CDCl3, 400 MHz) d
(ppm): 2.1 (s, 1H, –OH), 3.7–3.8 (m, 2H, –CH2OH),
5.1–5.2 (m, 1H, –CHPh), 6.4–6.5 (d, J = 6.7 Hz, 1H,
–NH), 7.2–7.7 (m, 19H, H aromatics); 13C NMR (CDCl3,
100.61 MHz) d (ppm): 57 (–CHCH2OH), 66.5 (–CH-
4. Experimental
CH2OH), 127–142 (C aromatics), 169 (–CO); MS (FAB,
m/z): 426.08 (M+1); ½aꢃD ¼ þ22:4 (c 0.41, CHCl3).
21
4.1. General information
4.3. Synthesis of (R)-N-(1-benzyl-2-hydroxyethyl)-2-
(diphenylphosphino)benzamide 5b
All reagents were obtained from Aldrich, Fluka, Alfa
Aesar or Acros. Solvents were dried using common labora-
tory methods.15
This compound was prepared in the same manner as for
5a. Giving the title compound 5b as a white solid (87%).
Mp 138.7–141.1 °C. IR [m (cmꢂ1)]: 1034.77, 1094.46,
1456.58, 1520.47, 1627.97, 2854.15, 2922.92, 3053.82,
3333.02; 1H NMR (CDCl3, 400 MHz) d (ppm): 2.2
(s, 1H, –OH), 2.6–2.8 (m, 2H, –CHCH2Ph), 3.55 (m, 2H,
–CHHOH), 3.75 (m, 2H, –CHHOH), 4.2 (m, 1H,
–CHCH2Ph), 5.8 (d, J = 6.8 Hz, 1H, –NH), 7.1–7.5 (m,
19H, H aromatics); 13C NMR (CDCl3, 100.61 MHz) d
(ppm): 37 (–CHCH2Ph), 54 (–CHCH2Ph), 64 (–CH2OH),
Column chromatography was carried out on silica gel (sds,
70–200 lm) and flash column chromatography (Merck,
40–63 lm and sds, 40–63 lm). TLC was carried out on
aluminium backed Kisel-gel 60 F254 plates (Merck). Plates
were visualised either by UV light or with phosphomolyb-
denic acid in ethanol.
High performance liquid chromatographic (HPLC) analy-
sis was performed on an Agilent 1100 series instrument.
The following conditions were used: pmax = 50 bar,
flux = 1 ml/min, detector = DAD (k = 210,10 nm), elu-
ent—n-hexane/isopropanol (98:2). The column used was
a Chiralcel OD-H (0.46 cm ꢁ 25 cm) fitted with a guard
column composed of the same stationary phase.
127–138 (C aromatics), 169 (–CO); MS (FAB, m/z):
23
440.08 (M+1). ½aꢃD ¼ ꢂ39 (c 0.55, CHCl3).
4.4. Preparation of [allylPd-5a]BF4 complex 7a and [allylPd-
5b]BF4 complex 7b
The chiral ligand (5a; 0.04 mmol, 5b; 0.09 mmol) was
added to a solution of [Pd(C3H5)Cl]2 (0.5 equiv) in CH2Cl2
under a nitrogen atmosphere. After stirring the solution for
10 min, AgBF4 (1 equiv) was added. The mixture was stir-
red for 5 min and then the precipitate (AgCl2) was removed
by filtration. The solvent was evaporated and the resultant
solid was washed with pentane and dried under vacuum.
Gas chromatographic (GC) analyses of the products were
performed on a Hewlett Packard (HP) 6890 series instru-
ment equipped with a flame ionisation detector (FID).
The chromatograph was fitted with a cyclosil-B capillary
column (30 m, 250 lm, 0.25 lm) (Agilent 112-2532).
In all cases, the acetate conversions were calculated by
determining the ratio of the peak areas for the substrate
(both enantiomers) and the alkylated product (both
enantiomers).
4.4.1. [AllylPd-5a]BF4 7a. Black solid (quantitative yield).
Mp 124.5–128.6 (°C). MS (FAB, m/z): 307.08 (M+2ꢂ
Pd+(allyl)ꢂC8H8O).
The melting points were recorded on a Barnstead Electro-
thermal 9100 apparatus and are uncorrected. The 1H NMR
spectra were recorded on either a Bruker AMX300 (1H:
300.13 MHz) instrument using CDCl3 as solvent and
TMS as internal standard. Mass spectra were recorded
on a VG Autospec M (Waters-Micromass) spectrometer
using the FAB technique. Infra-red spectra were measured
with a Perkin Elmer Paragon 1000 model.
4.4.2. [AllylPd-5b]BF4 7b. Black solid (quantitative yield).
MS (FAB, m/z): 584.20 (M+1ꢂH2O), 438.21 (M+1ꢂ
Pd+(allyl)), 305.07 (MꢂC8H8O).
4.5. Catalytic reactions
4.5.1. General procedure for the homogeneous catalytic
asymmetric allylic alkylation reactions. The ligand was
added to a solution of Pd (1–2 mol %) in solvent under a
nitrogen atmosphere. The reaction mixture was refluxed
for 2 h. A solution of rac-1,3-diphenylpropenyl acetate or
rac-3-acetoxycyclohexene (0.200 g), dimethyl malonate
(0.3 mL, 2.4 mmol), BSA (0.6 mL, 2.4 mmol) and KOAc
(0.800 mg, 1 mol %) in solvent (1 mL) was added to the
4.2. Synthesis of (S)-N-(2-hydroxy-1-phenylethyl)-2-
(diphenylphosphino)benzamide 5a
A solution of DCC (0.841 g, 4.1 mmol), DMAP (0.399 g,
3.3 mmol), amino alcohol (3.3 mol) and o-(diphenylphos-

458
V. R. Marinho et al. / Tetrahedron: Asymmetry 19 (2008) 454–458
Crosby, J.; Lloyd-Jones, G. C.; Stephen, S. C. Chem.
Commun. 1999, 1707.
4. Butts, C. P.; Crosby, J.; Lloyd-Jones, G. C.; Stephen, S. C.
Chem. Commun. 1999, 1707.
5. Amatore, C.; Jutand, A.; Mansah, L.; Ricard, L. J. Organo-
met. Chem. 2007, 692, 1457.
6. Clayden, J.; Johnson, P.; Pink, J. H.; Helliwell, M. J. Org.
Chem. 2000, 65, 7033.
catalyst solution at rt. The resulting mixture was stirred for
20 h at rt or at reflux. The reaction mixture was filtered
over a pad of Celite and silica gel and washed with
hexane/AcOEt (2:1). The solvent was evaporated and the
products were analysed by HPLC (malonate product
derived from rac-1,3-diphenylpropenyl acetate) or GC
(malonate product derived from rac-3-acetoxycyclo-
hexene).
7. Mino, T.; Kashihara, K.; Yamashita, M. Tetrahedron:
Asymmetry 2001, 12, 287.
8. (a) Konovets, A.; Glegola, K.; Penciu, A.; Framery, E.;
Jubault, P.; Goux-Henry, C.; Michal Pietrusiewicz, K.;
Quirion, J.-C.; Sinou, D. Tetrahedron: Asymmetry 2005, 16,
3183; (b) Johannesen, S. A.; Glegola, K.; Sinou, D.; Framery,
E.; Skrydstrup, T. Tetrahedron Lett. 2007, 48, 3569; (c)
Tollabi, M.; Framery, E.; Goux-Henry, C.; Sinou, D.
Tetrahedron: Asymmetry 2003, 14, 3329.
9. Jiang, B.; Huang, Z.-G. Tetrahedron lett. 2007, 48,
1703.
10. (a) Greenfield, S. J.; Agarkov, A.; Gilbertson, S. R. Org. Lett.
2003, 5, 3069; (b) Agarkov, A.; Gilbertson, S. R. Tetrahedron
Lett. 2005, 46, 181.
Acknowledgements
We are grateful for the financial support from the FCT
(project; POCI/QUI/55779/2004) through POCI 2010, sup-
ported by the European community fund, FEDER. FCT is
also acknowledged for the award of a PhD grant to
V.R.M. (SFRH/BD/25111/2005). The personnel of the
analytical service unit at C.A.C.T.I. (University of Vigo,
Spain) are also acknowledged for the chemical analyses.
11. Marques, C. S.; Burke, A. J. Tetrahedron: Asymmetry 2007,
18, 1804.
12. Bayardon, J.; Sinou, D.; Guala, M.; Desimoni, G. Tetra-
hedron: Asymmetry 2004, 15, 3195.
References
1. For a key review, see: Trost, B. M.; Lee, C. In Catalytic
Asymmetric Synthesis; Ojima, I., Ed., 2nd ed.; Wiley-VCH,
2000.
2. Trost, B. M.; Strege, P. E. J. Am. Chem. Soc 1977, 99, 1649.
3. (a) Trost, B. M.; Van Vranken, D. L. Chem. Rev. 1996, 96,
395; Trost, B. M. Acc. Chem. Res. 1996, 29, 355; (b) Trost, B.
M.; Crawley, M. L. Chem. Rev. 2003, 103, 2921; Butts, C. P.;
´
´
13. Montoya, V.; Pons, J.; Garcia-Anton, J.; Solans, X.; Font-
´
´
Bardıa, M.; Ros, J. Organometallics 2007, 26, 3183.
14. As suggested by one of the referees for this paper.
15. Armarego, W. L. F.; Perrin, D. D. Purification of
Laboratory Chemicals, 4th ed.; Butterworth Heinmann:
Oxford, 1996.