Parallel ligand screening on olefin mixtures in asymmetric hydroformylation reactions

Article abstract of DOI:10.1021/ol0487938

(Chemical Equation Presented) Herein we describe a new protocol for catalyst evaluation in asymmetric hydroformylation reactions where multisubstrate screening is performed in an array of parallel reactors. This method was successfully demonstrated using a mixture of styrene, allyl cyanide, and vinyl acetate. Using this screening methodology, a set of phosphite ligands was evaluated and led to the discovery of a bisphosphite ligand that gave 88% ee and unprecedented >100:1 branched:linear regioselectivity in asymmetric hydroformylation of vinyl acetate.

Full text of DOI:10.1021/ol0487938

ORGANIC
LETTERS
2
004
Vol. 6, No. 19
277-3280
Parallel Ligand Screening on Olefin
Mixtures in Asymmetric
Hydroformylation Reactions
†
3
,‡
Christopher J. Cobley, Jerzy Klosin,* Cheng Qin, and Gregory T. Whiteker*
Chemical Sciences, The Dow Chemical Company, 1776 Building,
Midland, Michigan 48674
jklosin@dow.com
Received June 24, 2004
ABSTRACT
Herein we describe a new protocol for catalyst evaluation in asymmetric hydroformylation reactions where multisubstrate screening is performed
in an array of parallel reactors. This method was successfully demonstrated using a mixture of styrene, allyl cyanide, and vinyl acetate. Using
this screening methodology, a set of phosphite ligands was evaluated and led to the discovery of a bisphosphite ligand that gave 88% ee and
unprecedented >100:1 branched:linear regioselectivity in asymmetric hydroformylation of vinyl acetate.
Asymmetric hydroformylation is a catalytic, atom-economic,
one-carbon homologation that produces chiral aldehydes.
chelating ligands can be prepared using achiral bridging diols.
This ligand design allows for the variation of ligand bite
angle, a structural feature that has been previously implicated
1
The historical challenge of asymmetric hydroformylation has
been to maximize both enantioselectivity and regioselectivity
through ligand design. Unfortunately, no single ligand has
been identified that leads to high selectivity with a wide range
of olefin substrates. As with many asymmetric catalytic
reactions, ligands required for a specific hydroformylation
substrate need to be optimized empirically. We recently
described a simple ligand design to access a wide range of
structurally diverse bisphosphite ligands for use in rhodium-
1
,3
in hydroformylation regiocontrol. To rapidly evaluate this
class of easily synthesized bisphosphite ligands, we utilized
parallel pressure reactors. To increase the ligand screening
throughput even further, we decided to explore the possibility
of a one-pot screening of multiple olefinic substrates. This
methodology was recently introduced by Kagan for screening
of a single catalyst in the asymmetric reduction of ketones.4
Herein, we demonstrate simultaneous screening of pooled
substrates in the asymmetric hydroformylation of olefins
using parallel reactors that allows screening throughput to
be dramatically increased.
2
catalyzed asymmetric hydroformylation. By use of optically
active biphenols as chiral auxiliaries, a wide variety of
†
Dowpharma, Chirotech Technology Limited, a subsidiary of The Dow
Chemical Company, Cambridge, U.K.
The phosphite ligands investigated in this study are
depicted in Figure 2. The synthesis and use of these
‡
The Dow Chemical Company, South Charleston, WV 25303.
(
1) Claver, C.; van Leeuwen, P. W. N. M. In Rhodium Catalyzed
Hydroformylation; Claver, C., van Leeuwen, P. W. N. M., Eds.; Kluwer
Academic Publishers: Dordrecht, 2000.
(3) Casey, C. P.; Whiteker, G. T.; Melville, M. G.; Petrovich, L. M.;
Gavney, J. A.; Powell, D. R. J. Am. Chem. Soc. 1992, 114, 5535.
(4) (a) Gao, X.; Kagan, H. B. Chirality 1998, 10, 120. (b) Wolf, C.;
Hawes, P. A. J. Org. Chem. 2002, 67, 2727.
(
2) Cobley, C. J.; Gardner, K.; Klosin, J.; Praquin, C.; Hill, C.; Whiteker,
G. T.; Zanotti-Gerosa, A.; Petersen, J. L.; Abboud, K. A. J. Org. Chem.
004, 69, 4031.
2
1
0.1021/ol0487938 CCC: $27.50 © 2004 American Chemical Society
Published on Web 08/20/2004

Figure 1. Gas chromatograms of aldehyde product mixture produced with Rh-Kelliphite catalyst. Enantioselectivity and regioselectivity
measurements of styrene and vinyl acetate hydroformylation products were obtained using a Supelco Beta Dex 225 column (chromatogram
a). Allyl cyanide hydroformylation products were analyzed using an Astec Chiraldex A-TA column (chromatogram b). Peaks marked with
an asterisk are due to dodecane (internal standard).
phosphites in asymmetric hydroformylation of allyl cyanide
has been recently described. All screening experiments were
parallel chiral GC analysis using an R-cyclodextrin column
was developed. Dodecane was used as internal standard in
all experiments. These methods also allowed quantitative
analysis of unreacted olefin as well as the products from
olefin hydrogenation.
2
performed in an Argonaut Endeavor reactor system consisting
of eight parallel stirred autoclaves. Ligands were tested
against an olefin mixture composed of styrene, vinyl acetate,
and allyl cyanide in 1:1:1 molar ratio. These olefins were
chosen because of their significance in the potential synthesis
of fine chemicals via asymmetric hydroformylation. For
example, styrene was studied because of its relevance in the
synthesis of antiinflammatory 2-aryl-propionic acid drugs,
such as (S)-Naproxen. Additionally, these three substrates
represent different types of terminal olefins with their unique
reactivity and selectivity in hydroformylation reactions.
Screening against all three substrates in a single run should
thus allow for identification of ligands applicable to a broad
range of olefins.
To ensure that the presence of additional olefins did not
influence catalyst performance, a series of control experi-
5
ments was performed at 35 and 70 °C using Chiraphite,
which was previously developed for hydroformylation of
6
vinylarenes (Table 1). Reactions were studied with separate
Table 1. Asymmetric Hydroformylation of Styrene (St), Allyl
Cyanide (AC), and Vinyl Acetate (VA) with Chiraphite at 70 °C
in Toluene; Control Experiment5
separate olefins
Simultaneous screening of pooled substrates can be
successfully applied only if appropriate analytical techniques
are developed and the presence of other components in the
reaction mixture does not influence reactivity and selectivity
of any individual reaction. Chiral GC methods were devel-
oped to rapidly analyze the aldehyde mixtures resulting from
hydroformylation of the mixture of three olefins (Figure 1).
Use of a â-cyclodextrin GC column allowed complete
separation of both enantiomers of branched aldehyde and
the achiral linear aldehyde derived from both styene and vinyl
acetate, thus allowing for measurements of enantioselectivity
and regioselectivity. As a result of incomplete separation of
the linear aldehyde regioisomer and chiral branched alde-
hydes derived from allyl cyanide in this first analysis, a
St
AC
VA
combined olefins
%
%
conv
ee
St
AC
VA
68(3)a
97.9(9)
100
96(1)
100
95.6(9)
52.2(1)
74(3)
St
AC
VA
54.8(3)
12.9(1)
57.7(5)
14.6(1)
52.3(1)
14.2(2)
6.2
b/l
St
AC
VA
14.3(3)
6.18(5)
95(3)
a
Numbers in parentheses are confidence intervals at a 95% confidence
level obtained from four independent runs.
3278
Org. Lett., Vol. 6, No. 19, 2004

Figure 2. Ligands used in this study.
olefins as well as their mixture in toluene and acetone as
solvents. All reactions were found to proceed at a faster
is perhaps the most widely used ligand in asymmetric
hydroformylation, was included in the screens (cell 8) for
7
10
rate in acetone. Acetone was found to be the optimal solvent
for reactions at 35 °C, since at 70 °C enantioselectivities of
products derived from styrene and allyl cyanide were reduced
by about 20% when using the mixture of olefins as compared
to experiments with separate olefins. In toluene, regardless
of temperature, enantioselectivities were found to be inde-
comparison. Data for this parallel, multisubstrate screening
are given in Table 2. Through use of this screening
methodology, the catalytic performance of eight different
ligands with three different substrates was obtained under
identical reaction conditions (24 reactions all together).
Inspection of the data in Table 2 revealed a range of
regioselectivities and enantioselectivities for the three olefins
studied. In particular, ligand 1, which we refer to as
Kelliphite, exhibited very high enantioselectivity (88% ee
at 35 °C) for hydroformylation of vinyl acetate. There are
very few ligands capable of inducing such high enantiose-
lectivity in rhodium-catalyzed hydroformylation of vinyl
8
pendent of whether hydroformylation reactions were per-
formed using a single olefin or the mixture of three olefins;
therefore toluene was selected as solvent for runs at 70 °C.
In addition, conversions were typically unchanged; however,
styrene hydroformylation using Chiraphite was slightly
accelerated for the mixed olefins reaction. Analogous control
experiments performed with ligand 1 further validated this
method.5
9
11
12
acetate as Binaphos, Esphos and ABDPP. What distin-
guishes Kelliphite from these other highly enantioselective
ligands, however, is its ability to induce unprecedented high
regioselectivity. This is very significant as the chemical yield
of chiral product in asymmetric hydroformylation depends
both on enantioselectivity and regioselectivity of the reaction.
Following the demonstration of the simultaneous hydro-
formylation of pooled olefins, we next applied this procedure
to parallel asymmetric hydroformylation screening of the
9
phosphite ligands shown in Figure 2. (R,S)-Binaphos, which
9,11,12
None of the ligands previously reported
for asymmetric
hydroformylation of vinyl acetate gave such high regiose-
lectivities. For example, the regioselectivity observed with
Kelliphite is considerably higher than that obtained with
(
5) See Supporting Information for details.
(6) (a) Babin, J. E.; Whiteker, G. T. Union Carbide U.S. Patent 5,360,938,
1
994. (b) Buisman, G. J. H.; Vos, E. J.; Kamer, P. C. J.; van Leeuwen, P.
W. N. M. J. Chem. Soc., Dalton Trans. 1995, 409.
7) The concentration of each olefin was held constant: 1.4 mM [styrene],
.4 mM [vinyl acetate], 1.4 mM [allyl cyanide]. For reactions of individual
olefins, the same olefin concentration was used.
8) Regioselectivities varied slightly presumably as a result of polarity
differences between individual and combined olefins solutions.
9) Nozaki, K.; Sakai, N.; Nanno, T.; Higashijima, T.; Mano, S.; Horiuchi,
(
1
(10) Data for Chiraphite were obtained from a separate set of comparative
experiments.
(11) Breeden, S.; Cole-Hamilton, D. J.; Foster, D. F.; Schwarz, G. J.;
Wills, M. Angew. Chem., Int. Ed. 2000, 39, 4106.
(12) Lu, S.; Li, X.; Wang, A. Catal. Today 2000, 63, 531. ABDPP )
1,6-anhydro-2,6-bis(diphenylphosphino)pyranose.
(
(
T.; Takaya, H. J. Am. Chem. Soc. 1997, 119, 4413.
Org. Lett., Vol. 6, No. 19, 2004
3279

Table 2. Asymmetric Hydroformylation of Styrene (St), Allyl Cyanide (AC), and Vinyl Acetate (VA) with Chiraphite, 1-7, and
Binaphosa
conversionb
regioselectivityc
(b/l) (St, AC, VA)
enantioselectivity
entry
1
ligand
T (°C)
(%) (St, AC, VA)
(% ee, configuration) (St, AC, VA)
Chiraphite
35
53.5, 67.2, 21.9
98.1, 100, 94.8
39.6, 95.1, 19.4
99.6, 100, 98.7
81.9, 99.6, 18.9
99.9, 100, 98.7
31.3, 57.9, 2.5
99.4, 100, 80.9
92.6, 99.6, 15.5
100, 100, 98.6
43.3, 61.1, 2.8
100, 100, 100
59.4, 86.7, 11.1
100, 100, 99.9
15.2, 23.4, 6.2
63.2, 99.9, 83.7
33.0, 49.3, 9.9
99.8, 99.9, 94.2
46.8, 6.7, 109.2
13.3, 6.1, 100.2
68.3, 15.9, 124.7
17.2, 10.8, 55.7
29.6, 12.2, 21.5
12.0, 10.5, 16.4
26.8, 5.4, 160.0
7.6, 5.3, 11.3
35.1, 6.5, 107.0
12.6, 7.1, 180.0
16.1, 5.3, 12.3
2.3, 4.2, 35.8
16.6, 5.0, 13.9
2.4, 4.0, 28.4
16.2, 5.0, 20.7
1.1, 3.4, 166.3
9.7, 2.7, 7.0
76.4 (S), 12.6 (R), 58.0 (S)
55.9 (S), 15.3 (R), 47.2 (S)
17.7 (R), 75.2 (S), 87.7 (S)
0.7 (S), 69.8 (S), 77.2 (S)
9.0 (S), 5.3 (S), 30.0 (S)
7.2 (R), 18.4 (S), 29.5 (S)
37.4 (S), 12.9 (R), 3.5 (S)
6.6 (S), 0.8 (R), 11.7 (S)
18.8 (R), 2.1 (R), 5.9 (S)
20.8 (R), 5.7 (S), 6.8 (S)
37.6 (S), 43.1 (R), 2.8 (R)
7.7 (R), 34.8 (R), 4.6 (R)
27.6 (S), 37.4 (R), 8.3 (R)
10.4 (R), 30.3 (R), 7.5 (R)
3.7 (R), 13.2 (R), 4.6 (R)
12.7 (R), 3.9 (R), 2.2 (R)
85.6 (R), 80.4 (S), 70.0 (S)
84.8 (R), 69.3 (S), 62.1 (S)
7
0
2
3
4
5
6
7
8
9
1
35
7
0
2
35
7
0
3
35
7
0
4
35
7
0
5
35
7
0
6
35
7
0
7
35
7
0
Binaphos
35
7
0
5.3, 2.2, 6.5
a
Pressure ) 150 psi (1:1 H2/CO), ligand:Rh ) 1.2:1 for bidentate and 2.2:1 for monodentate phosphites, solvent ) acetone (35 °C runs) and toluene (70
°
C) (3.66 mL), olefins:Rh ) 300:1 at 35 °C and 500:1 at 70 °C. b Percentage conversion of olefins after 3 h. c b/l ) branched-to-linear ratio. Each data point
is an average obtained from two independent runs.
Binaphos (b/l ) 125 vs 7.0 at 35 °C; Table 2, entries 2 and
Use of the Symyx Parallel Pressure Reactor15 (PPR-48)
allowed 48 different catalysts to be screened simultaneously
with styrene, vinyl acetate, and allyl cyanide (144 discrete
reactions). Using this methodology, we have screened a
library of 78 catalysts based on phosphites, phosphines, and
phosphoroamidites. Kelliphite remained the most selective
ligand uncovered by this screen for both allyl cyanide and
vinyl acetate hydroformylation.
9). Enantioselectivities obtained with Binaphos for styrene
and vinyl acetate (Table 2, entry 9) in this study are
9
somewhat lower than reported values. This is most likely
due to different reaction conditions used.
A larger scale reaction in neat vinyl acetate with the Rh-
13
Kelliphite catalyst system proceeded with satisfactory rates
-
1
(
minimum average TOF of 200 h ) and high enantio- and
In summary, we have demonstrated for the first time that
multisubstrate screening in combination with parallel tech-
niques leads to higher reactor throughput. Using this
methodology we have discovered a ligand that gives high
enantio- and regioselectivity for the asymmetric hydroformy-
lation of vinyl acetate. This method should be applicable to
other homogeneous catalyzed reactions.
regioselectivity (82% ee, b/l ) 86).
Ligand 1 was also recently found to lead to quite high
enantio- and regioselectivity in the asymmetric hydroformy-
2
lation of allyl cyanide. Surprisingly, Kelliphite also led to
very high regioselectivity (b/l ) 68.3 at 35 °C) for styrene
hydroformylation; however, the reaction occurred with only
marginal enantioselectivity.
In addition to increased throughput, an additional benefit
of multisubstrate screening in parallel reactors is that relative
reaction rates of different substrates can be measured in one
experiment. The data in Table 2 indicate that allyl cyanide
is the most reactive olefin for all hydroformylation catalysts
in this study. Vinyl acetate is a less reactive substrate and
exhibits lower conversions under these conditions.
Acknowledgment. We thank Drs. Cynthia Rand and Ian
Lennon from Dowpharma for helpful discussions and Mari-
lyn Bokota and Dr. Duane Romer for assistance with PPR-
48 experiments.
14
Supporting Information Available: Experimental pro-
cedures and hydroformylation data for control experiments
using Chiraphite and Kelliphite. This material is available
free of charge via the Internet at http://pubs.acs.org.
We have further increased the throughput of our multi-
substrate screens by utilizing larger arrays of parallel reactors.
OL0487938
(
13) Reaction conditions: VA:Rh ) 2000, neat VA (4 mL), temp ) 60
C, time ) 10 h, 100% conversion.
14) Conversions obtained from experiments with separate olefins (see
(15) (a) Turner, H.; Dales, G.; Vanerden, L.; Van Beek, J. Symyx
Technologies Inc. U.S. Patent 6,306,658, 2001. (b) Tuinstra, H. E.;
Cummins, C. H. AdV. Mater. 2000, 12, 1819. (c) Peil, K. P.; Neithamer, D.
R.; Patrick, D. W.; Wilson, B. E.; Tucker, C. J. Macromol. Rapid Commun.
2004, 25, 119.
°
(
Supporting Information) are very similar to those obtained for the mixture
of olefins.
3280
Org. Lett., Vol. 6, No. 19, 2004