Mass spectrometric screening of enantioselective Diels-Alder reactions

Article abstract of DOI:10.1002/anie.200705082

(Chemical Equation Presented) A rapid cat. scan: Mass spectrometric monitoring of reaction intermediates of the retro-Diels-Alder reaction has allowed the rapid screening of catalysts for enantioselective Diels-Alder reactions (see scheme). Copper catalysts as well as metal-free organocatalysts were tested. A protocol for the simultaneous screening of catalyst mixtures has also been developed, which offers new possibilities for high-throughput catalyst development.

Full text of DOI:10.1002/anie.200705082

Communications
DOI: 10.1002/anie.200705082
Catalyst Screening (1)
Mass Spectrometric Screening of Enantioselective Diels–Alder
Reactions**
Antje Teichert and Andreas Pfaltz*
In the preceding communication[1] we introduced a new
concept for screening chiral catalysts by mass spectrometric
monitoring of catalytic reactions in the reverse direction. The
use of mass-labeled quasienantiomeric[2] products as reactants
in the screening process makes it possible to distinguish
catalyst-bound intermediates with an opposite sense of
chirality by mass spectrometry. According to the principle
of microscopic reversibility, the transition states of the
forward and the back reaction are identical and, therefore,
the enantioselectivity determined from back-reaction screen-
ing is identical to that of the forward reaction. After having
demonstrated the feasibility of this method for palladium-
catalyzed allylic substitutions, we report now an extension to
metal-catalyzed and organocatalytic Diels–Alder (DA) reac-
tions.
The principle of our method is illustrated in Scheme 1.
When a mixture of two mass-labeled quasienantiomeric
Diels–Alder products 1a and 1b (R1 ¼ R2) is treated with a
chiral cationic Lewis acid L*M+, the two positively charged
catalyst complexes 2a and 2b are formed, which can be
distinguished by ESI-MS because of their different molecular
masses. ALewis acid which catalyzes the DAreaction of the
dienophiles 4a and 4b also catalyzes the retro-DAreaction.
Therefore, under appropriate conditions, the two catalyst–
dienophile complexes 3a and 3b formed by loss of cyclo-
pentadiene should be observable together with the complexes
2a and 2b. Given that the quasienantiomers 1a and 1b
behave like real enantiomers, and if formation of the
complexes 2a and 2b from 1a and 1b is fast and reversible,
Scheme 1. Retro-Diels–Alder reaction of two quasienantiomeric Diels–
and if the subsequent retro-DAreaction is slow and
irreversible, then the ratio 3a/3b directly reflects the enan-
Alder products 1a and 1b. R1 and R2 represent mass labels.
tioselectivity of the catalyst. At elevated temperature under
dilute conditions, especially during the initial phase of the
retro-DAreaction when the concentrations of 4a, 4b, and
cyclopentadiene are low compared to that of 1a and 1b, the
DAreaction of 3a and 3b with cyclopentadiene is expected to
be very slow, so the requirement of an essentially irreversible
retro-DAreaction should be fulfilled. As ESI-MS allows the
selective detection of charged species in the presence of a
large excess of neutral compounds, it should be possible to
monitor the signals of 3a and 3b even at low concentration.
As quasienantiomeric DA products we chose the 4-
ethylphenyl and 4-butylphenyl derivatives 1a and 1b
(Figure 1), which were readily obtained in high enantiomeric
purity by a [Cu(box)]-catalyzed DAreaction according to the
procedure described by Evans et al. (box = bisoxazoline,
(S,S)- or (R,R)-5).[3] As the para-substituted ethyl- and
butylphenyl groups of 1a and 1b have essentially identical
steric and electronic properties, we expected the two quasi-
enantiomers to behave like real enantiomers in the retro-DA
reaction.
Preliminary experiments showed that the retro-DAreac-
tion was sufficiently fast at 1008C to allow detection of the
dienophile–catalyst complexes 3a and 3b by ESI-MS. When a
1:1 mixture of quasienantiomers (2R)-1a and (2S)-1b was
treated with 20 mol% of the enantiomerically pure catalyst
(S,S)-5 in dichloromethane at 1008C for 1 h, ESI-MS analysis
[*] Dipl.-Chem. A. Teichert, Prof. Dr. A. Pfaltz
Department of Chemistry
University of Basel
St. Johanns-Ring 19, 4056 Basel (Switzerland)
Fax: (+41)61-267-1103
E-mail: andreas.pfaltz@unibas.ch
[**] Financial support fromthe Swiss National Science Foundation is
gratefully acknowledged.
Supporting information for this article is available on the WWW
under http://www.angewandte.org or fromthe author.
3360
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2008, 47, 3360 –3362

Angewandte
Chemie
Scheme 2. The preparative DA reaction.
phase, was 64:36, which is in good agreement with the
screening results.
We next screened a series of differently substituted box
ligands under these conditions (Table 1). The results obtained
Table 1: ESI-MS screening of CuII-box catalysts.[a]
Ligand
ESI-MS
HPLC[b]
Preparative reaction[c]
3a/3b
4a/4b
endo e.r.
5a (R=2-naphthyl)
5b (R=Ph)
5c (R=tBu)
5d (R=iPr)
5e (R=benzyl)
86:14
80:20
68:32
61:39
55:45
90:10
85:15
69:31
58:42
51:49
88:12
86:14
64:36
56:44
52:48
[a] Conditions for ESI-MS screening: 1008C, 1 h, 0.02m in CH2Cl2. The
ratio of 3a/3b was determined by integration of the signals correspond-
ing to the major isotope. [b] The reaction mixture was filtered through a
plug of silica gel and then subjected to HPLC analysis on a Chiralcel OD-
H column. [c] 0.1m solution of 4b (R2 =4-butylphenyl), cyclopentadiene
(10 equiv), 20 mol% CuII catalyst, CH2Cl2, 1 h, 1008C. Product 1b was
analyzed by HPLC on a Chiralcel OD-H column.
Figure 1. ESI-MS screening of the copper-catalyzed retro-DA reaction.
by ESI-MS were again in good agreement with the HPLC
analyses and showed the same selectivity order 5a > 5b >
5c > 5d > 5e for the different box ligands, with the highest
enantiomeric ratio for the 2-naphthyl-substitued derivative
5a. The small deviations observed between the ESI-MS and
HPLC data are likely due to a relatively high background
noise in the MS spectra caused by traces of unknown Cu
species and/or errors in the HPLC analyses. The same
selectivity order and very similar enantiomeric ratios for the
of the reaction mixture showed a 68:32 ratio of 3a/3b at
m/z 634 and m/z 662, thus indicating a preference of the
catalyst for (2R)-1a. Essentially the reverse signal ratio of
31:69 was observed with the enantiomeric R,R catalyst, as
expected. A50:50 ratio was observed for 2a/2b at m/z 700
and m/z 728. An additional signal cluster at m/z 651 was
attributed to a complex with the composition [Cu(5)2]+. A ll endo product were obtained in the analogous preparative DA
the copper complexes were readily identified by their
characteristic isotope pattern. Interestingly, only CuI com-
plexes were observed, even though the catalyst was prepared
from copper(II)triflate. Similar observations have previously
been reported and attributed to redox reactions occuring
during the electrospray process.[4]
The enantioselectivities determined by ESI-MS were
confirmed by HPLC analysis of the reaction mixtures, which
showed virtually the same dienophile ratios 4a/4b as the
corresponding mass spectrometric signal ratios 3a/3b
(Figure 1). The reaction was also carried out in the forward
direction on a preparative scale with the 4-butylphenyl-
substituted dienophile 4b (Scheme 2). Except for the higher
concentration of the dienophile (0.1m instead of 0.02m), the
conditions used were the same as in the screening experi-
ments. The enantiomeric ratio of the endo product 1b,
measured by HPLC on a column with a chiral stationary
reactions performed at the same temperature (1008C).
Higher ee values and a different selectivity order was
observed when the reactions were carried out at ꢀ358C (5a
(97% ee) > 5c (95% ee) > 5b (89% ee) > 5d (23% ee) > 5e
(17% ee)). These results are in contrast to the findings of
Evans et al.,[3] who found that the tert-butylbox derivative 5c
was by far the most efficient ligand of this series in the
analogous DAreaction with an unsubstituted acrylimide at
ꢀ358C. To demonstrate the scope of our screening method,
we tested a number of additional chiral ligands, including
phox[5] and bisimine[6] ligands, which all gave only moderate to
low enantioselectivities. Again, the results were in good
agreement with HPLC analyses of the reaction mixtures and
the corresponding preparative DAreactions.
We then turned to organocatalytic DAreactions, which
should be ideal for ESI-MS screening because metal-free
catalysts have a much narrower isotope distribution than Cu
Angew. Chem. Int. Ed. 2008, 47, 3360 –3362
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
3361

Communications
or other transition-metal catalysts and, therefore, cause less
problems with respect to overlapping signals. As a first test,
we chose the DAreaction of cinnamaldehyde with cyclo-
pentadiene in the presence of imidazolidinone 6·HCl as the
catalyst. This reaction, originally developed by MacMillan
and co-workers,[7] proceeds via iminium ion intermediates,
which should be observable by ESI-MS. The required
quasienantiomeric DAproducts (2 S)-7a and (2R)-7b were
readily prepared in enantiomerically pure form by reduction
of the corresponding DAproducts 1. Back-reaction screening
was performed in dichloromethane at room temperature by
using 20 mol% of the catalyst. After a reaction time of 24 h,
ESI-MS analysis of the reaction mixture showed the signals of
the iminium ions 8a and 8b at m/z 375 and m/z 389,
respectively, in a 88:12 ratio (Figure 2). The control experi-
Figure 3. Simultaneous screening of three organocatalysts.
screening individual catalysts. These results clearly show that
multicatalyst screening in solution is indeed feasible, which
opens up new possibilities for the high-throughput screening
of organocatalyst libraries of this type.
On the basis of the results obtained for the palladium-
catalyzed allylic substitution reactions[1] and DAreactions, we
expect back-reaction screening by ESI-MS to be applicable to
various other classes of reactions. In organocatalysis, for
example, there are many reactions which proceed through
cationic intermediates and, therefore, should be amenable to
ESI-MS screening. The option to evaluate mixtures of
catalysts simultaneously offers the possibility to screen
catalyst or ligand libraries prepared by combinatorial meth-
ods without the need to isolate each catalyst or ligand. Hence,
this strategy could speed up ligand development considerably.
Received: November 2, 2007
Revised: February 8, 2008
Published online: March 20, 2008
Figure 2. ESI-MS screening of organocatalysts.
Keywords: asymmetric catalysis · high-throughput screening ·
homogeneous catalysis · mass spectrometry · retro-Diels–
Alder reaction
.
ment with inversely labeled quasienantiomers (Pr and Bu
interchanged) gave the expected reverse ratio of 12:88. In line
with these results, an enantiomer ratio of 87:13 was measured
for the endo product of the corresponding preparative DA
reaction with cinnamaldehyde at room temperature. The ESI-
MS spectra were much cleaner than those observed for the
copper-catalyzed reactions, with no signal overlap and with
higher signal-to-noise ratios.
With a reliable screening protocol at hand, we tested
whether mixtures of different organocatalysts could be
screened simultaneously. The results obtained with a solution
of the three catalysts 9, 10·HCl, and 6·HCl are shown in
Figure 3. Spectra recorded after a reaction time of 1 h at 508C
showed signal ratios for the corresponding dienophile–
catalyst adducts that closely matched the data obtained by
[1] C. Müller, A. Pfaltz, Angew. Chem. 2008, 120, 3411 – 3414; Angew.
Chem. Int. Ed. 2008, 18, 3363 – 3366.
[2] Q. Zhang, D. P. Curran, Chem. Eur. J. 2005, 11, 4866 – 4880.
[3] D. A. Evans, S. J. Miller, T. Lectka, P. von Matt, J. Am. Chem. Soc.
1999, 121, 7559 – 7573.
[4] a) A. T. Blades, P. Jayaweera, M. G. Ikonomou, P. Kebarle, J.
Chem. Phys. 1990, 92, 5900 – 5906; b) L. Gianelli, V. Amendola, L.
Fabbrizzi, P. Pallavicini, G. G. Mellerio, Rapid Commun. Mass
Spectrom. 2001, 15, 2347 – 2353; c) J. Shen, J. Brodbelt, J. Mass
Spectrom. 1999, 34, 137 – 146.
[5] I. Sagasser, G. Helmchen, Tetrahedron Lett. 1998, 39, 261 – 263.
[6] D. A. Evans, T. Lectka, S. J. Miller, Tetrahedron Lett. 1993, 34,
7027 – 7030.
[7] K. A. Ahrendt, C. J. Borths, D. W. C. MacMillan, J. Am. Chem.
Soc. 2000, 122, 4243 – 4244.
3362
www.angewandte.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2008, 47, 3360 –3362