Enabling ligand screening for palladium-catalysed enantioselective aza-Michael addition reactions

Article abstract of DOI:10.1002/adsc.200505404

The bis(trifluoromethanesulfonate)palladium(II) dihydrate complex, Pd(OTf)₂·2 H₂O (1), is an active palladium(II) precursor for the generation of dicationic palladium(II) catalysts. Parallel ligand screening is enabled for the first

Full text of DOI:10.1002/adsc.200505404

UPDATE
DOI: 10.1002/adsc.200505404
Enabling Ligand Screening for Palladium-Catalysed
Enantioselective Aza-Michael Addition Reactions
a
a
b
Pim Huat Phua, Andrew J. P. White, Johannes G. de Vries,
a,
King Kuok (Mimi) Hii *
Department of Chemistry, Imperial College London, Exhibition Road, South Kensington, London SW7 2AZ,
a
United Kingdom
Fax: (þ44)-207-594-5804, e-mail: mimi.hii@imperial.ac.uk
DSM Research-Life Sciences, Advanced Synthesis, Catalysis and Development, P. O. Box 18, 6160 MD Geleen,
The Netherlands
b
Received: October 14, 2005; Accepted: January 16, 2006
[
2,3]
[4]
[5]
Abstract: The bis(trifluoromethanesulfonate)palla-
dium(II) dihydrate complex, Pd(OTf) ·2 H O (1),
(2), carbamates (3) and imides (4) with excellent
yields and enantioselectivities (Scheme 1).
2
2
is an active palladium(II) precursor for the genera-
tion of dicationic palladium(II) catalysts. Parallel li-
gand screening is enabled for the first time, and
twenty-four chiral ligands were evaluated for the
asymmetric aza-Michael addition of aromatic amines
to a,b-unsaturated N-alkenoylimides and carba-
mates. Enantioselectivities of >99% can be ob-
tained. Catalytic precursors generated from 1 using
the new protocol have been identified.
Dicationic palladium complexes are generally pre-
pared by halide abstraction from the corresponding li-
[
6]
gated palladium(II) dihalides using silver salts. Occa-
[7]
sionally, the halide abstraction is performed in situ,
but this still requires the preparation of (diphos-
phine)PdX , and the presence of silver cation in the re-
2
action mixtureis not always desirable. Most importantly,
the introduction of different ligands is an expensive and
time-consuming process. As the aza-Michael reactions
are highly sensitive to the steric and electronic structures
of the substrates, we wanted to developa method for li-
gand screening that will facilitate the identification of
more active and/or selective catalysts. Herein, we de-
scribe a method of generating these catalysts in situ,
and subsequent ligand screening in these aza-Michael
addition reactions.
Keywords: asymmetric catalysis; conjugate addition;
hydroamination; ligand effects; palladium
Introduction
Enantioselectivity of a catalytic process is dependent on Results and Discussion
the attainment of a kinetically favourable transition
state, which is often highly sensitive to stereoelectronic Preliminary examination of Pd precursors for the addi-
effects exerted by the ligand and substrate(s). Invaria- tion of aniline to butenoyl-N-imide 4a was conducted us-
bly, judicious modification of a ligandꢀs structure is nec-
essary to accommodate substrate changes. The problem
is alleviated to a certain extent by the increasing com-
mercial availability of chiral ligands, as well as the ad-
vent of easy-to-synthesise modular ligands, driven and
complemented by advances in high-throughput experi-
[
1]
mentation techniques and laboratory automation. Ide-
ally, a catalyst is prepared in situ by mixing a metal pre-
cursor with the requisite ligand prior to the introduction
of substrates. However, this is only feasible if the system
does not generate competing catalytic species that can
interfere with the intended reaction pathway.
Previously, we demonstrated that [R-(BINAP)Pd-
2
þ
À
2
(
solvate) ] [TfO ] (1) catalyses the addition of aro- Scheme 1. Asymmetric aza-Michael reactions catalysed by
2
matic amines to a,b-unsaturated N-oxazolidinones palladium complex 1.
Adv. Synth. Catal. 2006, 348, 587 – 592
ꢁ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
587

UPDATE
Pim Huat Phua et al.
[
a]
Table 1. Pd precursors for the addition of PhNH to 4a.
2
mately, bis(trifluoromethanesulfonato)palladium(II)
provided a result comparable to that obtained with com-
plex 1, and the system appeared to be insensitive to the
M:L ratio (entries 8–10).
The additions of the aromatic amines to substrates 3
[
b]
[c]
Entry [Pd] precursor
Time [h] Yield [%]
ee [%]
[
8]
1
2
3
4
5
6
7
8
9
0
Pd (dba) ·CHCl 18
–
–
–
–
–
37
–
86 (84)
89
83
–
–
–
–
–
2
–
2
3
3
PdCl2
18
18
18
18
18
18
Pd(NO3)2
and 4 were subsequently re-examined and the results
PdCl (NCCH )
2
3
2
were compared to that obtained previously with the iso-
Pd(acac)2
[4,5]
lated complex (Table 2).
Indeed, the reaction out-
Pd(OAc)2
[9]
comes proved to be largely similar. In most cases, ee
enhancement was obtained with the new catalyst sys-
tem; in two cases, optically pure products can be attained
(entries 2 and 5).
Pd(TFA)2
[d]
d
Pd(OTf) ·2 H O 18
91 (89)
89
2
2
[
[
e]
f]
Pd(OTf) ·2 H O 18
2
2
1
Pd(OTf) ·2H O 18
92
2
2
These results demonstrate that Pd(OTf) ·2 H O can
2
2
[
a]
PhNH /4a/Pd(OTf) /(R)-BINAP ratio of 1/1.1/0.05/0.055
2
2
be used to generate catalytically active dicationic palla-
dium(II) triflate complexes in situ without involving ex-
traneous reagents, thus enabling parallel ligand screen-
ing. Subsequently, twenty-four commercially available
chiral ligands were evaluated in parallel, for the addition
of aniline to the Michael acceptor 4a (Figure 1). The re-
sults show that the highest product yields are generally
afforded by diphosphine ligands. Also, those with biar-
yl-derived chirality are by far the most selective: Among
at 258C in toluene.
[
[
[
b]
1
Calculated by H integration.
c]
Determined by chiral HPLC.
d]
Value in parenthesis corresponds to result afforded by
[
5]
complex 1 (ref. ).
M:L ratio of 1: 2.
M:L ratio of 1 : 0.5.
[
[
e]
f]
ing the ligand (R)-BINAP (Table 1). Unsurprisingly, these, five ligands gave enantioselectivity of x90% (BI-
none of the common palladium(0) or palladium(II) NAP, Tol-BINAP, C3-Tunephos, Difluorphos and CTH-
complexes containing strongly coordinating anionic li- P-Phos). Remarkably, the reaction is fairly insensitive to
gands was found to be active under these conditions (en- electronic effects of these biaryl ligands.
tries 1–5). Interestingly, while palladium(II) acetate in-
Other C -symmetric diphosphine ligands are less se-
2
itiated some product formation, palladium(II) trifluoro- lective (DIOP, Phanephos). The Solvias family of di-
acetate was completely inactive (entries 6 and 7). Ulti- phosphines also offered low enantioselectivities: Josi-
[a]
Table 2. Comparison of isolated complex and that generated in situ.
1
[b]
[c]
Entry
Substrate
R
Ar
Time [h]
Yield [%]
Ee [%]
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
3a
3a
3a
3b
3b
3b
3c
3c
3c
4a
4a
4a
4a
4b
4b
4b
4b
4c
4c
4c
4c
Me
Me
Me
Et
Et
Et
Pr
Pr
Pr
Me
Me
Me
Me
Et
Et
Et
Et
Pr
Ph
4-Cl-C H
4-MeO-C H
Ph
4-Cl-C H
4-MeO-C H
Ph
4-Cl-C H
4-MeO-C H
Ph
4-Cl-C H
4-Me-C H
4-MeO-C H
Ph
4-Cl-C H
4-Me-C H
18
18
40
72
72
72
120
120
120
18
18
18
18
72
72
72
72
72
72
72
72
89 (>99)
82 (>99)
38 (99)
54 (98)
61 (92)
31 (>99)
50 (98)
45 (90)
30 (82)
86 (84)
81 (88)
86 (81)
81 (70)
86 (75)
75 (70)
22 (70)
67 (76)
79 (69)
73 (78)
70 (71)
60 (69)
97 (97)
>99 (>99)
76 (73)
88 (90)
>99 (85)
19 (16)
89 (89)
92 (89)
9 (17)
91 (89)
81 (73)
85 (83)
78 (80)
80 (63)
60 (54)
79 (59)
52 (51)
68 (67)
54 (58)
68 (70)
57 (55)
6
4
6
4
4
4
6
4
6
6
4
6
1
1
1
1
1
1
1
1
1
1
2
2
6
4
6
4
6
4
4
4
6
4
6 4
4-MeO-C H
Ph
4-Cl-C H
4-Me-C H
6
Pr
Pr
Pr
6
4
6 4
4-MeO-C H
6
[
a]
ArNH /alkene/Pd(OTf) ·2 H O/(R)-BINAP ratio of 1/1.1/0.05/0.055 at 258C in toluene. Values in parenthesis correspond
2
2
2
to results obtained with complex 1.
Calculated by H integration.
Determined by chiral HPLC.
[
[
b]
1
c]
588
asc.wiley-vch.de
ꢁ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2006, 348, 587 – 592

Palladium-Catalysed Enantioselective Aza-Michael Addition Reactions
UPDATE
Figure 1. Evaluation of twenty-four commercially available ligands in the addition of aniline to 4a.
þ
phos and Walphos induced the highest conversions. One lecular composition of [Pd(BINAP)(OTf)] }, and a
of the Walphos ligands (W002) offered opposite stereo- novel bis-ligated complex 6 (d ¼ þ18 ppm), identifi-
P
control (R), suggesting a possible synergistic effect be- able by a characteristic mass ion at m/z¼1499
þ
tween the elements of chirality. Use of the aminodiphos- {[Pd(BINAP) (OTf)] }.
2
phines was less successful; Taniaphos led to lower prod-
uct yield, while Mandyphos ligands gave the lowest
yields with a complete loss of selectivity. BOX and BI-
NOL ligands, previously shown to be effective in Lewis
acid-catalysed processes, are incompatible with cationic
palladium in this instance, showing low turnover with lit-
tle/no selectivity.
ð1Þ
A solution of Pd(OTf) ·2 H O and BINAP in di-
2
2
chloromethane was found to contain two catalytic pre- The formation of the complexes is dependent on the re-
cursors [Eq. (1)], the expected mono-ligated bis-aqua action conditions – complex 5 can be isolated by per-
complex 5 (d ¼ þ34 ppm) with an associated observed forming the reaction in the presence of a small amount
P
molecular ion at m/z¼877 {corresponding to a mo- of acetonitrile. Conversely, complex 6 may be obtained
Adv. Synth. Catal. 2006, 348, 587 – 592
ꢁ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
589

UPDATE
Pim Huat Phua et al.
planar with a marked tetrahedral twist, {Pd,P(1),P(2)}
and {Pd,O(1),O(2)} being inclined by ca. 168 (the twists
for the two independent dications in NATTEH are ca. 8
and 188). With the two trans ligands being identical [i.e.,
both water, PdÀO(1) 2.141(3) and PdÀO(2) 2.138(3) Š],
the PdÀP binding of the binapligand is symmetric
[
PdÀP(1) 2.2376(11), PdÀP(2) 2.2367(11) Š]; the bite
angle is 91.39(4)8. By contrast, in each of the two inde-
pendent dications in NATTEH the bonding is more
asymmetric with PdÀP distances of 2.183(3) and
2
2
.236(3) Š and PdÀO bond lengths of 2.18(1) and
.26(1) Š for dication A, whilst for dication B the
PdÀP distances are 2.177(3) and 2.241(3) Š with PdÀO
bond lengths of 2.18(1) and 2.24(1) Š.
The single crystal X-ray analysis of 6 revealed a chiral
structure (the absolute structure was determined unam-
biguously from the X-ray data; see Experimental Sec-
tion) with two R-BINAP ligands coordinated to a dis-
torted square planar palladium centre (Figure 3), the
two triflate counterions being non-coordinating. The di-
Figure 2. The molecular structure of 5. The aqua· · ·triflate
hydrogen bonds have O· · ·O separations (Š) of a)
2.716(6), b) 2.612(14), c) 2.716(6) and d) 2.994(10) Š.
Table 3. Selected bond lengths (Š) and angles (8) for 5.
cation has molecular C symmetry about an axis that
2
PdÀO(1)
2.141(3)
PdÀO(2)
2.138(3)
passes through the metal centre and bisects the naphtha-
PdÀP(1)
2.2376(11)
88.89(13)
90.06(10)
168.91(11)
PdÀP(2)
2.2367(11) lene-naphthalene bond in each of the diphosphine li-
O(1)ÀPdÀO(2)
O(1)ÀPdÀP(2)
O(2)ÀPdÀP(2)
O(1)ÀPdÀP(1)
O(2)ÀPdÀP(1)
P(1)ÀPdÀP(2)
167.98(11)
91.93(10)
91.39(4)
gands. Interestingly, the PdÀP coordination distances
for the two BINAP ligands are noticeably different (Ta-
ble 2); whereas the P(1)/P(2) ligand coordinates sym-
metrically [PdÀP(1) 2.4271(13), PdÀP(2) 2.4352(14)
Š], the coordination of the P(3)/P(4) ligand is distinctly
asymmetric with one bond shorter and one longer than
Table 4. Selected bond lengths (Š) and angles (8) for 6.
those seen for the P(1)/P(2) ligand [PdÀP(3)
PdÀP(1)
2.4271(13)
2.4678(13)
86.43(5)
172.21(5)
94.10(5)
PdÀP(2)
2.4352(14) 2.4043(14), PdÀP(4) 2.4678(13) Š]. It is not immediately
PdÀP(3)
PdÀP(4)
2.4043(14) apparent why this should be the case. [The associated
P(1)ÀPdÀP(2)
P(1)ÀPdÀP(4)
P(2)ÀPdÀP(4)
P(1)ÀPdÀP(3)
P(2)ÀPdÀP(3)
P(3)ÀPdÀP(4)
94.80(4)
170.31(4)
85.99(4)
PÀPdÀP bite angles for the two chelating BINAP li-
gands are 86.43(5) and 85.99(4)8 for the P(1)/P(2) and
P(3)/P(4) ligands, respectively.] There is a noticeable
tetrahedral twist in the metal coordination plane, the
{
Pd,P(1),P(2)} and {Pd,P(3),P(4)} planes being inclined
by employing two equivalents of the diphosphine ligand. by ca. 138.
The structures of both these dicationic palladium com-
plexes were subsequently determined by X-ray crystal- ic activity of complexes 5 and 6 were assessed for the ad-
lography.
Finally, in the last part of the present study, the catalyt-
The racemic diaquo complex 5 adopts a solid state
structure where both of the non-coordinated triflate
counterions hydrogen bond to both of the aqua ligands
[10]
(
Figure 2). The structure of this dication has been de-
termined previously as the optically pure BF salt (NAT-
4
[11]
TEH). In this literature structure, there were two in-
dependent dications and so four independent tetrafluor-
oborate anions, all four of which were disordered. How-
ever, it was still evident that these anions were hydrogen
bonding to the aqua ligands, the shortest of the O· · ·F
contacts being ca. 2.67 Š. Here in 5 the dication has ap-
proximate C symmetry about an axis that passes
2
through the metal centre and bisects the naphthalene-
naphthalene bond of the BINAP ligand, although the
hydrogen-bonded triflate anions break this symmetry. Figure 3. The molecular structure of the dication in 6 (the tri-
The geometry at the palladium centre is distorted square flate counterions are not shown).
590
asc.wiley-vch.de
ꢁ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2006, 348, 587 – 592

Palladium-Catalysed Enantioselective Aza-Michael Addition Reactions
UPDATE
dition of aniline to 4a. Rather surprisingly, identical re-
sults were obtained – complex 5 gave 88% yield, 89%
ee, whilst complex 6 afforded the product in 87% yield
and 88% ee. Given that these results are similar to that
obtained with complex 1, this observation suggests
that complex 6 undergoes dissociation of one of the BI-
NAP ligands, to generate the catalytically active mono-
ligated complex 5. This is supported by the X-ray crystal
Crystal data for 6: [C88
H P Pd](CF SO ) ·0.5 CH Cl ·
64 4 3 3 2 2 2
0
1
Š
1
3
.5 Et O, M¼1729.34, orthorhombic, P2 2 2 (no. 18), a¼
2
1 1
9.7973(5), b¼30.2266(8), c¼13.3181(3) Š, V¼7969.6(3)
3
À3
À1
, Z¼4, D ¼1.441 g·cm , m(Mo-Ka)¼0.470 mm , T¼
c
73 K, yellow block-like needles, Oxford Diffraction Xcalibur
2
diffractometer; 27409 independent measured reflections, F
refinement, R ¼0.079, wR ¼0.134, 15615 independent ob-
1
2
served absorption-corrected reflections [jF j >4s(jF j),
o
o
2
qmax ¼668], 1061 parameters. The absolute structure of 6 was
þ
data, which suggest weaker binding of one of the BINAP determined by a combination of R-factor tests [R ¼0.0785,
1
À
þ
ligands in complex 6.
R1 ¼0.0811] and by use of the Flack parameter [x ¼
À
þ0.02(2), x ¼ þ0.98(2)]. CCDC 286581.
Crystallographic data (excluding structure factors) for the
structures reported in this paper have been deposited with
the Cambridge Crystallographic Data Centre. Copies of the
data can be obtained free of charge on application to CCDC,
Conclusion
1
2 Union Road, Cambridge CB2 1EZ, UK [Fax: int. code
In summary, we have demonstrated that Pd(OTf)2 ·
þ44(1223)336–033; E-mail: deposit@ccdc.cam.ac.uk].
2
H O can be used as a precursor in the rapid constitu-
2
tion of asymmetric hydroamination catalysts. Hence,
identification of enantioselective ligands may be accom-
plished by parallel screening. In this work, important
structural aspects for high enantioselectivity were eluci-
Acknowledgements
dated for the aza-Michael addition of aromatic amines The authors thank The Committee of Vice-Principals and Chan-
to N-imide, and different catalytic precursors were iso- cellors (CVCP) for an Overseas Research Studentship Award to
lated and identified.
PHP, as well as Imperial College London and DSM Research
for additional financial support. Palladium salts are provided
by Johnson Matthey plc. A Solvias ligand kit (containing Josi-
phos, Mandyphos, Taniaphos and Walphos) was kindly donated
by Professor H.-U. Blaser.
Experimental Section
Typical Catalytic Reaction
References and Notes
In a glove box, Radleyꢀs reaction tubes were charged with
[12]
Pd(OTf) ·2 H O (6 mg), an appropriate amount of the cor-
2
2
[
1] See, for example: a) T. Satyanarayana, H. B. Kagan, Adv.
Synth. Catal. 2005, 347, 737; b) K. Ding, H. Du, Y. Yuan,
J. Long, Chem. Eur. J. 2004, 10, 2872; c) A. Duursma, L.
Lefort, J. A. F. Boogers, A. H. M. de Vries, J. G. de Vries,
A. J. Minnaard, B. L. Feringa, Org. Biomol. Chem. 2004,
responding ligand and a magnetic stirrer. Each tube was then
fitted with a PTFE screw capwith integrated gas inlet valve
and fitted rubber septum, placed in the carousel, and purged
and filled successively with dry N . Dry toluene (0.5 mL) was
2
introduced, and the solution of the catalytic precursor was stir-
red for 30 min. The temperature of the reaction block was
maintained at 258C by means of a thermostat, before the addi-
tion of the Michael acceptor, amine and additional amount of
toluene (0.5 mL). Periodically, small amounts of reaction ali-
2
, 1682; d) L. Lefort, J. A. F. Boogers, A. H. M. de Vries,
J. G. de Vries Org. Lett. 2004, 6, 1733; e) J. G. de Vries,
A. H. M. de Vries, Eur. J. Org. Chem. 2003, 799; f) C.
Gennari, U. Piarulli, Chem. Rev. 2003, 103, 3071;
g) M. T. Reetz, Angew. Chem. Int. Ed. 2001, 40, 284.
1
quots were extracted and analysed by H NMR spectroscopy.
At the end of the reaction, the solvent was evaporated, and
the residue was subjected to column chromatography. Charac-
terisation data of the products are contained in previous re-
ports.
[2] a) K. Li, K. K. Hii, Chem. Commun. 2003, 1132; b) P. H.
Phua, K. Li, K. K. Hii, Tetrahedron 2005, 61, 6237.
[
3] The enantioselective addition of aromatic amine salts to
can also be achieved effectively using a dimeric Pd
2
complex: Y. Hamashima, H. Somei, Y. Shimura, T. Ta-
mura, M. Sodeoka, Org. Lett. 2004, 6, 1861.
[4] K. Li, X. Cheng, K. K. Hii, Eur. J. Org. Chem. 2004, 959.
X-Ray Crystallographic Study
[
[
[
5] P. H. Phua, J. G. de Vries, K. K. Hii, Adv. Synth. Catal.
005, 347, 1775.
6] K. Li, P. N. Horton, M. B. Hursthouse, K. K. Hii, J. Orga-
nomet. Chem. 2003, 665, 250.
7] For example, see: Y. Uozumi, K. Kato and T. Hayashi,
Tetrahedron Asymmetry 1998, 9, 1065.
Crystal data for 5: [C H O P Pd](CF SO ) ·0.5 CH Cl , M¼
4
4
36
2
2
3
3
2
2
2
2
¯
1
105.67, triclinic, P1 (no. 2), a¼11.2008(8), b¼13.4496(12), c¼
6.6057(12) Š, a¼98.787(7), b¼105.486(6), g¼98.946(7)8,
1
3
À3
V¼2331.3(3) Š , Z¼2, D ¼1.575 g·cm
,
m(Mo-Ka)¼
c
À1
0
.690 mm , T¼173 K, yellow plate-like needles, Oxford Dif-
fraction Xcalibur 3 diffractometer; 15356 independent meas-
2
ured reflections, F refinement, R ¼0.101, wR ¼0.266, 13562
[8] As far as we are aware, Pd(OTf) ·2 H O has only been
2 2
1
2
independent observed absorption-corrected reflections [jF j
directly employed once before in an asymmetric aldol
process, but yield and ee were very low: A. Yanagisawa,
o
>
4s(jF j), 2q ¼668], 625 parameters. CCDC 266252.
o
max
Adv. Synth. Catal. 2006, 348, 587 – 592
ꢁ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
591

UPDATE
Pim Huat Phua et al.
K. Kimura, Y. Nakatsuka, Y. Yamamoto, Synlett 1998,
[10] Similar hydrogen bonding motif has been previously ob-
served in analogous achiral bis(aqua)palladium(II) tri-
flate complexes: P. J. Stang, D. H. Cao, G. T. Poulter,
A. M. Arif, Organometallics 1995, 14, 1110.
[11] M. Sodeoka, R. Tokunoh, F. Miyazaki, E. Hagiwara, M.
Shibasaki, Synlett 1997, 463.
9
58.
[
9] Addition to N-Boc carbamate esters in the present study
were conducted in a lower amine:3 ratio (1:1.1) than that
employed previously (1:1.5), which could account for
the lower yields. The limited solubility of Pd(OTf) in
2
toluene could also be a contributing factor.
[12] S. Murata, Y. Ido, Bull. Chem. Soc. Japan 1994, 67, 1746.
592
asc.wiley-vch.de
ꢁ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2006, 348, 587 – 592