Quantitatively correlating the effect of ligand-substituent size in asymmetric catalysis using linear free energy relationships

Article abstract of DOI:10.1002/anie.200704257

Is bigger better? By systematically evaluating the size of the substituent G in a modular ligand (see picture), a linear free energy relationship was observed relating the product enantiomeric ratio to steric parameters developed by Charton. The reaction

Full text of DOI:10.1002/anie.200704257

Angewandte
Chemie
DOI: 10.1002/anie.200704257
Enantioselectivity
Quantitatively Correlating the Effect of Ligand-Substituent Size in
Asymmetric Catalysis Using Linear Free Energy Relationships**
Jeremie J. Miller and Matthew S. Sigman*
In the past three decades, the discovery and application
of effective asymmetric catalysts has flourished in both
academic and industrial settings.[1] This is especially true
within the recent decade where the field of asymmetric
catalysis has witnessed an incredible expansion, and the
benchmarks for acceptable asymmetric induction have
significantly risen. However, the approach to catalyst
identification remains mainly empirical, wherein evalu-
ation of a considerable number of catalyst structures is
required to develop a mature chiral catalyst. While this
process has been expedited by technological advance-
ments,[2] the understanding of the relationship between
catalyst structure and the enantioselectivity of the
catalyzed reaction is generally precarious. This can be
attributed to the small energy differences in the
activation barrier leading to the enantiomeric products
and/or the complicated preequilibria in most catalytic
reactions.[3]
An approach to developing greater predictive power
in asymmetric catalysis is based on the use of modular
ligands[4] in lieu of a detailed understanding of the
Figure 1. Effect of the proline module on facial selectivity in the Cr-catalyzed
addition of allyl bromide to carbonyl compounds. Boc=tert-butyloxycarbonyl,
TEA=triethylamine, TMS=trimethylsilyl, TBAF=tetrabutylammonium fluo-
ride.
microscopic processes involved in asymmetric induc-
tion. We believe the ability to make systematic changes
to the ligand structure not only leads to enhanced
product enantiomeric excess,[5] but allows for the
elucidation of structure–enantioselectivity relationships.[6]
Herein, we report the use of a modular ligand set to
systematically study the effect of ligand size and the first
observation of a linear free energy relationship between steric
parameters and enantiomeric ratio. Additionally, our analysis
of several previously reported enantioselective reactions
reveals similar linear free energy relationships.
We have recently disclosed the use of modular oxazoline
ligands[7] in enantioselective Nozaki–Hiyama–Kishi reactions
of allylic halides[8,9] with both aldehydes[5a] and ketones[5c]
(Figure 1). A key observation from these studies is that the
facial selectivity of the reaction is directly correlated to the
absolute configuration of the proline module of the ligand.
For example, in comparing ligands 1a and 1b, which differ
only in the configuration of the proline module, a consid-
erable difference in enantiomeric ratio is observed for
allylation of both aldehydes and ketones. Most notably, the
allylation of benzaldehyde is promoted in nearly 90% ee
(96:4 e.r. and 94.5:5.5 e.r.) using both ligands but with
different senses of asymmetric induction. A similar effect is
also found for the allylation of acetophenone. Systematic
changes to the configuration of the other chiral centers in the
ligand structure had only a subtle impact. These observations
prompted us to explore the role of the proline module in
greater detail.
In order to initiate the investigation, we decided to make
systematic changes to the proline module and probe the effect
on the enantioselective outcome of the reaction. This was
reasoned to be most easily accomplished through modifica-
tion of the proline nitrogen, as proline carbamates can be
either purchased from commercial sources or readily pre-
pared by treatment of proline with the corresponding
chloroformate.[10] Ligand 1a, the optimal diastereomer for
benzaldehyde allylation, was chosen as a starting structure to
which carbamate modifications were made. Ligands 2a–2d
were synthesized from a common precursor,[10] by varying the
size of the carbamate substituent from a small group (methyl)
[*] J. J. Miller, Prof. M. S. Sigman
Department of Chemistry
University of Utah
315 South 1400 East
Salt Lake City, UT 84112 (USA)
Fax: (+1)801-581-8433
E-mail: sigman@chem.utah.edu
[**] This work was supported by the National Science Foundation
through a CAREER award to M.S.S. (CHE-132905). M.S.S. thanks
the Dreyfus foundation (Teacher-Scholar award) and Pfizer for their
support.
Supporting information for this article is available on the WWW
under http://www.angewandte.org or from the author.
Angew. Chem. Int. Ed. 2008, 47, 771 –774
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
771

Communications
to a large group (adamantyl); the ligands were then initially
evaluated for benzaldehyde allylation (Table 1). A trend
between the relative size of the carbamate substituents and
Table 1: Systematic evaluation of proline carbamate ligands in Nozaki–
Hiyama–Kishi allylation reactions of three carbonyl substrates.
Figure 2. Three linear free energy relationships between Charton steric
parameters and log of enantiomeric ratio (R/S).
rates)[11] and modified by Charton (based on van der Waals
radii, plotted in Figure 2) were utilized.[12] Taft steric param-
eters have been commonly relied upon in the optimization of
drug candidates.[13] Excellent correlations are observed for
allylation of benzaldehyde, acetophenone, and the aliphatic
aldehyde hydrocinnamaldehyde. The slope (y as defined by
Charton) is a sensitivity factor much like a 1 value in a
Hammett plot. The sensitivity is the greatest for benzalde-
hyde, but considerable for the other two substrates as well.
Varying the size of the carbamate substituent could cause a
number of structural changes, both subtle and global, in the
diastereomeric transition states. However, the observation of
linear free energy relationships is consistent with only one
structural perturbation occurring upon modification of the
carbamate. To the best of our knowledge, this is the first such
reported correlation observed in asymmetric catalysis.[14]
Considering the volume of reported asymmetric catalytic
reactions and variety of chiral ligands used, we were curious
whether our current observation is unique. Perusal of the
literature revealed several examples where a general trend
was observed for the ligand-substituent size versus enantio-
meric ratio. A noteworthy example was found by performing
a similar analysis of the enantiomeric ratios from the seminal
report by Pfaltz and von Matt concerning the use of
phosphine oxazoline ligands in asymmetric Pd-catalyzed
allylic alkylation reactions (Figure 3a).[15] We found that the
size of the ligand substituent on the oxazoline can be
correlated to product enantiomeric ratio with y = 0.35.[16]
Another example, the enantioselective cyclopropanation of
allylic alcohols with bis(sulfonamide)-derived chiral diamines
and bis(halomethyl)zinc reagents reported by Denmark and
co-workers, was analyzed (Figure 3b).[17] Again, a linear
correlation in the Charton plot is observed. However, in this
case, y is negative, consistent with a smaller ligand substituent
leading to a higher enantiomeric ratio. This is in contrast to a
paradigm in ligand design that bigger is better.
Entry
G
R1
R2 e.r. (R/S)[a] Charton value (u)[b]
1
2
3
4
5
6
7
8
Me
Et
iPr
tBu
Ph
Ph
Ph
Ph
H
H
H
H
H
H
H
H
H
H
1.5
1.9
3.5
23
23
1
0.52
0.56
0.76
1.24
1.33
0.52
0.56
0.76
1.24
1.33
0.52
0.56
0.76
1.24
1.33
1-adamantyl Ph
Me
Et
iPr
tBu
1-adamantyl PhCH2CH2
Me
Et
iPr
tBu
PhCH2CH2
PhCH2CH2
PhCH2CH2
PhCH2CH2
1.1
1.3
2.8
2.8
9
10
11
12
13
14
15
Ph
Ph
Ph
Ph
Me 0.3
Me 0.3
Me 0.6
Me 2.2
Me 2.6
1-adamantyl Ph
[a] Enantiomeric ratio determined by HPLCequipped with a chiral
stationary phase. [b] See reference [12a] for all Charton values except for
1-adamantyl.[12b]
the enantiomeric ratio of the product was observed. The
range is noteworthy (entries 1–5, Table 1) in that 20% ee
(60:40 e.r.) is observed for the methyl substituent (entry 1)
versus 92% ee (96:4 e.r.) for the 1-adamantyl substituent
(entry 5); this corresponds to a difference of 1.62 kcalmolÀ1 at
258C in the diastereomeric transition states.
Based on this initial success, we also evaluated two other
substrates of interest: an aliphatic aldehyde, a particularly
difficult substrate class for our current catalytic systems, and a
simple ketone, acetophenone. Again, a correlation between
the size of the group and the enantiomeric ratio is observed in
both cases. A change in facial selectivity (e.r. < 1) is observed
for acetophenone allylation (entries 11–15, Table 1) when
going from smaller to larger G substituents. Notably, a similar
magnitude of asymmetric induction for the catalyst contain-
ing the smallest group, 2a, and the largest group, 2d, is
observed. This result highlights the capricious nature of the
optimization process in asymmetric catalysis.
Correlating the size of the substrate to product enantio-
meric ratio should also be possible, especially considering that
the origin of Taft/Charton values is based on the effect of
substituent size on the rate of ester hydrolysis.[11,12] As an
example, the enantioselective aziridination of styrenes with
Mn salen complexes reported by Komatsu et al. was analyzed
(Figure 4).[18] In this case, the reaction of styrene results in a
The most compelling aspect of these data is the ability to
observe a linear free energy relationship by quantitatively
correlating steric parameters to the log of enantiomeric ratio
(a relative rate; Figure 2). To accomplish this, the steric
parameters reported by Taft (based on ester hydrolysis
772
www.angewandte.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2008, 47, 771 –774

Angewandte
Chemie
significantly lower enantiomeric ratio than that resulting from
substrates with substitution trans in the b position. An
excellent correlation of the size of the group in the b position
of the styrene is observed with y = 1.4.
We believe that the use of a Charton analysis should
provide the opportunity to consider modifications to the
ligand structure in a more predictable manner where it could
become a useful tool for asymmetric catalyst optimization. Of
particular note, the sensitivity factor y provides the practi-
tioner with the ability to estimate the impact of making a
change in substituent size and whether the synthetic endeav-
our may be worthwhile. For example, in our evaluation of
hydrocinnamaldehyde, we found that the y value was
relatively small and even significant changes to the size of
the carbamate substituent are not anticipated to lead to high
enantiomeric excess. This information is prompting us to
think about other changes to the ligand design to obtain high
enantioselectivity. Additionally, a Charton plot could provide
insight into the transition state much like a Hammett plot.
Specifically, the observation of a linear free energy relation-
ship implies only a single structural change in the diastereo-
meric transition state is occurring by modification of the
substituent.
In conclusion, we have reported the ability to construct
linear free energy relationships of steric parameters and
enantiomeric ratio for enantioselective carbonyl allylation
reactions using modular oxazoline ligands developed within
our laboratory. We have also extended this type of analysis to
previously reported asymmetric reactions where substituent
effects on the catalyst or the substrate can be correlated.
Experiments to explore the origin of the Charton correlation
in our system as well as others and exploitation of the linear
free energy relationship in catalyst design are currently under
investigation.
Figure 3. a) Charton plot of Pfaltz’s enantioselective Pd-catalyzed allylic
allylation using phosphine oxazoline ligands. b): Charton plot of
Denmark’s enantioselective cyclopropanation with bis(halomethyl)zinc
reagents using bis(sulfonamide)-derived chiral diamines.
Received: September 14, 2007
Published online: December 11, 2007
Keywords: allylation · asymmetric catalysis · enantioselectivity ·
.
linear free energy relationships
[1] For monographs, see: a) H. U. Blaser, E. Schmidt, in Asymmetric
Catalysis on Industrial Scale: Challenges, Approaches and
Solutions, Wiley-VCH, Weinheim, 2004; b) Catalytic Asymmet-
ric Synthesis, 2nd ed. (Ed.: I. Ojima), Wiley-VCH, Weinheim,
2000; c) E. N. Jacobsen, A. Pfaltz, H. Yamamoto, Comprehen-
sive Asymmetric Catalysis I–III, Vol. 1–3, 1999; d) R. Noyori,
Asymmetric Catalysis in Organic Synthesis, Wiley-Interscience,
New York, 1994.
[2] a) C. Jäkel, R. Paciello, Chem. Rev. 2006, 106, 2912 – 2942;
b) J. P. Stambuli, J. F. Hartwig, Curr. Opin. Chem. Biol. 2003, 7,
420 – 426; c) K. D. Shimizu, M. L. Snapper, A. H. Hoveyda,
Comprehensive Asymmetric Catalysis I–III, Vol. 3, Springer,
Berlin, 1999, pp. 1389 – 1399.
[3] a) J. A. Mueller, A. Cowell, B. D. Chandler, M. S. Sigman, J. Am.
Chem. Soc. 2005, 127, 14817 – 14824; b) J. Halpern, Science 1982,
217, 401 – 407.
[4] For examples: a) A. H. Hoveyda, A. W. Hird, M. A. Kacprzyn-
ski, Chem. Commun. 2004, 1779 – 1785; b) S. J. Miller, Acc.
Chem. Res. 2004, 37, 601 – 610.
Figure 4. Charton plot of enantioselective alkene aziridination with
Mn-salen reported by Komatsu et al.
Angew. Chem. Int. Ed. 2008, 47, 771 –774
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
773

Communications
[5] a) J.-Y. Lee, J. J. Miller, S. S. Hamilton, M. S. Sigman, Org. Lett.
2007, 5, 763 – 766; j) M. Inoue, M. Nakada, Heterocycles 2007, 72,
133 – 138.
2005, 7, 1837 – 1839; b) S. Rajaram, M. S. Sigman, Org. Lett.
2005, 7, 5473 – 5475; c) J. J. Miller, M. S. Sigman, J. Am. Chem.
Soc. 2007, 129, 2752 – 2753.
[10] See the Supporting Information for details.
[11] a) R. W. Taft, Jr., J. Am. Chem. Soc. 1952, 74, 3120 – 3128;
b) R. W. Taft, Jr., J. Am. Chem. Soc. 1953, 75, 4538 – 4539;
c) R. W. Taft, Jr., Steric Eff. Org. Chem. 1956, 556 – 675; d) J. A.
MacPhee, A. Panaye, J.-E. Dubois, Tetrahedron 1978, 34, 3553 –
3562.
[12] a) M. Charton, J. Am. Chem. Soc. 1975, 97, 1552 – 1556; b) D. S.
Kristol, R. C. Parker, H. D. Perlmutter, K.-C. H. Chen, D. H.
Hawes, G. H. Wahl, Jr., J. Org. Chem. 1976, 41, 3205 – 3206.
[13] C. Hansch, A. Leo, Exploring QSAR: Fundamentals and
Applications in Chemistry and Biology, American Chemical
Society, Washington, DC, 1995.
[14] For an example in a diastereoselective process, see: B. Meyn-
hardt, U. Luning, C. Wolff, C. Nather, Eur. J. Org. Chem. 1999,
2327 – 2335.
[15] P. von Matt, A. Pfaltz, Angew. Chem. 1993, 105, 614 – 615;
Angew. Chem. Int. Ed. Engl. 1993, 32, 566 – 568.
[16] It should be noted that phenyl has two Charton values, and the
value that resulted in the best data fit was utilized.
[17] S. E. Denmark, B. L. Christenson, S. P. OꢀConnor, Tetrahedron
Lett. 1995, 36, 2219 – 2222.
[6] K. H. Jensen, M. S. Sigman, Angew. Chem. 2007, 119, 4832 –
4834; Angew. Chem. Int. Ed. 2007, 46, 4748 – 4750.
[7] S. Rajaram, M. S. Sigman, Org. Lett. 2002, 4, 3399 – 3401.
[8] For reviews of CrII-mediated reactions, see: a) L. A. Wessjo-
hann, G. Scheid, Synthesis 1999, 1 – 36; b) A. Fürstner, Chem.
Rev. 1999, 99, 991 – 1045; c) P. Cintas, Synthesis 1992, 248 – 257.
[9] For examples of catalytic asymmetric Nozaki–Hiyama–Kishi
aldehyde allylation reactions, see: a) M. Bandini, P. G. Cozzi, P.
Melchiorre, A. Umani-Ronchi, Angew. Chem. 1999, 111, 3558 –
3561; Angew. Chem. Int. Ed. 1999, 38, 3357 – 3359; b) M.
Bandini, P. G. Cozzi, A. Umani-Ronchi, Chem. Commun. 2002,
919 – 927; c) H.-w. Choi, K. Nakajima, D. Demeke, F.-A. Kang,
H.-S. Jun, Z.-K. Wan, Y. Kishi, Org. Lett. 2002, 4, 4435 – 4438;
d) A. Berkessel, D. Menche, C. A. Sklorz, M. Schroder, I.
Paterson, Angew. Chem. 2003, 115, 1062 – 1065; Angew. Chem.
Int. Ed. 2003, 42, 1032 – 1035; e) M. Inoue, M. Nakada, Org. Lett.
2004, 6, 2977 – 2980; f) A. Berkessel, M. Schroeder, C. A. Sklorz,
S. Tabanella, N. Vogl, J. Lex, J. M. Neudoerfl, J. Org. Chem. 2004,
69, 3050 – 3056; g) G. Xia, H. Yamamoto, J. Am. Chem. Soc.
2006, 128, 2554 – 2555; h) H. A. McManus, P. G. Cozzi, P. J.
Guiry, Adv. Synth. Catal. 2006, 348, 551 – 558; i) G. C. Hargaden,
H. A. McManus, P. G. Cozzi, P. J. Guiry, Org. Biomol. Chem.
[18] S. Minakata, T. Ando, M. Nishimura, I. Ryu, M. Komatsu,
Angew. Chem. 1998, 110, 3596 – 3598; Angew. Chem. Int. Ed.
1998, 37, 3392 – 3394.
774
www.angewandte.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2008, 47, 771 –774