Journal of the American Chemical Society
COMMUNICATION
derived from ligand 4f, no intermediates were observed. Pre-
parative experiments were consistent with this result, as they
showed that under the conditions used for screening, less than
5% product had formed after 24 h. Apparently, the sterically
demanding anthracenyl-substituted ligand does not form an
active catalyst.
’ ACKNOWLEDGMENT
Financial support by the Swiss National Science Foundation is
gratefully acknowledged. We thank Simon Allmendinger for his
synthetic contributions.
’ REFERENCES
The results in Table 1 and the very good linear correlation
between the calculated ee values from racemic catalyst screening
and the actual ee values from the corresponding enantiopure
catalysts demonstrate that it is possible to determine the en-
antioselectivity of a chiral catalyst from the racemic form in a
reliable manner. Especially for catalysts providing ee's in the
50ꢀ100% range, the results are remarkably accurate.7 However,
two questions remain: why are the ee values obtained from the
racemic catalysts about 25% lower than the actual enantioselec-
tivities determined from the enantiopure catalysts, and why do
the data from the racemic and the enantiopure catalysts show this
linear correlation? We assume that the ee difference results from
a deviation from ideal pseudo-zeroth-order kinetics.3,4 If the
turnover rates of the two enantiomeric catalysts show some
dependence on the substrate concentration, the proportion of
intermediate 3b derived from the more abundant quasienantio-
mer 2b should be higher than that under pseudo-zeroth-order
conditions, resulting in a lower predicted selectivity when eq 1 is
applied. In this case, the observed linear correlation would imply
that the deviation from pseudo-zeroth-order conditions affects
the ee values of the different catalysts by the same degree.
Our assumption is supported by the observed dependence of
the ee difference on the substrate/catalyst (S/C) ratio (see Table 2
in the SI).10 The calculated ee values for the racemic ligand 2e
increased from 58 to 76% when the S/C ratio was raised from
25:1 to 200:1. As expected, at lower catalyst loading, the devia-
tion from the enantioselectivity of the enantiopure catalyst was
smaller. However, because of the low signal intensities at S/C
ratios of g100:1, the results proved to be less accurate than with
the standard 50:1 ratio generally used. These results indicate that
even under conditions that deviate from an ideal pseudo-zeroth-
order regime, reliable ee values can be obtained.
(1) For reviews, see: (a) Reetz, M. T. Angew. Chem. 2008, 120, 2592;
Angew. Chem., Int. Ed. 2008, 47, 2556. (b) de Vries, J. G.; de Vries,
A. H. M. Eur. J. Org. Chem. 2003, 799. (c) Hagemeyer, A.; Jandeleit, B.;
Liu, Y.; Poojary, D. M.; Turner, H. W.; Volpe, A. F.; Weinberg, W. H.
Appl. Catal., A 2001, 221, 23.
(2) Lagasse, F.; Tsukamoto, M.; Welch, C. J.; Kagan, H. B. J. Am.
Chem. Soc. 2003, 125, 7490.
(3) Dominguez, B.; Hodnett, N. S.; Lloyd-Jones, G. C. Angew. Chem.
2001, 113, 4419; Angew. Chem., Int. Ed. 2001, 40, 4289.
(4) Blackmond, D. G.; Hodnett, N. S.; Lloyd-Jones, G. C. J. Am.
Chem. Soc. 2006, 128, 7450.
(5) (a) Markert, C.; Pfaltz, A. Angew. Chem. 2004, 116, 2552; Angew.
Chem., Int. Ed. 2004, 43, 2498; (b) Markert, C.; R€osel, P.; Pfaltz, A. J. Am.
Chem. Soc. 2008, 130, 3234; (c) M€uller, C. A.; Pfaltz, A. Angew. Chem.
2008, 120, 3411; Angew. Chem., Int. Ed. 2008, 47, 3363; (d) Teichert, A.;
Pfaltz, A. Angew. Chem. 2008, 120, 3408; Angew. Chem., Int. Ed. 2008,
47, 3360. (e) Fleischer, I.; Pfaltz, A. Chem.—Eur. J. 2010, 16, 95. (f) For a
review, see: M€uller, C. A.; Markert, C.; Teichert, A. M.; Pfaltz, A. Chem.
Commun. 2009, 1607.
(6) Koch, G.; Lloyd-Jones, G. C.; Loiseleur, O.; Pfaltz, A.; Pretot, R.;
Schaffner, S.; Schnider, P.; von Matt, P. Recl. Trav. Chim. Pays-Bas 1995,
114, 206.
(7) The error margins in the data of the less selective catalysts
(ligands 4a and 4b) are relatively high because the difference between
the measured signal ratio, I(3a)/I(3b), and the ratio of quasienantio-
meric substrates, Q, is extremely small in the 0ꢀ30% ee range [e.g., for
20% ee and Q = 25:75, I(3a)/I(3b) is 26:74]. Consequently, small
measurement errors lead to large errors in the calculated ee values. In five
reactions, the ee values for 4a varied from 17 to 26% (see Table 1 in the
SI). Values in the 50ꢀ100% ee range are much more accurate because
they are based on a substantially larger difference between the signal
ratio and Q. For ligand 4e, the values from four screening experiments
were between 61 and 62% ee (see Table 1 in the SI).
(8) Theoretically, the line should go through the zero point. The
observed deviation probably results from the relatively large error
margins of the ee values below 30% ee (see ref 7).
In summary, we have shown that on the basis of the concept of
Lloyd-Jones,3 the enantioselectivity of a chiral catalyst can be
readily determined from its racemic form by mass spectrometric
screening of a nonequal mixture of two quasienantiomeric
substrates. The experimental protocol is simple and fast, as no
workup or product isolation is necessary. As we have demon-
strated for the screening of enantiopure catalysts,5 it should also
be possible to evaluate mixtures of racemic catalysts simulta-
neously. Our method should be a valuable addition to existing
screening methods, especially for evaluating new catalyst struc-
tures that are not readily available in enanantiomerically pure form.
(9) (a) Dotta, P.; Kumar, P. G. A.; Pregosin, P. S.; Albinati, A.;
Rizzato, S. Organometallics 2004, 23, 2295. (b) Dotta, P.; Magistrato, A.;
Rothlisberger, U.; Pregosin, P. S.; Albinati, A. Organometallics 2002,
21, 3033. (c) Selvakumar, K.; Valentini, M.; Pregosin, P. S.; Albinati, A.;
Eisentr€ager, F. Organometallics 2000, 19, 1299. (d) Tschoerner, M.;
Pregosin, P. S.; Albinati, A. Organometallics 1999, 18, 670. (e) Trabesinger,
G.; Albinati, A.; Feiken, N.; Kunz, R. W.; Pregosin, P. S.; Tschoerner, M.
J. Am. Chem. Soc. 1997, 119, 6315.
(10) We thank one of the reviewers for suggesting these
experiments.
’ ASSOCIATED CONTENT
S
Supporting Information. Experimental procedures, char-
b
acterization data for all reactions and products, and detailed
screening procedure and results. This material is available free of
’ AUTHOR INFORMATION
Corresponding Author
4713
dx.doi.org/10.1021/ja111700e |J. Am. Chem. Soc. 2011, 133, 4710–4713