Synthesis of Versatile Chiral N,P Ligands Derived from Pyridine and Quinoline

Article abstract of DOI:10.1002/anie.200352755

Potent transition-metal complexes for asymmetric catalysis are formed from readily accessible N,P ligands constructed from basic N-heteroaryl building blocks (see structure). Their simple assembly should not be misconstrued: they posses several handles by which to tune both steric and electronic parameters. The potential of these ligands is demonstrated by the high levels of enantioselection they induce in such divergent processes as asymmetric hydrogenation and the Heck reaction.

Full text of DOI:10.1002/anie.200352755

Communications
Chiral N,P Ligands
Synthesis of Versatile Chiral N,P Ligands Derived
from Pyridine and Quinoline**
William J. Drury III, Nicole Zimmermann,
Martine Keenan, Masahiko Hayashi, Stefan Kaiser,
Richard Goddard, and Andreas Pfaltz*
Chiral phosphanyl oxazolines 1 (PHOX)[1] and related com-
pounds such as 2[2] and 3[3] have established themselves as
highly versatile, readily accessible ligands for controlling the
stereochemistry of metal-catalyzed reactions. The success of
these ligands in the Ir-catalyzed hydrogenation of olefins[2–4]
and imines,[5] and the intention to mimic the coordination
sphere of Crabtree's catalyst ([(Cy3P)(pyridine)Ir(cod)]PF6)[6]
(Cy = cyclohexyl, cod = 1,5-cyclooctadiene) with a chiral
bidentate ligand, motivated us to prepare chiral pyridyl
phosphanes having the general structure 4.[7] On the basis of
simple force-field calculations carried out with a fixed
standard geometry for the P-Ir-N core and the known
propensity of 2-(oxymethyl)pyridines to adopt an antiperi-
planar conformation of the N-C-C-O fragment,[8] we antici-
pated that these ligands would form a rigid chelate ring in a
[*] Dr. W. J. Drury III, Dr. N. Zimmermann, Dipl.-Chem. S. Kaiser,+
Prof. Dr. A. Pfaltz
Department of Chemistry
University of Basel
St. Johanns-Ring 19, 4056 Basel (Switzerland)
Fax: (+41)61-267-1103
E-mail: andreas.pfaltz@unibas.ch
Dr. M. Keenan, Dr. M. Hayashi, Dr. R. Goddard+
Max-Planck-Institut für Kohlenforschung
Kaiser-Wilhelm-Platz 1
45470 Mülheim an der Ruhr (Germany)
[+] X-Ray analyses.
[**] Financial support for this work from the American and Swiss
National Science Foundations is gratefully acknowledged. N.Z.
thanks the Fonds der Chemischen Industrie for a KekulØ fellowship.
The authors kindly thank Dr. Thomas Lectka (The Johns Hopkins
University) for helpful discussions.
Supporting information for this article is available on the WWW
under http://www.angewandte.org or from the author.
70
ꢀ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
DOI: 10.1002/anie.200352755
Angew. Chem. Int. Ed. 2004, 43, 70 –74

Angewandte
Chemie
boat conformation (6), thus generating a similar steric
environment around the Ir atom as the PHOX ligands.
However, as oxazoline and pyridine ligands are expected to
have different electronic effects on a coordinated metal, we
thought that pyridine-derived ligands, such as 4, could induce
new patterns of reactivity and selectivity. We report here the
synthesis and application of a series of pyridyl phosphanes 4
along with the analogous quinoline derivatives and pyridyl
phosphinites 5.
line analogue of 8 with Ipc2BCl was sluggish and gave only
70% ee. Therefore, an alternative route based on the Sharp-
less dihydroxylation[12] of 2-vinylquinoline[13] (12) was devised
(Scheme 2).
We felt that the ready availability of all starting materials
was of utmost importance. With this in mind, we chose ethyl
picolinate (7) as an appropriate entry point (Scheme 1). This
Scheme 1. Synthesis of pyridyl phosphanes 11a–c. a) Ph2PCH3·BH3
(1 equiv), TMEDA(1.1 equiv), sBuLi (1.1 equiv), À788C, 2 h, then 7
(1.1 equiv), À788C to RT, 18 h; b) (À)-Ipc2BCl (2.5 equiv), THF, 08C,
19 h; c) tBuMe2SiOTf (1.4 equiv), 2,6-lutidine, CH2Cl2, RT, 2 h, (after
recrystallization, 59%); d) (Bu4N)F (2 equiv), THF, RT, 4 h, 99%;
e) (iPr)3SiOTf (1.4 equiv), 2,6-lutidine, CH2Cl2, RT, 18 h, 94%; or
tBuPh2SiCl (1.25 equiv), imidazole (4.3 equiv), CH2Cl2, 80%; f) diethyl-
amine, RT, 12 h, 11a (63%), 11b (81%), 11c (89%). TMEDA=
N,N,N’,N’-tetramethylethylenediamine, Tf=trifluoromethanesulfonyl.
Scheme 2. Synthesis of quinolyl phosphanes 17a–c. a) K2OsO4·H2O
(1 mol%), (DHQD)2PHAL (1 mol%), K3[Fe(CN)6] (3 equiv), K2CO3
(3 equiv), tBuOH/H2O (1:1), 08C to RT; b) TsCl (1 equiv), pyridine,
À108C, 4 h; c) tBuMe2SiCl (1.3 equiv), imidazole (2.5 equiv), CH2Cl2,
RT, 5 h, (after recrystallization, 74%); d) Ph2PH·BH3 (1.2 equiv), nBuLi
(1.2 equiv), THF, À788C to 08C, 1 h, then 15, THF/DMF (3:2), RT,
24 h; e) (Bu4N)F (2 equiv), THF, RT, 4 h, 99%; f) (iPr)3SiOTf
(1.4 equiv), 2,6-lutidine, CH2Cl2, RT, 18 h, 99%; or tBuPh2SiCl
(1.25 equiv), imidazole (4.3 equiv), CH2Cl2, 79%; g) diethylamine, RT,
12 h, 17a (69%), 17b (86%), 17c (96%). (DHQD)2PHAL=hydroqui-
nidine-1,4-phthalazinediyl diether, Ts=toluene-4-sulfonyl.
was readily alkylated by lithiated BH3-protected methyldi-
phenylphosphane to yield ketone 8.[9] This material could be
reduced to 9 by (À)-chlorodiisopinocampheyl borane [(À)-
Ipc2BCl],[10] in good yield and high enantioselectivity. As this
product is an oil, we directly carried it forward in hopes of
recrystallizing a later intermediate. We capped the benzyl
alcohol by straightforward protection with tBuMe2SiOTf to
provide 10a, in which the ethereal oxygen atom has been
rendered noncoordinating. As this material could be obtained
enantiomerically pure after recrystallization, it was chosen as
the common intermediate for the construction of all other
analogues. At this point, it was unclear what role the silane
might play during catalysis. To gain insight, we cleaved the
silyl group and reprotected the alcohol with either (iPr)3-
SiOTf (10b) or tBuPh2SiCl (10c). Finally, the free phosphanes
11a–c were obtained by stirring the corresponding boranes in
diethylamine overnight, and filtering the reaction mixture
through silica gel.[11]
The oxidation of 12 reproduciblity provided diol 13 in
acceptable yield and high ee value on a multigram scale (46%,
94% ee).[14] Selective tosylation of the primary alcohol,
followed by silylation provided enantiomerically pure 15
after recrystallization from ethyl acetate/hexanes. Sulfonate
displacement by LiPPh2·BH3 gave the protected ligand 16a.
At this point, the overall diversity of the ligand set can be
expanded by judicious choice of nucleophile, and we have
obtained similar results with di-o-tolyl, di-m-xylyl, and
dicyclohexylphosphane–borane complexes.[15] As outlined
for the pyridyl phosphanes, the free ligands 17a–c were
obtained by exchange of the silyl group and deborination.
Cationic Ir complexes were made from N-heteroaryl
phosphanes 11a–c and 17a–c. Heating a dichloromethane
solution of the appropriate ligand in the presence of
0.5 equivalent of [Ir(cod)Cl]2 for 2 h followed by counterion
exchange with sodium tetrakis[3,5-bis(trifluoromethyl)phe-
nyl]borate (NaBArF)[16] (1.6 equiv), in the presence of water,
To increase the steric bulk of the coordinating N-hetero-
cycle, we wanted to replace the pyridine ring by a quinoline
unit. However, the enantioselective reduction of the quino-
Angew. Chem. Int. Ed. 2004, 43, 70 –74
www.angewandte.org
ꢀ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
71

Communications
provided the complexes as orange solids. These are generally
with pyridyl alkoxides were unsuccessful. However, N,N-
diethylaminodiaryl phosphanes in the presence of acidic
heterocycle salts[20] led to the desired phosphinites in all but
air and moisture stable, as well as being amenable to
chromatography on silica gel (dichloromethane/pentanes).[17]
We were able to obtain crystals of (S)-19a suitable for X-
ray analysis.[18] The crystal structure confirmed the absolute
configuration originally assigned to ligands 17a–c and 11a–c,
based on heuristic rules for Ipc2BCl reduction[10] and asym-
metric dihydroxylation.[12] It also showed the expected boat
conformation of the chelate ring (see structure 6). We also
À
determined the crystal structures of racemic PF6 analogues
of 19b and 19c,[18] which show a much higher tendency to
crystallize than the corresponding enantiomerically pure
BArF salts. The chelate ring and the silyloxy groups in these
two complexes, and in complex (S)-19a, adopt virtually the
same geometry. The chiral equatorial/axial arrangement of
À
the two P-phenyl groups is very similar to that found in Ir
PHOX complexes (see Figure 1).[5] Moreover, space-filling
models show that the quinoline and the isopropyloxazoline
moieties occupy similar regions in space. The striking
similarity between the chiral coordination spheres in the
quinolyl phosphane and PHOX complexes suggests that,
although the ligand structures are distinctly different, they
should allow comparable levels of enantiocontrol.
Scheme 3. Synthesis of pyridyl phosphinite complexes 21a–k.
a) Ar2PNEt2/4,5-dichloroimidazole/triethylamine (1:1:1; 1.5—
3.0 equiv), CH2Cl2, RT, 1–3 d; b) Alk2PCl (1 equiv), KH (1.1 equiv),
Et2O/DMF (9:1), 08C to RT, 18 h; c) [Ir(cod)Cl]2 (0.5 equiv), CH2Cl2,
508C, 2 h, then NaBArF (1.5 equiv), 508C, 30 min; d) [Ir(cod)Cl]2
(0.5 equiv), CH2Cl2, 0–508C, 2 h, then NaBArF (1.5 equiv), 508C,
30 min.
In the three crystal structures mentioned above, the silyl
groups are bent toward the coordination sphere. As shown in
Figure 1, one of the Si-phenyl groups is located above the
Ir atom (distance between the essentially parallel Ph and P-Ir-
N planes ca. 3.7 Š), close enough to interact with the active
the most congested systems.[21] The products were then
isolated by column chromatography, under argon, in greater
than 97% purity (31P NMR). In contrast to their aryl
conjoiners, dialkyl chlorophosphanes readily react with
pyridyl alkoxides. The instability of the resultant phosphinites
meant that it was imperative to carefully control reagent
stoichiometry, and when this was done simple filtration of the
reaction mixture through degassed silica provided > 90%
pure material. The corresponding Ir complexes were made by
a slightly modified procedure, because of the higher ligand
sensitivity. Again, the resultant complexes are stable red to
orange solids that can be purified by chromatography or
crystallization.
Figure 1. Comparison of [Ir(1)(cod)]PF6 (R1 =iPr, R2 =Ph) and the PF6
À
analogue of rac-19c. Hydrogen atoms, cod, and PF6À ions in both
structures have been omitted for clarity.
site of the catalyst. Therefore, it is not surprising that in
catalytic studies the enantioselectivity was found to depend
on the structure of the silyl group (see below).
The success with phosphinite ligands 3[3] prompted us to
replace the R2PCH2 fragment in ligands 4 by a R2PO group
(Scheme 3). The requisite alcohols were readily prepared by
reduction of commercially available ketones (20a, b) with
sodium borohydride or alkylating the appropriate aldehyde
with 2-lithiopyridine (20c, d). By this route, a diverse set of
ligands with many different R groups, ranging from simple
methyl to colossal trityl, could be readily prepared. The
racemic alcohols were then resolved into their component
enantiomers by chiral HPLC (see Supporting Information).[19]
To our surprise, all attempts to react chlorodiarylphosphanes
The iridium complexes synthesized above were used to
hydrogenate several representative alkenes. The efficacy of
the complexes in the hydrogenation of trans-a-methylstilbene
(22) are summarized in Table 1. In general, the phosphinites
are superior in terms of both enantioselectivity and con-
version. In the case of the pyridyl phosphanes (18a–c) the
nature of the silane seems to play a disproportionately large
role. This can be rationalized in light of the crystal structures
discussed above. High enantioselectivity was also obtained
with other substrates such as monoaryl alkenes 24 and 25,
72
ꢀ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
Angew. Chem. Int. Ed. 2004, 43, 70 –74

Angewandte
Chemie
Table 1: Iridium-catalyzed hydrogenation of 22.[a]
The results presented indicate considerable
potential for these structurally simple and readily
available heteroaryl phosphanes and phosphinites.
Their modular nature, which enables easy tuning
of steric and electronic properties, suggests that
these ligands should be widely applicable in
asymmetric catalysis.
Catalyst
Configuration
Conv. (%)[b]
ee (%)[c]
18a
18b
18c
19a
19b
19c
21a
21b
21c
21d
21e
21 f
21g
21h
21i
R
R
R
R
R
R
S
S
R
S
S
R
S
R
S
S
R
93
>99
>99
80
92
69
>99
>99
>99
>99
>99
>99
>99
>99
>99
>99
44
88 (R)
57 (R)
4 (R)
Received: September 1, 2003 [Z52755]
45 (R)
45 (R)
56 (R)
90 (S)
93 (S)
94 (R)
95 (S)
87 (S)
96 (R)
90 (S)
96 (R)
96 (S)
97 (S)
38 (R)
Keywords: asymmetric catalysis · Heck reaction ·
.
hydrogenation · ligand design · N,P ligands
[1] a) G. Helmchen, A. Pfaltz, Acc. Chem. Res. 2000,
33, 336.
[2] a) F. Menges, M. Neuburger, A. Pfaltz, Org. Lett.
2002, 4, 4713; for related ligands, see: b) C. A.
Busacca, D. Grossbach, R. C. So, E. M. O'Brien,
E. M. Spinelli, Org. Lett. 2003, 5, 595.
[3] a) J. Blankenstein, A. Pfaltz, Angew. Chem. 2001,
113, 4577; Angew. Chem. Int. Ed. 2001, 40, 4445;
b) F. Menges, A. Pfaltz, Adv. Synth. Catal. 2002,
344, 40.
21j
21k
[a] Conditions: see Equation (1); experimental procedures: see Ref. [3]. [b] Deter-
mined by GC.[3] [c] Determined by HPLC.[3]
[4] A. Pfaltz, J. Blankenstein, R. Hilgraf, E. Hörmann,
S. McIntyre, F. Menges, M. Schönleber, S. P. Smidt,
B. Wüstenberg, N. Zimmermann, Adv. Synth. Catal.
2003, 345, 33.
[5] P. Schnider, G. Koch, R. PrØtôt, G. Wang, F. M.
Krüger, A. Pfaltz, Chem. Eur. J. 1997, 3, 887.
[6] R. H. Crabtree, Acc. Chem. Res. 1979, 12, 331.
[7] For other chiral pyridine and quinoline derived N,P
ligands, see: a) T. Bunlaksananusorn, K. Polborn, P.
Knochel, Angew. Chem. 2003, 115, 4071; Angew.
Chem. Int. Ed. 2003, 42, 3941; b) P. Espinet, K.
allylic alcohol 26, and ethyl 2-methylcinnamate 27. Of
particular note is the tetrasubstituted alkene 28, a substrate
Soulantica, Coord. Chem. Rev. 1999, 293, 499; c) J. M. Brown,
D. I. Hulmes, P. J. Guiry, Tetrahedron 1994, 50, 4493; d) B. Wolfe,
T. Livinghouse, J. Am. Chem. Soc. 1998, 120, 5116; e) K. Ito, R.
Kashiwagi, K. Iwasaki, T. Katsuki, Synlett 1999, 10, 1563; f) G.
Franciò, F. Faraone, W. Leitner, Angew. Chem. 2000, 112, 1486;
Angew. Chem. Int. Ed. 2000, 39, 1428; g) A. V. Malkov, M. Bella,
that gives low conversion and poor enantioselectivity with
most catalysts. Overall, the complexes rival the best catalysts
developed so far, and in some cases match the highest
reported ee values.
To further demonstrate the generality and utility of N-
heteroaryl phosphanes and phosphinites we tested the ligands
as catalyst precursors for several palladium-catalyzed proc-
esses. While the phosphinites formed rather unselective
complexes in this regard, the phosphanes proved to be far
superior, especially in the enantioselective Heck reaction
(Scheme 4). While the product mixtures contain slightly more
of isomer 30 than when PHOX complexes are used, the
enantiomeric excess equaled the highest reported values.[22]
´
I. G. Starµ, P. Ko›ovsky, Tetrahedron Lett. 2001, 42, 3045.
[8] C. Moberg, H. Adolfsson, K. Wärnmark, P. O. Norrby, K. M.
Marstokk, H. Møllendal, Chem. Eur. J. 1996, 2, 516.
[9] See, for example, a) T. Imamoto, T. Kusumoto, N. Suzuki, K.
Sato, J. Am. Chem. Soc. 1985, 107, 5301; b) T. Imamoto, T.
Oshiki, T. Onozawa, T. Kusumoto, K. Sato, J. Am. Chem. Soc.
1990, 112, 5244.
[10] H. C. Brown, J. Chandrasekharan, P. V. Ramachandran, J. Am.
Chem. Soc. 1988, 110, 1539.
[11] a) R. Ewalds, E. B. Eggeling, A. C. Hewat, P. C. J. Kamer,
P. W. N. M. van Leeuwen, D. Vogt, Chem. Eur. J. 2000, 6, 1496 .
[12] H. C. Kolb, M. S. VanNieuwenhze, K. B. Sharpless, Chem. Rev.
1994, 94, 2483.
[13] 2-Vinylquinoline is commercially available from Acros Chem-
ical, or alternately, can be prepared from quinoline in one step,
see: M. A. Fakhfakh, X. Franck, A. Fournet, R. Hocquemiller, B.
Figad›re, Tetrahedron Lett. 2001, 42, 3847.
[14] Dihydroxylation of 2-vinylpyridine afforded the corresponding
diol in only 15% yield (94% ee).
[15] N. Zimmermann, Dissertation, University of Basel, 2001.
[16] D. L. Reger, T. D. Wright, C. A. Little, J. J. S. Lambda, M. D.
Smith, Inorg. Chem. 2001, 40, 3810. See also:J. L. Leaser, Jr., R.
Cvetovich, F.-R. Tsay, U. Dolling, T. Vickery, D. Bachert, J. Org.
Chem. 2003, 68, 3695.
Scheme 4. Enantioselective Heck reaction. dba=trans,trans-dibenzyl-
ideneacetone, proton sponge=1,8-bis(dimethylamino)naphthalene.
Angew. Chem. Int. Ed. 2004, 43, 70 –74
www.angewandte.org
ꢀ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
73

Communications
[17] For long-term storage (> 2 weeks), complexes were kept under
argon at À258C.
[18] CCDC-217683 (19a), CCDC-218218 (rac-19b), CCDC-218219
(rac-19c) contains the supplementary crystallographic data for
this paper. These data can be obtained free of charge via
www.ccdc.cam.ac.uk/conts/retrieving.html (or from the Cam-
bridge Crystallographic Data Centre, 12, Union Road, Cam-
bridge CB21EZ, UK; fax: (+ 44)1223-336-033; or deposit@
ccdc.cam.ac.uk). rac-19c: crystallized from CH2Cl2/Et2O,
C47H50F6Ir1N1O1P2Si1, Mr = 1041.16, orange block 0.20 ” 0.22 ”
¯
0.24 mm, triclinic, space group P111, a = 10.9930(8), b =
12.8619(7), c = 15.8814(12) Š, a = 98.936(9), b = 104.955(6),
g = 95.555(7)8, V= 2120.9(3) Š3, 1calcd = 1.630 gcmÀ3
,
m =
3.316 mmÀ1
, F(000) = 1044.000, Z = 2, MoKa radiation (l =
0.71073 Š, graphite monochromator), T= 173 K, V scan range
3.168 < V< 33.0008. 108123 reflections collected (Æ h, Æ k, Æ l),
15971 independent reflections, 12910 observed reflections [I >
3s(I)], 532 refined parameters. R = 0.0345, Rw = 0.0242, max.
residual electron density À2.55(3.06) eÀ ŠÀ3, located near the
metal center. There were no restraints applied. Hydrogen atoms
were calculated and refined as riding atoms. Data collection was
performed with a Kappa CCD diffractometer. Programs used:
structure solution: SIR79 (A. Altomare, M. C. Burla, M.
Camalli, G. L. Cascarano, C. Giacovazzo, A. Guagliardi, A.
Grazia, G. Moliterni, G. Polidori, R. Spagna, J. Appl. Crystallogr.
1999, 32, 115), structure refinement: CRYSTALS (D. J. Watkin,
Crystals, Issue 11, Chemical Crystallography Laboratory,
Oxford, 2001), weighting scheme: Chebychevpolynomial (J. R.
Carruthers, D. J. Watkin, Acta Crystallogr. Sect. A 1979, 35, 698).
We thank Markus Neuburger for assistance in solving the crystal
structures of rac-19b and rac-19c.
[19] While we have chosen to separate the enantiomers on expe-
diency grounds, this is by no means necessary as enantioselective
routes to all but one are known. 20a: a) B. Lipshutz, L. Asher, K.
Noson, Org. Lett. 2002, 4, 4045; b) R. Seemayer, M. P. Schneider,
Tetrahedron: Asymmetry 1992, 3, 827; c) M. D. Garret, R. Scott,
G. N. Sheldrake, Tetrahedron: Asymmetry 2002, 13, 2201; 20b:
d) A. G. Davies, J. Kenyon, K. Thaker, J. Chem. Soc. 1956, 3394;
e) N. A. Salvi, S. Chattopadhyay, Tetrahedron 2001, 57, 2833;
20c: f) C. Bolm, M. Zehnder, D. Bur, Angew. Chem. 1990, 102,
206; Angew. Chem. Int. Ed. Engl. 1990, 29, 205; g) C. Bolm, M.
Ewald, M. Felder, G. Schlingloff, Chem. Ber. 1992, 125, 1169;
g) G. Cheulucci, F. Soccolini, Tetrahedron: Asymmetry 1992, 3,
1235.
[20] E. E. Nifantiev, M. K. Grachev, S. Y. Burmistrov, Chem. Rev.
2000, 100, 3755.
[21] We were unable to prepare the pyridyl phosphinite (R1 = trityl,
R2 = o-Tol) even in the presence of a large excess of reagents.
[22] a) O. Loiseleur, P. Meier, A. Pfaltz, Angew. Chem. 1996, 108,
218; Angew. Chem. Int. Ed. Engl. 1996, 35, 20; b) O. Loiseleur,
M. Hayashi, N. Schmees, A. Pfaltz, Synthesis 1997, 1339; c) O.
Loiseleur, M. Hayashi, M. Keenan, N. Schmees, A. Pfaltz, J.
Organomet. Chem. 1999, 576, 16.
74
ꢀ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
Angew. Chem. Int. Ed. 2004, 43, 70 –74