Asymmetric hydrogenation of unsaturated ureas with the BIPI ligands

Article abstract of DOI:10.1021/ol7028542

(Chemical Equation Presented) Asymmetric hydrogenation of unsaturated urea esters with the BIPI Ligands has been examined. Optimization of the P-N ligand structure has led to the development of chiral rhodium catalysts capable of producing the targets wit

Full text of DOI:10.1021/ol7028542

ORGANIC
LETTERS
2008
Vol. 10, No. 2
341-344
Asymmetric Hydrogenation of
Unsaturated Ureas with the BIPI Ligands
Carl A. Busacca,* Jon C. Lorenz, Nelu Grinberg, Nizar Haddad, Heewon Lee,
Zhibin Li, Mary Liang, Diana Reeves, Anjan Saha, Rich Varsolona, and
Chris H. Senanayake
Department of Chemical DeVelopment, Boehringer-Ingelheim Pharmaceuticals, Inc.,
900 Ridgebury Road, Ridgefield, Connecticut 06877
cbusacca@rdg.boehringer-ingelheim.com
Received December 6, 2007
ABSTRACT
Asymmetric hydrogenation of unsaturated urea esters with the BIPI Ligands has been examined. Optimization of the P−N ligand structure has
led to the development of chiral rhodium catalysts capable of producing the targets with >99% ee. The critical phosphine borane SNAr reaction
needed for ligand synthesis has been optimized to give the adducts in high yield at ambient temperature with no racemization. An extremely
concise, economical, and scalable sequence to these important new ligands for catalysis of asymmetric hydrogenation has been developed.
Several molecules of biological interest to us possess a chiral
urea ester moiety. We envisaged that this important subunit
could be accessed by asymmetric hydrogenation of the
requisite dehydro-urea esters. Enantioselective hydrogena-
tions of unsaturated aminoesters with acyl- and carbamoyl-
substitution on nitrogen are well precedented,1 although the
analogous ureas do not appear to have been studied. We
report herein an asymmetric hydrogenation of these substrates
using newly engineered chiral phosphinoimidazolines, the
BIPI ligands,2 with extremely high levels of stereocontrol
(>99% ee).
During our process research analysis, we needed to
develop an economical, practical, and scalable process for
production of urea 2 in optically pure form. Initial ligand
screening and optimization was done using the functionalized
urea substrate 1 as shown in Scheme 1. We used cationic
rhodium in the form of Rh(NBD)2BF4, due to the higher
hydrogenation rates observed for NBD relative to COD in
rhodium-catalyzed hydrogenations.3 The screening reactions
were performed in methanol at 30 °C, under 100 psi H2,
and 0.6 mol % (0.006 equiv) of catalyst. Most of the BIPI
ligands containing one or two aryl groups on phosphorus2,4
failed to give any turnover in this hydrogenation reaction,
(1) For reviews, see: (a) Brown, J. M. In ComprehensiVe Asymmetric
Catalysis; Jacobsen, E., Pfaltz, A., Yamamoto, H., Eds.; Springer: Berlin,
1999; Vol. 1, pp 121-182. (b) Drexler, H.-J.; You, J.; Zhang, S.; Fischer,
C.; Baumann, W.; Spannenberg, A.; Heller, D. Org. Process Res. DeV. 2003,
7 (3), 355. (c) Gridnev, I. D.; Imamoto, T. Acc. Chem. Res. 2004, 37 (9),
633. (d) Noyori, R.; Kitamura, M.: Ohkuma, T. Proc. Natl. Acad. Sci. U.S.A.
2004, 101 (15), 5356. (e) McGarrity, J. F.; Brieden, W.; Fuchs, R.; Mettler,
H.-P.; Schmidt, B.; Werbitzky, O. In Asymmetric Catalysis on Industrial
Scale; Blaser, H. U., Schmidt, E., Eds.; Wiley:Weinheim, 2004; pp 283-
308.
(2) (a) Busacca, C. A. U.S Patent 6,316,620, 2001. (b) Busacca, C. A.;
et al. J. Org. Chem. 2004, 69 (16), 5187. (c) Busacca, C. A.; Grossbach,
D.; So, R. C.; O’Brien, E. M.; Spinelli, E. M. Org. Lett. 2003, 5 (4), 595.
(3) Drexler, H.-J.; Baumann, W.; Spannenberg, A.; Fischer, C.; Heller,
D. J. Organomet. Chem. 2001, 621 (1-2), 89.
(4) (a) Menges, F.; Neuburger, M.; Pfaltz, A. Org. Lett. 2002, 4 (26),
4713. (b) Guiu, E.; Claver, C.; Benet-Buchholz, J.; Castillon, S. Tetrahedron:
Asymmetry 2004, 15 (21), 3365.
10.1021/ol7028542 CCC: $40.75
© 2008 American Chemical Society
Published on Web 12/21/2007

We ultimately discovered that stoichiometric formation of
two equivalents of the phosphine borane anion prior to
addition of the arylfluoride was the key to obtaining both
100% conversion and complete suppression of racemization.
As shown in Scheme 3, two equivalents of dicyclohexy-
Scheme 1. Asymmetric Hydrogenation
Scheme 3. Typical Ligand Synthesis
leading to recovered starting material. Dialkylphosphine
substitution was thus found to be a general prerequisite for
reactivity.
After evaluation of many available chiral diamines (alkyl
and aryl) we chose to hold the chiral backbone of the ligand
structure constant as the 1,2-diphenylethylenediamine (DPE-
DA) derivative. Importantly, both enantiomers of DPEDA
are commercially available on a large scale. We decided to
form the critical carbon-phosphorus bond by an SNAr
reaction. This type of nucleophilic substitution of arylfluo-
rides is well-known for metal phosphides,5 although SNAr
of phosphide borane anions has been far less studied.
Recently, Imamoto6 has described phosphine borane SNAr
reactions under very mild conditions. The latter method
employed electrophile activation by formation of the sto-
ichiometric arene chromium tricarbonyl complex. However,
this method was not desirable in the process chemistry
environment for large-scale operations. As a result, a
significant amount of experimentation was required to
develop a robust fluoroimidazoline SNAr reaction with
dialkylphosphide boranes. Specifically, two serious problems
had to be overcome: incomplete conversion and unexpected
and pronounced racemization of the SNAr product. A typical
SNAr result before optimization is shown in Scheme 2.
lphosphine borane anion were generated in DMAc at ambient
temperature using 60% NaH. When the aryl fluoride was
subsequently charged, a deep orange- or red-colored reaction
mixture indicative of the SNAr reaction was immediately
formed. The “obvious” explanation for this 2:1 stoichiometry
requirement appeared to be that one equivalent of the
phosphide served to deprotonate the imidazoline NH. This
concept was readily invalidated, however. When one equiva-
lent of phosphide anion was added to a separate flask
containing one equivalent of deprotonated (NaH) fluoroimi-
dazoline, less than 10% conversion occurred. The true
explanation for this requirement is more subtle and is an
active area of investigation. These SNAr reactions could be
performed at 23 °C, requiring from 4 to 16 hours to be
completed (Scheme 3). Reactions in less polar solvents, such
as THF, were far more sluggish. The isolated phosphine
borane imidazolines8 (5) were then generally deprotected
with DABCO in toluene at 50 °C (6), and ligand synthesis
was completed by addition of an electrophile to functionalize
the nitrogen atom.
Scheme 2. SNAr Reaction Prior to Optimization
Utilizing KOH as the base in DMSO, the reaction stalled at
75% conversion. More importantly, the product was found
to have undergone extensive racemization under the condi-
tions employed. Solvent, base, cation, and temperature were
all found to be critical parameters controlling racemization.7
We felt it was critical to develop a ligand synthesis that
was very direct, highly scalable, and inexpensive to allow
for large-scale production. The synthesis outlined here is
much shorter than typical routes to furnish highly enanti-
oselective P-P ligands and relies upon a relatively inex-
pensive chiral diamine rather than a costly chiral diol. The
route also avoids the need for optical resolution of intermedi-
(5) (a) Coote, S. J.; Dawson, G. J.; Frost, C. G.; Williams, J. M. J. Synlett
1993, 509. (b) Dawson, G. J.; Frost, C. G.; Williams, J. M. J. Tetrahedron
Lett. 1993, 34 (19), 3149. (c) Tang, X.; Zhang, D.; Jie, S.; Sun, W.-H.;
Chen, J. J. Organomet. Chem. 2005, 690 (17), 3918.
(6) Katagiri, K.; Danjo, H.; Yamaguchi, K.; Imamoto, T. Tetrahedron
2005, 61 (19), 4701.
(7) The mechanism of racemization is under close scrutiny and will be
reported elsewhere in due course.
(8) An X-ray crystal structure of 5 is included in the Supporting
Information.
342
Org. Lett., Vol. 10, No. 2, 2008

ates common to many P-P ligand syntheses. We have
recently demonstrated scalability by rapidly preparing >1
kg of BIPI 69 in a very facile campaign.
mization involved changing the “benzo core” from phenyl
to naphthyl (entries 5-6). This modification led to a dramatic
breakthrough in stereoselectivity, generating urea 2 with 98%
ee and >99% ee, respectively, for BIPI 166 and BIPI 153.
The reasons for this improvement are not immediately clear,
yet we speculate that the peri hydrogen of the naphthyl ring
restricts the number of conformations available to the two
cyclohexyl substituents. The remaining conformations may
be more inherently enantioselective, leading to the extreme
selectivity observed.
We next decided to screen our best ligand, BIPI 153,
against a variety of structurally diverse dehydrourea esters
to gain an understanding of the scope of this new asymmetric
hydrogenation. The results for six different substrates are
collected in Figure 2.
Figure 1. Ligands used.
The ligands used in this study are shown in Figure 1. The
BIPI ligands were designed to allow for ready electronic
tuning by varying the nitrogen substituent. We had expected
that alkyl substitution on nitrogen would be beneficial, since
most of the successful P-P ligands for asymmetric hydro-
genation of olefins, such as DuPhos, Tangphos, and others,9
are quite electron rich. To our surprise, alkyl substitution
led to catalysts that gave poor turnover and low selectivity
(0-10% ee). We then examined different types of acyl
substitution on nitrogen. As shown in Table 1, electron rich
Table 1. Asymmetric Hydrogenation of Substrate 1
entry
ligand no.
er
% ee
Figure 2. Asymmetric hydrogenation of additional substrates.
1
2
3
4
5
6
BIPI 41
BIPI 39
BIPI 164
BIPI 69
BIPI 166
BIPI 153
85.5:14.5
92.5:7.5
97.2:2.8
97.7:2.3
99.1:0.9
>200:1
71
85
94
95
98
The phenylalanine analogues 7b-9b were each generated
in >99% ee, and the isoleucine target 10b, in 99% ee.
Variation of the urea functionality (11b) was also completely
tolerated, as was introduction of a terminal olefin (12b), as
less than 5% hydrogenation of this moiety was observed,
and the product still formed in >99% ee.
We wanted to determine the absolute configuration of
these urea products. This was accomplished by converting
both (S)-phenylalanine methyl ester and (S)-isoleucine methyl
ester to their morpholine urea derivative using commercial
morpholine carbonyl chloride.
>99
p-isopropoxybenzoyl was more enantioselective (85% ee)
than the p-CF3 analogue (71% ee, entries 1-2). We therefore
reasoned that an alkyl amide might prove to be better still.
This was indeed the case, as N-acetyl and N-cyclohexan-
ecarbonyl gave the product with 94% ee and 95% ee,
respectively (entries 3-4). The final stage of ligand opti-
In both cases, these materials coeluted on chiral HPLC
with the major enantiomer formed through the use of the
(S,S)-BIPI ligands (7b, 10b). Although we did not rigorously
assign the absolute stereochemistry of the other products, it
seems quite likely that the (S,S) ligand series will furnish
ureas with the (S) absolute configuration in most cases.
In conclusion, we have applied the BIPI ligands to the
asymmetric hydrogenation of unsaturated urea esters and
have uncovered a catalyst, BIPI 153, that proceeds with near-
perfect enantioselection. The synthetic process to access this
ligand class is short, simple, and scalable, and leads to a
very practical asymmetric hydrogenation sequence. The high
(9) For examples, see: (a) Cobley, C. J.; Johnson, N. B.; Lennon, I. C.;
McCague, R.; Ramsden, J. A.; Zanotti-Gerosa, A. In Asymmetric Catalysis
on Industrial Scale; Blaser, H. U., Schmidt, E., Eds.; Wiley:Weinheim, 2004;
269-282. (b) Burk, M. J. Acc. Chem. Res. 2000, 33 (6), 363. (c) Tang,
W.; Zhang, X. Org. Lett. 2002, 4 (23), 4159. (d) Liu, D.; Zhang, X. Eur.
J. Org. Chem. 2005, 4, 646. (e) Almena, J.; Monsees, A.; Kadyrov, R.;
Riermeier, T. H.; Gotov, B.; Holz, J.; Boerner, A. AdV. Synth. Cat. 2004,
346 (11), 1263. (f) Hoge, G.; Wu, H.-P.; Kissel, W. S.; Pflum, D. A.; Greene,
D.; Bao, J. J. Am. Chem. Soc. 2004, 126 (19), 5966. (g) Lefort, L.; Boogers,
J. A. F.; de Vries, A. H. M.; de Vries, J. G. Top. Catal. 2006, 40 (1-4),
185. (h) Blaser, H.-U.; Brieden, W.; Pugin, B.; Spindler, F.; Studer, M.;
Togni, A. Top. Catal. 2002, 19 (1), 3. For P-N ligands, see: (i)
Bunlaksananusorn, T.; Polborn, K.; Knochel, P. Angew. Chem., Int. Ed.
2003, 42, 3941. (j) Liu, D.; Tang, W.; Zhang, X. Org. Lett. 2004, 6 (4),
513.
Org. Lett., Vol. 10, No. 2, 2008
343

levels of stereocontrol observed here have previously been
obtained almost exclusively with P-P ligands.10 Appropri-
ately designed P-N ligands such as these can now clearly
be used for this important class of asymmetric transforma-
tions as well. Application of the BIPI ligands to a wide range
of unsaturated substrates is under active investigation and
will be reported in due course.
Supporting Information Available: Full experimental
procedures and compound characterization; X-ray crystal-
lographic data for 5; complete ref 2b. This material is
available free of charge via the Internet at http://pubs.acs.org.
(10) For P-N ligands in asymmetric hydrogenations of functionalized
olefins, see: (a) Bunlaksananusorn, T.; Polborn, K.; Knochel, P. Angew.
Chem., Int. Ed. 2003, 42, 3941. (b) Yamada, I.; Ohkouchi, M.; Yamaguchi,
M.; Yamagishi, T. J. Chem. Soc., Perkin Trans. 1 1997, 1869. (c) Brauer,
D. J.; Kottsieper, K. W.; Rossenbach, S.; Stelzer, O. Eur. J. Inorg. Chem.
2003, 9, 1748.
OL7028542
344
Org. Lett., Vol. 10, No. 2, 2008