Chiral bis(N-tosylamino)phosphine- and TADDOL-phosphite-oxazolines as ligands in asymmetric catalysis

Article abstract of DOI:10.1055/s-1999-2939

A series of P,N-ligands containing a chiral oxazoline ring and a chiral bis(N-tosylamino)phosphine group derived from a 1,2-diamine or a chiral cyclic phosphite group derived from TADDOL has been prepared. These compounds proved to be efficient ligands for enantiocontrol of palladium-catalyzed allylic alkylations and iridium-catalyzed hydrogenations of olefins.

Full text of DOI:10.1055/s-1999-2939

1
814
LETTER
Chiral Bis(N-tosylamino)phosphine- and TADDOL-Phosphite-Oxazolines as
Ligands in Asymmetric Catalysis
Robert Hilgraf, Andreas Pfaltz*
Department of Chemistry, University of Basel, St. Johanns-Ring 19, CH-4056 Basel, Switzerland
Fax +41-61 267-1103; E-mail: pfaltz@ubaclu.unibas.ch
Received 31 August 1999
6
by Heldmann and Seebach. They found that rhodium
Abstract: A series of P,N-ligands containing a chiral oxazoline ring
complexes of these ligands are efficient catalysts for the
and a chiral bis(N-tosylamino)phosphine group derived from a 1,2-
enantioselective hydrosilylation of ketones.
diamine or a chiral cyclic phosphite group derived from TADDOL
has been prepared. These compounds proved to be efficient ligands
for enantiocontrol of palladium-catalyzed allylic alkylations and iri-
dium-catalyzed hydrogenations of olefins.
The synthesis of ligands 3-5 was commenced from the
readily available enantiomerically pure diamines 1,2-
diphenylethylene-diamine and 1,2-diaminocyclohexane.
Reaction with tosyl chloride in the presence of diisopro-
Key words: asymmetric catalysis, chiral P,N-ligands, allylic alkyl-
ation, Ir-catalyzed hydrogenation
pylethylamine in CH Cl gave the corresponding bistosy-
2
2
7
lated amines in good yields. Treatment with PCl and
3
triethylamine afforded the bis(N-tosylamino)phosphoro-
chloridites as pale yellow solids, which were character-
ized by mass spectroscopy. Because of their moisture- and
air sensitivity these intermediates were converted directly
to the desired P,N-ligands 3-5 by reaction with the oxazo-
line alcohol, DMAP and triethylamine. Since all attempts
to purify the products by column chromatography failed,
the crude products were recrystallized from chloroform/
hexane. The trace amounts of remaining DMAP were then
removed by carefully heating (90°C) under high vacuum.
This yielded analytically pure P,N-ligands 3-5 in 60-80%
overall yield from the bistosylated amines.
In the course of our studies of enantioselective palladium-
1
-3
catalyzed allylic alkylation, we have developed chiral
3
oxazoline-phosphite ligands of type 2 as a new variant of
2
,4
the phosphinooxazoline (PHOX) ligands 1. Although,
originally, these ligands were designed specifically for
enantio- and regiocontrol in allylic alkylation with 1- and
3
-monosubstituted allyl systems, they have also given
promising results in the enantioselective 1,4-addition of
5
organozinc reagents to cyclic enones.
By this route, a wide variety of different derivatives with
the structural motif of ligands 3-5 should be readily acces-
sible. The modular construction of these ligands allows
independent structural variation of the oxazoline ring, the
backbone, the sulfonamide groups and the substituents on
the diazaphospholidine ring. Although the two stereogen-
ic centers derived from the 1,2-diamine are too far away
from the P-atom to directly influence a reaction taking
place at a P-bound metal, they are expected to induce a
chiral non-planar arrangement of the TsN-P-NTs unit, re-
sulting in a chiral environment near the coordination
sphere of the ligand. Therefore, both the oxazoline and the
diazaphospholidine ring are expected to have a distinct ef-
fect on the enantioselectivity of a metal-catalyzed process.
Here, we describe the synthesis and application of the re-
lated ligands 3-5, containing a bis(N-tosylamino)phos-
phine group derived from a chiral 1,2-diamine, and ligand
6
, possessing a TADDOL-derived phosphite unit. Analo-
gous TADDOL-based ligands containing either an (R)- or
S)-4-isopropyloxazoline ring have been reported recently
(
Synlett 1999, No. 11, 1814–1816 ISSN 0936-5214 © Thieme Stuttgart · New York

LETTER
Chiral Bis(N-tosylamino)phosphine- and TADDOL-Phosphite-Oxazolines
1815
regioselectivity in most cases. In particular, ligand 3a
proved to be very effective for regio- and enantiocontrol
in reactions of 3-aryl-2-propenyl acetates. The TADDOL-
derived ligand 6 gave inferior results with one exception
(
(
Entry 5). In the case of (E)-3-phenyl-2-propenyl acetate
Entry 1), ligand 3a gave distinctly higher enantio- and re-
gioselectivities than the phosphite-oxazoline 2 (R = tBu)
94 vs 86% ee; 84:16 vs 69:31). As with ligands of type 2,
(
the tert-butyloxazoline derivative 3a was superior to the
phenyl-substituted analogue 4. Ligand 5, containing cy-
clohexanediamine as the chiral backbone afforded even
better enantioselectivities than ligand 3a in two cases (En-
tries 1 and 2) but with lower regioselectivity.
The TADDOL-derived ligand 6 was synthesized in a sim-
ilar way to the bis(N-tosylamino)phosphine ligands 3-5.
8
(
R,R)- or (S,S)-TADDOL was treated with PCl in the
3
presence of triethylamine to give the phosphorochloridite
which was converted directly to ligand 6 by reaction with
the oxazoline alcohol, DMAP and triethylamine. Purifica- Using (1’-naphthyl)prop-2-enylacetate as substrate (Entry
tion by recrystallisation from chloroform/hexane afforded 2), ligand 3a again gave significantly higher enantio- and
the P,N-ligand 6 in 71% overall yield.
regioselectivity than BINOL-derived ligands 2 (R = tBu:
5% ee, 90:10) or other Pd catalysts reported in the liter-
ature. As expected, the ee's and regioselectivities with 2-
butenyl acetate (Entry 3) were lower, however ligand 3a
still showed better enantio- and regioselectivity than the
phosphite ligands 2 (43% ee, 30:70).
9
With these ligands in hand, we evaluated their scope in the
Pd-catalyzed allylic alkylation of various substrates (Ta-
ble 1). Of special interest were the aryl-allyl derivatives
3
(
Entries 1-2) because of the promising results obtained
previously with ligand 2 (R = tBu).
With cyclohexenyl acetate (Entry 6), ligand 3a proved to
be superior to the analogous phosphite ligands 2 or phos-
phinooxazolines 1. However, the ee was still moderate
compared to the best values reported so far.4 In reac-
tions with 1,3-disubstituted acyclic substrates (Entries 4
and 5), all new ligands gave unsatisfactory ee's. Ligand 3b
as well as the corresponding diastereomers of ligands 5
and 6 gave lower ee's and regioselectivities in the reaction
of phenylallyl acetate (Entry 1; 3b: 93% ee, 53:47).
a,10
In view of the promising results obtained with phosphi-
nooxazolines 1 in the Ir-catalyzed hydrogenation of ole-
9
fins, we decided to test our new ligands in this type of
reaction (Table 2). The required [Ir(COD)(L)]TFPB com-
plexes (TFPB = tetrakis[3,5-bis(trifluoromethyl)phe-
nyl]borate) were readily obtained as air-stable, red-orange
crystals by the standard procedure previously reported.9
All reactions were carried out with 4 mol% of catalyst at
1
00 bar hydrogen pressure. Lower catalyst loading or hy-
drogen pressure resulted in reduced conversions and
enantioselectivities. Interestingly, ligand 3b was more ef-
ficient than the diastereomer 3a in the hydrogenation of
(
E)-1,2-diphenyl-1-propene (Entry 3; 87 vs 80% ee, 49 vs
1
6% conv.), in contrast to allylic substitution where ligand
3
a gave better results. However, replacement of ligands 5
and 6 by the corresponding diastereomers resulted in low-
er ee's and conversions.
In comparison to iridium catalysts derived from phosphi-
nooxazolines of type 1, the novel bis(N-tosylamino)phos-
phine-derived ligands 3b and 5 gave comparable
enantioselectivities but mostly lower conversions. How-
ever, in the hydrogenation of ethyl ß-methylcinnamate
and (E)-2-(4-methoxyphenyl)but-2-ene (Entries 4 and 5),
higher ee's than those previously reported were achieved.
In comparison to oxazoline-phosphite ligands of type 2, The TADDOL-derived ligand 6, in general, gave higher
the new bis(N-tosylamino)phosphine-derived ligands 3-5 conversions. The best results were achieved in the hydro-
gave comparable or better results in terms of enantio- and genation of (Z)-2-(4-methoxyphenyl)but-2-ene (Entry 6),
Synlett 1999, No. 11, 1814–1816 ISSN 0936-5214 © Thieme Stuttgart · New York

1
816
R. Hilgraf, A. Pfaltz
LETTER
where the ee was improved from 42% (with phosphinoox- solution of NEt
(784 mg, 7.74 mmol) and DMAP (96 mg, 0.78
mmol) in 8 ml of toluene was added dropwise over 20 minutes, fol-
3
azolines) to 90% ee.
lowed by a solution of (-)-(S)-2-[1’-hydroxy-1’-methylethyl]-4-
1
2
tert-butyloxazoline (141 mg, 0.76 mmol) in 4 ml of toluene. The
mixture was slowly warmed to r.t. overnight. After filtration under
argon and evaporation of the solvent, the crude product was recrys-
tallized from anhydrous chloroform/hexane. Traces of DMAP were
removed at 90°C/0.01 mbar to yield 400 mg (71 % yield) of analyt-
ically pure 3a as a white solid.
1
H NMR: d (CDCl ) 0.90 (s, 9H), 1.88 (s, 6H), 2.24 (s, 3H), 2.34 (s,
3
3
H), 3.90 (m, 1H), 4.21 (m, 2H), 4.61 (d, J=8.2 Hz, 1H), 4.84 (d,
3
1
J=8.3 Hz, 1H), 6.84-7.55 (m, 18H). P NMR: d (CDCl ) 121.25.
Anal. calc. for C H N O PS : C 62.19, H 6.04, N 5.73; found C
3
3
8
44
3
6
2
6
2.08, H 6.09, N 5.66.
Acknowledgement
R.H. thanks the Fonds der Chemischen Industrie for a Kekulé Sti-
pendium. This work was supported by the Swiss National Science
Foundation.
References and Notes
(
(
1) A. Pfaltz Acc. Chem. Res. 1993, 26, 339.
2) a) A. Pfaltz Acta Chem. Scand. B 1996, 50, 189. b) P. von
Matt, A. Pfaltz Angew. Chem. Int. Ed. 1993, 32, 566; Angew.
Chem. 1993, 105, 614.
(
(
3) a) R. Prétôt, G. C. Lloyd-Jones, A. Pfaltz Pure Appl. Chem.
1
1
998, 70, 1035. b) R. Prétôt, A. Pfaltz Angew. Chem. Int. Ed.
998, 37, 323; Angew. Chem. 1998, 119, 337. See also: c) M.
Kawatsura, Y. Uozumi, T. Hayashi Chem. Commun. 1998,
17.
2
4) a) S. Kudis, P. Sennhenn, H. Steinhagen, G. Helmchen Pure
Appl. Chem. 1997, 69, 513. b) J. M. J. Williams Synlett 1996,
705.
Our results show that the new P,N-ligands described
above can induce high enantioselectivities in Pd- and Ir-
catalyzed reactions, in some cases surpassing the selectiv-
ities obtained with phosphinooxazolines or other ligands.
In view of the many applications that have been reported
for phosphinooxazolines, it will be worthwhile exploring
the potential of bis(N-tosylamino)phosphine- and phos-
phite-oxazoline ligands for other metal-catalyzed process-
es.
(5) A. K. H. Knöbel, I. H. Escher, A. Pfaltz Synlett 1997, 1429.
(
(
(
6) D. K. Heldmann, D. Seebach Helv. Chim. Acta 1999, 82,
1096.
7) H. Takahashi, T. Kawakita, M. Ohno, M. Yoshioka, S.
Kobayashi Tetrahedron 1992, 48, 5691.
8) R. Dahinden, A. K. Beck, D. Seebach Encyclopedia of
Reagents for Organic Synthesis John Wiley & Sons:
Chichester, 1995, Vol. 3, 2167.
(9) A. Lightfoot, P. Schnider, A. Pfaltz Angew. Chem. Int. Ed.
998, 37, 2897; Angew. Chem. 1998, 110, 3047.
1
(
(
10) B. M. Trost Acc. Chem. Res. 1996, 29, 355.
11) P. von Matt, G. C. Lloyd-Jones, A. B. E. Minidis, A. Pfaltz, L.
Macko, M. Neuburger, M. Zehnder, H. Rüegger, P. S.
Pregosin Helv. Chim. Acta 1995, 78, 265.
Experimental procedure
Ligand 3a: Freshly destilled PCl (152 mg, 1.1 mmol) was added to
3
a solution of NEt (328 mg, 3.24 mmol) in 4.8 ml of toluene at –78
3
°
C under argon, followed by dropwise addition, via a dropping fun-
(12) J.V. Allen, J.M.J. Williams, Tetrahedron: Asymmetry 1994, 5,
nel, of a suspension of (1R,2R)-1,2-N,N’-bis(p-toluenesulfonylami-
7
277.
no)-1,2-diphenylethane (400 mg, 0.77 mmol) in 10.4 ml of toluene.
The dropping funnel was washed with 4 ml of toluene. The mixture
was allowed to slowly warm up to r.t. overnight. After filtration un-
der argon and evaporation of the solvent, the residue was dissolved
in 19 ml of toluene, cooled to –78 °C and stirred for 5 minutes. A
Article Identifier:
437-2096,E;1999,0,11,1814,1816,ftx,en;G22099ST.pdf
1
Synlett 1999, No. 11, 1814–1816 ISSN 0936-5214 © Thieme Stuttgart · New York