New highly effective phosphite-phosphoramidite ligands for palladium-catalysed asymmetric allylic alkylation reactions

Article abstract of DOI:10.1002/adsc.200600489

We have designed a new family of readily I available highly-modular phosphite-phosphoramidite ligands for asymmetric allylic aJkylation reactions. The introduction of a phosphoramidite moiety in the ligand design is highly advantageous in the product outcome. Thus, this ligand series affords high reaction rates and enantioselectivities and, at the same time, shows a broad scope for disubstituted hindered and unhindered substrate types.

Full text of DOI:10.1002/adsc.200600489

COMMUNICATIONS
DOI: 10.1002/adsc.200600489
New Highly Effective Phosphite-Phosphoramidite Ligands for
Palladium-Catalysed Asymmetric Allylic Alkylation Reactions
Oscar Pàmies,a,* Montserrat DiØguez,a,* and Carmen Clavera
a
Departament de Química Física i Inorgànica, Universitat Rovira i Virgili, C/Marcel·lí Domingo s/n, 43007 Tarragona,
Spain
Fax : (+34)-977-559-563; e-mail: Montserrat.dieguez@urv.cat or oscar.pamies@urv.cat
Received: September 26, 2006; Revised: December 12, 2006
Supporting information for this article is available on the WWW under http://asc.wiley-vch.de/home/.
Abstract: We have designed a new family of readily
available highly-modular phosphite-phosphorami-
dite ligands for asymmetric allylic alkylation reac-
strates and low ees are obtained in unhindered sub-
strates, or vice versa).[1]
Therefore, research into more versatile ligand sys-
tems from simple starting materials in this reaction is
of great importance nowadays. In this light, a group
of less electron-rich phosphorus homodonor com-
pounds (diphosphite ligands) has demonstrated their
usefulness in this process, overcoming these limita-
tions.[3] However, a high backbone dependence has
been observed. Thus, enantioselectivities were excel-
lent with 1,3-diphosphites (ees up to 99%), with the
1,2-diphosphites being less enantioselective (ees up to
80%) (Scheme 1).[3a]
tions. The introduction of
a phosphoramidite
moiety in the ligand design is highly advantageous
in the product outcome. Thus, this ligand series af-
fords high reaction rates and enantioselectivities
and, at the same time, shows a broad scope for di-
substituted hindered and unhindered substrate
types.
Keywords: allylic alkylation; asymmetric catalysis;
P ligands; palladium
Because of the high enantioselectivity induced by
mixed bidentate donor ligands in this process,[1] the
decision was made to design a new family of ligand
based on 1,2-diphosphite ligands [Scheme 1 (b)] in
Palladium-catalysed asymmetric allylic substitution is which one of the phosphite groups is replaced by a
a versatile, widely used process in organic synthesis phosphoramidite moiety [Scheme 1 (c)]. This new
[1]
À
for the enantioselective formation of C C bonds.
A
ligand design therefore offers the opportunity of an
large number of chiral ligands, mainly P and N li- electronic differentiation whilst maintaining a similar
gands, which possess either C2- or C1-symmetry, have spatial disposition around the metal centre. Moreover,
provided high enantiomeric excesses.[1] Most of the
chiral ligands developed for asymmetric allylic substi-
tution are mixed bidentate donor ligands (such as P-
N, P-S and S-N).[1,2] The efficiency of this type of
hard-soft heterodonor ligands has been mainly attrib-
uted to the different electronic effects of the donor
atoms. However, although to a lesser extent, homodo-
nor ligands (e.g., Trostꢀs diphosphines) have also dem-
onstrated their potential usefulness in this process
mainly based on the chiral discrimination induced by
the C2 or C1 backbone symmetry.[1] However, one dis-
advantage of using these ligand systems is that they
are often synthesised either from expensive chiral
sources or in tedious synthetic steps. Another
common disadvantage of the most successful ligands
developed for this process is that they have low reac-
tion rates and a high substrate specificity (i.e., high
ees are obtained in disubstituted linear hindered sub- Scheme 1.
836
ꢁ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2007, 349, 836 – 840

Phosphite-Phosphoramidite Ligands for Pd-Catalysed Asymmetric Allylic Alkylation
COMMUNICATIONS
Table 1. Pd-catalysed allylic alkylation of S1 using ligands 1–
4.[a]
Entry
Ligand
% Conv. (min)[b]
% ee[c]
1
2
3
4
1a
1b
1c
2a
3a
4a
1a
1a
100 (15)
100 (25)
41 (15)
100 (15)
100 (15)
100 (15)
100 (120)
100 (65)
96 (R)
92 (R)
6 (R)
95 (R)
96 (S)
95 (S)
99 (R)
96 (R)
5
6
7[d]
8[e]
[a]
All reactions were run at room temperature, 0.5 mol%
[PdCl
(h3-C3H5)]2, CH2Cl2 as solvent.
Reaction time in minutes shown in parentheses.
AHCTREUNG
[b]
[c]
Enantiomeric excesses. Absolute configuration shown in
parentheses.
T=58C.
S1/Pd=500.
[d]
[e]
Scheme 2. Synthesis of phosphite-phosphoramidite ligands
1–4.
tivities, while ligand 1c, with two unsubstituted bi-
phenyl moieties, provided almost no enantioselectivi-
the high activities obtained with diphosphite ligands[4] ty. Moreover, the substituents in the para positions of
are expected to be maintained because the phosphor- the biphenyl moieties have a slight but important
amidite moiety is also a good p-acceptor group.[5] effect on both activity and enantioselectivity (en-
These new phosphite-phosphoramidite ligands 1–4 tries 1 vs. 2). Activities and enantioselectitivies are
(Scheme 2) also provide a flexible ligand scaffold be- therefore highest with ligand 1a, which contains tert-
cause they can be easily tuned in the amino alcohol butyl groups at both ortho and para positions of the
backbone and in the biphenyl moieties to explore biphenyl moieties.
how they affect the catalytic performance. As a result,
The effect of the amino alcohol backbone was stud-
the highly enantioselective and active Pd-allylic sub- ied with ligands 1–4a (entries 1, 4–6). Ligands 1–4,
stitution reactions of several substrates with different which differ in the substituents at the carbon adjacent
steric properties are reached.
to the amino group (ligands 1–3) or in the amino
The synthesis of ligands 1–4 is straightforward group (ligand 4), provided fairly similar enantioselec-
(Scheme 2). They are synthesised very efficiently in tivities (ees up to 96%) and activities. This indicates
one step from the corresponding commercially avail- that variations in the amino alcohol backbone hardly
able, cheap 1,2-amino alcohols by reaction with affect the catalytic performance. Interestingly, when
2 equivs. of the desired in situ formed phosphoro- this new family of ligands is used both enantiomers of
chloridite in the presence of base.
the product can be accessed in high enantioselectivi-
These new ligands were first investigated in the Pd- ties simply by changing the absolute configuration of
catalysed allylic substitution of rac-1,3-diphenyl-3-ace- the amino alcohol unit (ligands 1a and 2a vs. ligands
toxyprop-1-ene S1 with dimethyl malonate as a model 3a and 4a, entries 1 and 4 vs. 5 and 6).
reaction using standard conditions (Table 1).[1] In gen-
Enantioselectivity was further improved (ees up to
eral, activities [TOFs>700 mol S1 ” (mol Pd ” h)À1] 99%) with ligand 1a by lowering the reaction temper-
and enantioselectivities (ees up to 99%) were high. ature to 58C (entry 7). The reaction was also per-
The effect of the biaryl substituents was investigat- formed at a low catalyst concentration (S1/Pd=500)
ed with ligands 1a–c (entries 1–3). It was observed using ligand 1a (entry 8). The enantioselectivity [96%
that these moieties affect both activity and enantiose- (R) ee] and activity [TOF>700 mol S1 ” (mol Pd ”
lectivity. Bulky substituents in the ortho positions of h)À1)[6] were high.
the biphenyl moieties are needed for high enantiose-
It should be pointed out that these phosphite-
lectivity (entries 1 and 2 vs. 3). Thus, ligands 1a and phosphoramidite ligands showed a much higher
1b with bulky tert-butyl groups in the ortho positions degree of enantioselectivity and higher reaction rates
of the biphenyl moieties provided high enantioselec- than their corresponding phosphinite-aminophosphine
Adv. Synth. Catal. 2007, 349, 836 – 840
ꢁ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.asc.wiley-vch.de
837

Oscar Pàmies et al.
COMMUNICATIONS
Scheme 3. Pd-catalysed allylic alkylation reaction of substrates S2–S5.
[Scheme 1 (d), ees up to 60%][7] and 1,2-diphosphite
[Scheme 1 (b), ees up to 80%][3a] analogues under
similar reaction conditions.
To further study the potential of these readily avail-
able ligands, ligand 1a was also tested in the allylic al-
kylation of the cyclic substrates S2 and S3, unhin-
dered linear substrate S4 and for the more sterically
hindered linear substrate S5. The results are summar-
ised in Scheme 3.
The enantioselectivity in unhindered cyclic S2, S3
and linear S4 substrates is usually more difficult to
control than for the hindered substrate S1, mainly be-
cause of the presence of less sterically syn substitu-
ents.[1] Although highly enantioselective catalysts have
been developed for these unhindered substrates, they
generally provide low enantiocontrol in hindered sub-
Scheme 4. Diastereoisomeric Pd-allyl intermediates for S1
with ligand 3a.
strates like S1. The development of enantioselective mers, the carbon NMRchemical shifts indicate that
catalysts for both hindered and unhindered substrates the more electrophilic allyl carbon terminus is trans
is therefore still a challenge. Interestingly, under non- to the phosphoramidite moiety. Assuming that the nu-
optimised conditions, for these sterically undemand- cleophilic attack takes place at the more electrophilic
ing substrates, high enantioselectivities (92% ee for carbon terminus and based on the observed stereo-
S2, 81% for S3 and 82% ee for S4) were also ach- chemical outcome of the reaction, 96% (S) in product
ieved (Scheme 3). These results are among the best 5, and the fact that the enantiomeric excess of 5 is
reported for these types of unhindered substrates.[8] higher than the diastereoisomeric excesses of the Pd
Again, these results indicate that the replacement of a intermediates, the A isomer must react faster than the
phosphite moiety by a phosphoramidite group in the B isomer. In addition, the reactivity of the Pd inter-
ligand design has led to higher enantioselectivities mediates with sodium malonate at low temperature
than with related 1,2-diphosphite ligands (ees up to was also studied by in situ NMR. The study indicated
39%).[3a]
that the fast reacting isomer is the major one. There-
The allylic alkylation of substrate S5, which is more fore, it can be concluded that the nucleophilic attack
sterically demanding than the previously used sub- takes place at the allyl terminus located trans to the
strate S1, also produced high enantioselectivity phosphoroamidite moiety of the major A Pd inter-
(Scheme 3). Although, as expected, the activities were mediate.
lower than in the alkylation reaction of S1, they were
In summary, a new series of easily accessible phos-
much higher than those obtained with other successful phite-phosphoramidite ligands for allylic alkylation
ligands.[1]
The NMRstudy of the Pd-allyl intermediate con-
reactions has been designed. These ligands can be
prepared in one step from commercially available 1,2-
taining ligand 3a, [Pd(h3-1,3-diphenylallyl)(3a)]BF4, amino alcohols as chiral source. The introduction of a
showed a mixture of two isomers in a ratio of 2:1. No phosphoramidite moiety in the ligand design has been
changes were observed down to À808C. Both isomers advantageous, providing higher enantioselectivities
were unambiguously assigned by NOE to the two syn/ (ees up to 99%) than with the 1,2-diphosphite and
syn endo and exo isomers (Scheme 4). For both iso- phosphinite-aminophosphine analogues. In addition,
838
www.asc.wiley-vch.de
ꢁ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2007, 349, 836 – 840

Phosphite-Phosphoramidite Ligands for Pd-Catalysed Asymmetric Allylic Alkylation
COMMUNICATIONS
enantiomeric excess was determined by HPLC (Chiralcel-
OD, 0.5% 2-propanol/hexane, flow 0.5 mLminÀ1). For sub-
strates S2–S4, conversion and enantiomeric excess were de-
termined by GC using an FS-Cyclodex b-I/P 25 m column,
internal diameter 0.2 mm, film thickness 0.33 mm, carrier
gas: 100 kPa He, F.I.D. detector). For substrate S5, conver-
the high activities were maintained. It should be
noted that both enantiomers of the product can be
obtained in high ee by using both enantiomers of the
amino alcohol backbone. The preliminary study of the
allyl intermediate indicates that the nucleophilic
attack takes place trans to the phosphoramidite
moiety.
sion was determined by 1H NMRand the enantiomeric
1
excess was determined by H NMRusing Eu
A
The combination of high enantioselectivities (ees up
to 99%), high activities and high substrate versatility
as well as the low cost and easy synthesis of the li-
gands makes these catalyst systems very attractive for
further research. These results open up the allylic al-
kylation of a wide range of substrates to the potential
effective use of readily available and highly modular
1,2-amino alcohol-based phosphite-phosphoramidite
ligands. Moreover, because of the modular construc-
tion, these ligands can be easily tuned in three regions
(biaryl substituents, backbone substituents at C-1 and
C-2, and at the amino group) to explore their effect
on catalytic performance. Studies of this kind, as well
as mechanistic studies, are currently under way.
Preparation of [Pd(h3--1,3-diphenylallyl)(3a)]BF4
Ligand 3a (54.5 mg, 0.05 mmol) and the complex [Pd(m-
A
Cl)(h3--1,3-diphenylallyl)]2 (17.3 mg, 0.025 mmol) were dis-
solved in CD2Cl2 (1.5 mL) at room temperature under
argon. AgBF4 (9.8 mg, 0.5 mmol) was added after 30 min
and the mixture was stirred for 30 min. The mixture was
then filtered over celite under argon and the resulting solu-
tion was analysed by NMR. Characterisation details are col-
lected in the Supporting Information.
Supporting Information
Experimental details for the preparation and characterisa-
tion of all new compounds and the allylic substitution reac-
tion procedures are collected in the Supporting Information.
Experimental Section
General Procedure for the Preparation of Ligands 1–
4a–c
Acknowledgements
We are indebted to Prof. P. W. N. M. van Leeuwen, Universi-
ty ofAmsterdam, ofr his comments and suggestions. We
thank the Spanish Government (Consolider Ingenio
CSD2006–0003, CTQ2004–04412/BQU and Ramon y Cajal
fellowship to O.P.) and the Generalitat de Catalunya
(2005SGR007777 and Distinction to M.D.) for financial sup-
port.
The phosphorochloridite (2.2 mmol) produced in situ was
dissolved in toluene (5 mL) and pyridine (0.36 mL,
4.6 mmol) was added. The amino alcohol (1 mmol) was
azeotropically dried with toluene (3”1 mL) and then dis-
solved in toluene (10 mL), to which pyridine (0.36 mL,
4.6 mmol) was added. The phosphorochloridite solution was
transferred slowly at 08C to the solution of the amino alco-
hol. The reaction mixture was warmed up to 808C and stir-
red overnight, and the pyridine salts were removed by filtra-
tion. Evaporation of the solvent gave a white foam, which
was purified by flash chromatography (toluene/NEt3 =100/
1) to afford the corresponding ligand as a white powder.
Yields and characterisation details of all new ligands 1–4a–c
are collected in the Supporting Information.
References
[1] For reviews, see: a) J. Tsuji, Palladium Reagents and Cat-
alysis, Innovations in Organic Synthesis Wiley, New
York, 1995; b) B. M. Trost, D. L. van Vranken Chem.
Rev. 1996, 96, 395; c) A. Pfaltz, M. Lautens, in: Compre-
hensive Asymmetric Catalysis, (Eds.: E. N. Jacobsen, A.
Pfaltz, H. Yamamoto), Springer-Verlag, Berlin, 1999,
Vol. 2, Chapter 24; d) B. M. Trost, M. L. Crawley Chem.
Rev. 2003, 103, 2921.
Typical Procedure for Allylic Alkylation of Substrates
S1–S5
A
degassed solution of [Pd(p-C3H5)Cl]2 (0.9 mg,
A
0.0025 mmol) and the corresponding phosphite-phosphora-
midite (0.0055 mmol) in the appropriate solvent (0.5 mL)
was stirred for 30 min. Subsequently, a solution of corre-
sponding substrate (0.5 mmol) in the appropriate solvent
(1.5 mL), dimethyl malonate (171 mL, 1.5 mmol), N,O-bis-
(trimethylsilyl)acetamide (370 mL, 1.5 mmol) and a pinch of
KOAc were added. The reaction mixture was stirred at
room temperature. After the desired reaction time (see Sup-
porting Information for details), the reaction mixture was
diluted with Et2O (5 mL) and saturated aqueous NH4Cl so-
lution (25 mL) was added. The mixture was extracted with
Et2O (3”10 mL) and the extract dried over MgSO4. For
[2] A. M. Masdeu-Bultó, M. DiØguez, E. Martin, M.
Gómez, Coord. Chem. Rev. 2003, 242, 159.
[3] a) M. DiØguez, O. Pàmies, C. Claver, J. Org. Chem.
2005, 70, 3363; b) M. DiØguez, O. Pàmies, C. Claver
Adv. Synth. Catal. 2005, 347, 1257.
[4] Diphosphite ligands have provided high activities, see
ref.[3]
[5] G. P. F. van Strijdonck, M. D. K. Boele, P. C. J. Kamer,
J. G. de Vries, P. W. N. M. van Leeuwen, Eur. J. Inorg.
Chem. 1999, 1073.
[6] TOF measured at around 40% conversion.
[7] L. Gong, G. Chen, A. Mi, Y. Jiang, F. Fu, X. Cui,
A. S. C. Chan, Tetrahedron: Asymmetry 2000, 11, 4297.
1
substrate S1, conversion was measured by H NMRand the
Adv. Synth. Catal. 2007, 349, 836 – 840
ꢁ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.asc.wiley-vch.de
839

Oscar Pàmies et al.
COMMUNICATIONS
[8] For some successful applications, see: a) B. M. Trost,
R. C. Bunt J. Am. Chem. Soc. 1994, 116, 4089; b) P. Di-
erkes, S. Randechul, L. Barloy, A. De Cian, J. Fischer,
P. C. J. Kamer, P. W. N. M. van Leeuwen, J. A. Osborn,
Angew. Chem. Int. Ed. 1998, 37, 3116; c) B. M. Trost,
A. C. Krueger, R. C. Bunt, J. Zambrano, J. Am. Chem.
Soc. 1996, 118, 6520; d) B. Wiese, G. Helmchen, Tetrahe-
dron Lett. 1998, 39, 5727; e) O. Pàmies, M. DiØguez, C.
Claver J. Am. Chem. Soc. 2005, 127, 3646; f) Y. Mata,
M. DiØguez, O. Pàmies, C. Claver, C. Adv. Synth. Catal.
2005, 347, 1943.
840
www.asc.wiley-vch.de
ꢁ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2007, 349, 836 – 840