Mass Spectrometric Screening of Racemic Amine Catalysts for Enantioselective Michael Additions

Article abstract of DOI:10.1002/adsc.201500224

In extension of a concept of Lloyd-Jones, based on the combination of a racemic catalyst with a scalemic substrate, we have recently developed a method for determining the enantioselectivity of a chiral catalyst from its racemic form by mass spectrometric screening of a non-equal mixture of two mass-labeled quasi-enantiomeric substrates. After an initial proof of principle using palladium-catalyzed allylic substitution as test reaction, we report now the successful application of this approach to the screening of chiral amines as catalysts for the enantioselective Michael addition to α,β-unsaturated aldehydes. The results confirm that our method allows fast and reliable evaluation of chiral racemic catalysts. This opens up new possibilities for investigating catalyst structures that are not easily available in enantiomerically pure form.

Full text of DOI:10.1002/adsc.201500224

FULL PAPERS
DOI: 10.1002/adsc.201500224
Very Important Publication
Mass Spectrometric Screening of Racemic Amine Catalysts for
Enantioselective Michael Additions
Florian Bächle,a,b Ivana Fleischer,a,c and Andreas Pfaltza,*
a
Department of Chemistry, University of Basel, St. Johanns-Ring 19, 4056 Basel, Switzerland
Phone: +41-61 267-1108, Fax: +41-61 267-1103
E-mail: andreas.pfaltz@unibas.ch
Current address: Dottikon Exclusive Synthesis AG, P.O. Box, 5605 Dottikon, Switzerland
Current address: Institute of Organic Chemistry, University of Regensburg, Universitätsstrasse 31, 93040 Regensburg,
b
c
Germany
Received: March 4, 2015; Revised: March 25, 2015; Published online: May 5, 2015
This paper is dedicated to Stephen Buchwald, a good friend and eminent scientist, on the occasion of his 60th
birthday.
Supporting information for this article is available on the WWW under http://dx.doi.org/10.1002/adsc.201500224.
Abstract: In extension of a concept of Lloyd-Jones, tioselective Michael addition to a,b-unsaturated al-
based on the combination of a racemic catalyst with dehydes. The results confirm that our method allows
a scalemic substrate, we have recently developed fast and reliable evaluation of chiral racemic cata-
a method for determining the enantioselectivity of lysts. This opens up new possibilities for investigating
a chiral catalyst from its racemic form by mass spec- catalyst structures that are not easily available in
trometric screening of a non-equal mixture of two enantiomerically pure form.
mass-labeled quasi-enantiomeric substrates. After an
initial proof of principle using palladium-catalyzed
allylic substitution as test reaction, we report now Keywords: enantioselectivity; mass spectrometry;
the successful application of this approach to the Michael addition; organocatalysis; racemic catalyst
screening of chiral amines as catalysts for the enan- screening
Introduction
We have recently devised a mass spectrometric
method of this kind and, as a proof of principle, dem-
During the last two decades, organocatalysis has onstrated its viability by screening racemic palladium
evolved into one of the major fields of asymmetric complexes as catalysts for enantioselective allylic sub-
synthesis.[1] A major class of organocatalysts are pri- stitution.[3] Our method is based on a concept origi-
mary or secondary amines derived from the chiral nally developed by Lloyd-Jones and coworkers,[4] who
pool, typically from naturally occurring amino acids.[2] showed that the enantiodiscrimination ability of
Catalysts of this type are readily available from inex- a chiral catalyst can be deduced from kinetic resolu-
pensive precursors but their structural diversity is lim- tion experiments with the racemic form.[5,6] In
ited. For the exploration of new applications and a proof-of-concept study they evaluated a series of
novel catalysts, it would be desirable to extend the racemic Pd catalysts in the kinetic resolution of a sca-
range of possible catalyst candidates to amines that lemic (enantioenriched but not enantiopure) allylic
are not accessible from the chiral pool. However, the acetate (Figure 1). Kinetic analysis predicts that if the
preparation of enantiopure chiral amines by asym- reaction shows zero-order rate dependence on sub-
metric synthesis or resolution is often time-consuming strate concentration (saturation conditions), the enan-
and, therefore, may not be worth the effort, consider- tiomeric excess (ee) of the scalemic substrate increas-
ing the high uncertainty in predicting the performance es with conversion. In the extreme case of a perfectly
of a new catalyst. Hence, a method for determining enantioselective catalyst (selectivity factor s=1)
the enantioselectivity of a catalyst by screening its each catalyst enantiomer only reacts with one sub-
racemic form would strongly enhance the range of strate enantiomer at the same rate until all of the
possible structures that can be explored.
minor enantiomer is converted to the product, so the
Adv. Synth. Catal. 2015, 357, 2247 – 2254
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2247

FULL PAPERS
Florian Bächle et al.
Figure 1. Evaluation of a racemic catalyst based on the con-
cept of Lloyd-Jones.[4] The diagram shows the theoretical
curves for the increase of ee with conversion calculated for
different selectivity factors (s) starting from scalemic acetate
with 60% ee.
ee of the unreacted acetate eventually reaches 100%.
An unselective catalyst (s=1), on the other hand,
does not distinguish between the substrate enantio-
mers and, consequently, the ee of the allylic acetate
remains constant throughout the reaction. Compari-
son of the experimentally observed correlation be-
tween the ee and conversion with theoretical charts,
calculated for different selectivity factors, enabled es-
timation of the enantioselectivity of the correspond-
ing enantiopure catalyst.
Scheme 1. ESI-MS screening of racemic Pd-catalysts.
The relative selectivity order between different cat-
alysts could be clearly established in this way. Howev-
er, as the experimental data deviated from the calcu- lated from the ratio of the two allyl intermediates
lated curves, expected for clean pseudo zero-order ki- (Iminor/Imajor), which can be determined from the inten-
netics, a quantitative analysis was not possible. Re- sities of the ESI-MS signals of the major isotopomers,
striction to kinetic resolutions and the labor-intensive using equation (1).[3]
data collection over the whole reaction course further
limit the practicality of the method.
Under ideal conditions the calculated selectivity
factor should be identical to the enantiomeric ratio
By combining the principle of the Lloyd-Jones obtained with the corresponding enantiopure catalyst
method with a mass spectrometric back-reaction in the forward reaction with acetylacetone as nucleo-
screening protocol developed in our lab,[7] we were phile. However, the calculated s values from racemic
able to overcome these limitations. As a first applica- catalyst screening were found to be systematically
tion of our method, we studied the Pd-catalyzed allyl- lower than the values determined for the enantiopure
ic substitution with acetylacetone as nucleophile catalysts, due to deviation from pseudo zero-order ki-
(Scheme 1).[3]
netics. Surprisingly, the ee values from racemate
When a 75:25 mixture of two quasi-enantiomeric screening showed a very good linear correlation with
mass-labeled allylation products[8] was treated with enantioselectivities of the corresponding enantiopure
a racemic catalyst under the conditions previously de- catalysts (Figure 2). Based on this correlation it
veloped for back-reaction screening,[7a] the corre- became possible to predict the enantioselectivity of
sponding allyl intermediates were detected by electro- a new catalyst with high accuracy from screening its
spray ionization mass spectrometry (ESI-MS). As ex- racemic form.
plained above, a perfectly selective catalyst is expect-
In order to evaluate the scope of this approach and
ed to produce the two allyl intermediates in equal to examine whether the linear correlation shown in
amounts under saturation conditions, whereas an un- Figure 2 is a general phenomenon, we decided to in-
selective catalyst does not alter the initial ratio of vestigate other reactions. Herein, we present the suc-
75:25. Thus, with increasing selectivity the ratio cessful extension of our method to the screening of
should decrease from 75:25 to 50:50. Assuming clean racemic amines as catalysts for organocatalyzed enan-
zero-order rate dependence on the substrate concen- tioselective Michael additions to a,b-unsaturated al-
trations, the selectivity factor s can be directly calcu- dehydes.
2248
asc.wiley-vch.de
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2015, 357, 2247 – 2254

FULL PAPERS
Mass Spectrometric Screening
Figure 2. Correlation of the ee values from racemic catalyst
screening with the enantioselectivities determined for the
corresponding enantiopure catalysts.
Principle of the screening protocol for
racemic Michael addition catalysts
We have recently developed an efficient protocol for
the evaluation of chiral amines as catalysts for enan-
tioselective Michael additions of malonates to a,b-un-
saturated aldehydes,[9] based on our ESI-MS back-re-
action screening approach (Figure 3).[7d] Initial studies
showed that the reaction is reversible under catalytic
conditions. Starting from a 1:1 mixture of mass-la-
beled quasi-enantiomeric Michael adducts 2 and 2’,
the signals of the corresponding iminium salts (1-Im
Figure 3. a) Simplified mechanism of the amine-catalyzed
Michael addition. b) ESI-MS screening with enantiopure
catalysts.
and 1-Imꢀ) derived from the unsaturated aldehydes various chiral catalysts and catalyst mixtures con-
were clearly visible in the ESI mass spectrum, despite firmed this expectation.[7d] The results were in excel-
the fact that the back-reaction is endergonic implying lent agreement with the enantioselectivities deter-
that the retro-Michael products must be formed in mined for the corresponding preparative reactions,
very low concentration (Figure 3b). Under the condi- demonstrating that the enantioselectivity can be relia-
tion that the enantioselectivity-determining step in bly determined in this way.
À
the forward and reverse reaction is C C bond forma-
Based on this protocol we investigated the analo-
tion and cleavage, respectively, the selectivity ka/kb in gous reaction of racemic catalysts starting from a sca-
the retro-Michael reaction should be identical to the lemic 75:25 mixture of the same quasi-enantiomers 2
enantioselectivity of the forward reaction according and 2ꢀ (Figure 4). As explained above, a perfectly
to the principle of microscopic reversibility. This con- enantioselective catalyst is expected to produce
dition is met if the Michael adducts and the catalyst a 50:50 ratio of the quasi-enantiomeric retro-Michael
are in rapid equilibrium with the corresponding imini- products Im and Imꢀ under saturation conditions. For
um and enamine intermediates (Curtin-Hammet con- less selective catalysts the ratio will increase, reaching
ditions). In this case, the ratio 1-Im/1-Imꢀ monitored 75:25 for a completely unselective catalyst. Assuming
in the back-reaction and the enantiomeric ratio result- an ideal zero-order rate dependence on the concen-
ing from the preparative Michael addition should be trations of 2 and 2’, the enantioselectivity of the cor-
the same. The results from back-reaction screening of responding enantiopure catalysts can be calculated
Adv. Synth. Catal. 2015, 357, 2247 – 2254
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
2249

FULL PAPERS
Florian Bächle et al.
Figure 4. Principle of ESI-MS screening of racemic catalysts.
through equation 1 from the Im/Imꢀ ratio determined
by ESI-MS.
Figure 5. ESI-MS back reaction screening with 1 mol% of
racemic catalyst 1a. Below, typical ESI-MS spectra obtained
under screening conditions.
Results and Discussion
In the back-reaction screening of enantiopure cata-
lysts for the Michael addition of benzyl malonate to
cinnamaldehyde the TBS-protected prolinol 1a[10]
(TBS=tBuMe2Si) was identified as a highly selective
catalyst affording the product with an enantiomeric
ratio of 97:3. Since amine 1a also showed high activity
and produced well detectable iminium intermediates
in the back reaction, we chose this catalyst for our ini-
tial experiments. According to eq. 1 an enantiomeric
ratio of 97:3 (s=32) corresponds to a Im/Imꢀ ratio of
54:46 in the racemate screening with a 75:25 mixture
(Q=3) of quasi-enantiomeric Michael products. How-
ever, under the standard protocol for ESI-MS back
reaction screening with 10 mol% of catalyst we mea-
sured a ratio of 74:26, very close to the value expect-
ed for a completely unselective catalyst. We attributed
this discrepancy to a deviation from pseudo zero-
order kinetics at this relatively high catalyst loading,
Figure 6. Imꢀ/Im ratios monitored by ESI-MS. In brackets:
in line with observations made in racemic catalyst enantioselectivity of catalysts 1a-d calculated from eq. (1).
screening for allylic substitution (Scheme 1 and
Figure 2), which showed that the discrepancy between
the ee values from racemic and enantiopure catalyst
Although the ee value determined from racemic
screening increased with higher catalyst loading. catalyst screening was still substantially lower than
Indeed, reduction of the catalyst loading to 1 mol% the actual enantioselectivity of the enantiopure cata-
brought the ratio to 65:35 corresponding to 71% ee lyst (71 vs. 93% ee), we decided to go ahead with test-
(Figure 5). Such a low catalyst loading is challenging ing additional catalysts. In view of the results ob-
because of the low intensity of the ESI-MS signals de- tained for allylic substitution (Figure 2), we hoped
rived from catalytic intermediates. However, modified that the racemic and enantiopure catalyst data would
sample preparation, involving dilution with acetoni- again show a linear correlation that could be used as
trile, resulted in clean spectra with a very good signal a correction function. For a proof of principle, we
to noise ratio and strong signals of the iminium inter- chose four known silyl- and methyl-protected prolinol
mediates (1a-Im/1a-Imꢀ) (Figure 5).
derivatives 1a–d (Figure 6).[10]
2250
asc.wiley-vch.de
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2015, 357, 2247 – 2254

FULL PAPERS
Mass Spectrometric Screening
Figure 7. Calculated enantioselectivity from racemic catalyst
screening (light grey bars) vs. the enantioselectivity deter-
mined for the corresponding enantiopure catalyst (dark grey
bars).
Figure 8. Linear correlation between the results from race-
mate and enantiopure catalyst screening.
The racemic catalysts were synthesized from N-
Boc-protected pyrrolidine by lithiation and subse-
quent reaction with benzophenone followed by O-
protection (see supporting information). The enantio-
pure catalysts are commercially available or easily ac-
cessible starting from enantiopure diphenyl prolinol
or proline using established procedures.[10,11] The
enantioselectivities are known to increase with grow-
ing steric demand of the protecting group,[7d,12] so the
four catalysts were expected to provide suitable data
points in the moderate to high ee range.
Racemate screening was conducted under the con-
ditions described in Figure 5. The detected ratios of
the iminium intermediates Im and Imꢀ derived from
the four catalysts are depicted in Figure 6 with the
corresponding enantioselectivities calculated from
equation 1.[13] According to the signal ratios the enan-
Figure 9. Dependence of the calculated theoretical enantio-
selectivity on the Imꢀ/Im ratio in racemate screening (with
Q=1/3).
tioselectivity should increase in the sequence 1b <1c graph showing the calculated ee as a function of the
<1a <1d. For comparison back-reaction screening of Im/Imꢀ ratio (Figure 9). Due to the strong curvature
the corresponding enantiopure catalysts was carried merging into an almost vertical slope between 20%
out, using a 1:1 mixture of the mass-labeled substrates and 0% ee, very small measuring errors in the ratios
2 and 2ꢀ under otherwise identical conditions (catalyst 2/2ꢁ and Im/Imꢀ result in very large errors of the ee
loading, reaction time, ESI-MS parameters). As calculated according to the linear correlation equation
shown in Figure 7 the same selectivity order was 2 (a change of the Im/Imꢀ ratio from 75:25 to 74:26
found for the enantiopure and the racemic catalysts corresponds to an increase in the calculated ee from
confirming the validity of the racemate screening pro- 0% to 20%). However, the observed inadequate cor-
tocol for identifying the most selective catalysts relation in the low ee region does not impair the utili-
among a set of racemic catalyst candidates.
ty of the screening method, which primarily serves to
As observed before in our allylic alkylation studies, identify the most selective members of a catalyst li-
the enantioselectivities derived from racemate screen- brary. The exact values in the low ee region are not
ing data showed a good linear correlation with the relevant, as long as they allow clear distinction of in-
enantioselectivities determined from the enantiopure ferior catalysts that are not worth to be further inves-
catalysts (Figure 8). While equation 2 displays an ex- tigated.
cellent fit to the four data points (R2 =0.99), it is only
To examine the scope of our screening protocol we
valid for the moderate to high ee region (enantiopure synthesized a series of new organocatalysts based on
catalysts inducing >50–60% ee). For less selective an isoindoline framework. As these compounds
catalysts this linear relationship breaks down as can cannot be synthesized from readily available precur-
be seen from the y-intercept at 52% ee which in reali- sors from the chiral pool, they are difficult to obtain
ty should be at 0% ee. This deviation from ideal be- in enantiopure fashion. Compared to pyrrolidine-de-
havior is not surprising, considering the course of the rived catalysts, the annulated phenyl ring of the isoin-
Adv. Synth. Catal. 2015, 357, 2247 – 2254
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
2251

FULL PAPERS
Florian Bächle et al.
the two enantiopure amines (1R,2S)- and (1S,2S)-(7)
from N-Boc-(S)-prolinol by oxidation to the aldehyde
and subsequent Grignard reaction with phenylmagne-
sium bromide. Samples of enantiopure catalysts 6a–
c were obtained from the racemates by semi-prepara-
tive HPLC on a chiral stationary phase.
The racemic and enantiopure forms of the six new
organocatalysts were then subjected to ESI-MS back-
reaction screening using the standard protocol de-
scribed above (Table 1). As a further structural var-
iant, the commercially available racemic 2-phenylpyr-
rolidine 8 and an enantiopure sample obtained by
HPLC resolution on a chiral phase were also included
in this study. For both the enantiopure and racemic
catalysts, the ee values listed in the table were aver-
Scheme 2. Synthesis of racemic catalysts 6a–c. Conditions:
a) N’-tert-butyl-N,N-dimethylformamidine, toluene, reflux.
b) For 5a: sec-BuLi, THF, À788C; then benzophenone,
À788C to r.t. For 5b: sec-BuLi, THF, À788C; then benzalde-
hyde, À788C to r.t. c) (1) LiAlH4, THF, reflux. (2) For 6a:
TMSOTf, NEt3, CH2Cl2, 08C to r.t. For 6b: TBSOTf, NEt3,
CH2Cl2, 08C to r.t. For 6c: TBSOTf, NEt3, CH2Cl2, À788C
to r.t.
Table 1. ESI-MS screening results of the new catalysts.[a]
entry catalyst
racemate corrected enantiopure
sceening value
ee [%]
catalyst
ee [%]
ee [%][b]
Scheme 3. Synthesis of racemic catalysts 7. Conditions: a)
sec-BuLi, TMEDA, Et2O, À788C; then benzaldehyde,
À788C to r.t., separation of diastereoisomers by FC. b) (1)
AcOH/3m HCl, r.t. (2) TBSOTf, NEt3, CH2Cl2, 08C to r.t.
1
2
47
78
85
79
87
doline derivatives alters the geometry of the nitrogen
containing heterocycle as well as the conformation of
the substituent at the stereogenic center through
steric interactions. In order to see how these effects of
the extra phenyl ring influence the enantioselectivity,
we prepared isoindoline derivatives 6a and 6b, which
are analogues of the previously evaluated catalysts 1c
and 1a bearing TMS and TBS ether substituents. In
addition, we synthesized both diastereomers of isoin-
doline 6c (Scheme 2) and pyrrolidine 7 containing an
additional stereogenic center (Scheme 3).[14]
The isoindoline derivatives were synthesized start-
ing from N-protected isoindoline by lithiation and
subsequent trapping with benzophenone or benzalde-
hyde, respectively. Initial experiments with tert-butyl-
carbamate as directing group failed as they led to de-
composition. Eventually the desired products were
obtained using a formamidine protecting group as re-
ported by Beeley and Rockell (Scheme 2).[15] The re-
action with benzaldehyde led to a mixture of the dia-
stereomeric isoindoline derivatives (S*,R*)- and
(S*,S*)-6c, which was separated by flash column chro-
matography. The relative configuration of the two iso-
mers was established by X-ray analysis of the HCl
salt of the (S*,S*)-isomer (see supporting informa-
tion).
58
44
3
4
77
78[c]
23
33
65
71
66
70
5
6
[d]
0
0
–
–
49
2
[d]
7
[a]
For screening conditions see Figure 5.
Values corrected using the linear regression depicted in
Figure 8 (y=0.56x+52.43).
2 mol% of catalyst used.
No correction possible due to low catalyst selectivities.
[b]
[c]
[d]
In order to correlate the results from racemic and
enantiopure catalyst screening, we also synthesized
2252
asc.wiley-vch.de
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2015, 357, 2247 – 2254

FULL PAPERS
Mass Spectrometric Screening
The additional stereogenic center in catalysts (R*,S*)-
7/(S*,S*)-7 and (S*,R*)-6c/(S*,S*)-6c causes a clear
match/mismatch effect. The additional annulated
phenyl ring in catalysts 6a and 6b has a negative
effect on the ee (83 vs. 79% for 1c/6a; 93 vs. 87% for
1a/6b). For the monophenyl-substituted analogues,
on the other hand, the effect is positive (Table 1, en-
tries 3 vs. 5 and 4 vs. 6), which can be explained by
steric repulsion between the substituent at the stereo-
genic center and the adjacent aromatic H atom of the
annulated phenyl ring. This interaction forces the
phenyl and silyloxy groups to orient themselves
toward the active site of the catalyst, resulting in
greater steric hindrance in this region compared to
the conformationally more flexible pyrrolidine ana-
logues. In summary, among the various structural var-
iants evaluated in this study, the proline-derived pyr-
Figure 10. Linear correlation for all catalysts with an enan-
tioselectivity>60%.
aged from four independent measurements (see sup- rolidine derivatives 1a and 1c, of the structural type
porting information).
originally introduced by Hayashi and Jørgensen,[10]
For all catalysts investigated, the back reaction remain the catalysts of choice for the investigated
took place and the intermediates Im and Imꢀ were enantioselective Michael addition.
clearly detected by ESI-MS. Using equation 1, the
Im/Imꢀ ratios obtained from the racemic catalysts
were converted to the theoretical ee values expected Conclusions
under ideal zero-order kinetics. These values were
then corrected using the linear correlation function In an initial proof of principle study using palladium-
shown in Figure 8. Gratifyingly, the results confirmed catalyzed allylic substitution as test reaction,[3] we had
the validity of this protocol. For the moderately to demonstrated that our mass spectrometric screening
highly selective catalysts (ee> 60%; entries 1–5) the method combined with the scalemic substrate ap-
corrected ee values from racemic catalyst screening proach of Lloyd-Jones[4] enabled determination of the
closely matched the data obtained from the corre- enantioselectivity of a chiral catalyst from its racemic
sponding enantiopure catalysts. These results con- form. We now wanted to explore other applications in
firmed the validity of the linear correlation function order to evaluate the scope of our racemic catalyst
established in our initial experiments with four cata- screening protocol. The results presented here for the
lysts (Figure 7). Linear regression analysis including enantioselective organocatalyzed Michael addition
all catalysts in the ee range of >60% again showed confirm that our method allows fast and reliable eval-
an excellent fit with an R2 value of 0.99. Apparently, uation of chiral racemic catalysts. As observed in the
the enantioselectivity of a new moderately to highly allylic substitution, the ee values from racemic catalyst
enantioselective catalyst can be reliably determined screening were smaller than those determined for the
with remarkable accuracy using this protocol. The enantiopure catalysts, but could be corrected by
data in Table 1 also allowed clear distinction of the means of a linear correlation function. The corrected
less selective catalysts (S*,S*)-7 and 8 (entries 6 and ee values were in excellent agreement with the actual
7) from the group of potentially useful catalysts (en- enantioselectivities of the respective enantiopure cat-
tries 1–5).
alysts. Because racemic compounds that are not de-
From the screening results the following conclu- rived from the chiral pool are in general more readily
sions can be drawn regarding the influence of the sub- accessible than the enantiopure forms, our screening
stituent at the stereogenic center and the annulated protocol considerably extends the possibilities for
phenyl ring. Consistent with our previous study,[7d] the rapid evaluation of new catalyst structures.
enantioselectivity increases with the steric size of the
oxygen protecting group (TIPS 1d> TBS 1a> TMS
1c> Me 1b; TBS 6b> TMS 6a; Figure 10). The two
Experimental Section
diastereomeric pyrrolidine derivatives (S*,S*)-7 and
(R*,S*)-7) with only one phenyl substituent are con-
siderably less selective than the diphenyl-substituted
analogue (1d). Apparently, the steric hindrance exert-
ed by the two phenyl groups and the silyloxy substitu-
General Remarks
ESI-MS spectra were measured on a Varian 1200 L Quadru-
pol MS/MS spectrometer using mild desolvation conditions
ent is essential for efficient enantiodiscrimination. (39 psi nebulizing gas, 4.9 kV spray voltage, 19 psi drying
Adv. Synth. Catal. 2015, 357, 2247 – 2254
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
2253

FULL PAPERS
Florian Bächle et al.
gas at 2008C, 50 V capillary voltage). The samples were di-
luted immediately with CH3CN prior to their analysis and
measured using direct injection. The spectra were acquired
in the centroid mode. Every spectrum consisted of at least
30 scans and the selectivity was calculated from the ratios of
the peak heights of the major isotopomers of Im and Imꢀ.
H. B. Kagan, J. Am. Chem. Soc. 2003, 125, 7490–7491;
b) L. A. Evans, N. S. Hodnett, G. C. Lloyd-Jones,
Angew. Chem. 2012, 124, 1556–1564; Angew. Chem.
Int. Ed. 2012, 51, 1526–1533; c) K. D. Collins, T.
Gensch, F. Glorius, Nature Chem. 2014, 6, 859–871.
[6] For selective poisoning of one enantiomer of a racemic
catalyst, see: a) J. W. Faller, J. Parr, J. Am. Chem. Soc.
1993, 115, 804–805; b) J. W. Faller, A. R. Lavoie, J.
Parr, Chem. Rev. 2003, 103, 3345–3368.
[7] a) C. A. Müller, A. Pfaltz, Angew. Chem. 2008, 120,
3411–3414; Angew. Chem. Int. Ed. 2008, 47, 3363–3366;
b) A. Teichert, A. Pfaltz, Angew. Chem. 2008, 120,
3408–3410; Angew. Chem. Int. Ed. 2008, 47, 3360–3362;
c) C. A. Müller, C. Markert, A. M. Teichert, A. Pfaltz,
Chem. Commun. 2009, 1607–1618; d) I. Fleischer, A.
Pfaltz, Chem. Eur. J. 2010, 16, 95–99.
[8] The term ꢂquasi-enantiomericꢁ means that the com-
pounds behave like enantiomers but have different mo-
lecular masses due to mass labels introduced at posi-
tions where they do not alter the reactivity of the mole-
cule. For a review on quasi-enantiomers, see: Q.
Zhang, D. P. Curran, Chem. Eur. J. 2005, 11, 4866–4880.
[9] a) M. Yamaguchi, N. Yokota, T. Minami, J. Chem. Soc.
Chem. Commun. 1991, 1088–1089; b) S. Brandau, A.
Landa, J. Franzen, M. Marigo, K. A. Jørgensen, Angew.
Chem. 2006, 118, 4411–4415; Angew. Chem. Int. Ed.
2006, 45, 4305–4309; c) C. Palomo, A. Landa, A.
Mielgo, M. Oiarbide, A. Puente, S. Vera, Angew.
Chem. 2007, 119, 8583–8587; Angew. Chem. Int. Ed.
2007, 46, 8431–8435; d) Y. Wang, P. Li, X. Liang, J. Ye,
Adv. Synth. Catal. 2008, 350, 1383–1389; e) A. Ma, S.
Zhu, D. Ma, Tetrahedron Lett. 2008, 49, 3075–3077;
f) O. V. Maltsev, A. S. Kucherenko, S. G. Zlotin, Eur. J.
Org. Chem. 2009, 5134–5137; g) E. Alza, S. Sayalero, P.
Kasaplar, D. Almas¸i, M. A. Pericàs, Chem. Eur. J. 2011,
17, 11585–11595; h) S. K. Ghosh, K. Dhungana, A. D.
Headley, B. Ni, Org. Biomol. Chem. 2012, 10, 8322–
8325; i) K. S. Feu, A. M. Deobald, S. Narayanaperumal,
A. G. CorrÞa, M. Weber Paix¼o, Eur. J. Org. Chem.
2013, 5917–5922; j) K. Akagawa, N. Sakai, K. Kudo,
Angew. Chem. 2015, 127, 1842–1846; Angew. Chem.
Int. Ed. 2015, 54, 1822–1826.
General procedure for the ESI-MS screening of
racemic organocatalysts
A GC-vial was charged with a scalemic 3:1 mixture of (R)-2
(4.30 mg 9.38 mmol, 0.75 equiv) and (S)-2ꢀ (1.39 mg, 3.13
mmol, 0.25 equiv). The mixture was dissolved in EtOH/
CH2Cl2 (85 mL/10 mL) and a solution of the corresponding
racemic organocatalyst in EtOH (5 mL, 0.025m, 1 mol%)
was added. The mixture was stirred for 15 min at room tem-
perature. The reaction mixture was diluted with CH3CN
(1 mL) and subjected to ESI-MS analysis.
General procedure for the ESI-MS screening of
enantiopure organocatalysts
A GC-vial was charged with an equimolar mixture of (R)-2
(2.87 mg 6.25 mmol, 0.50 equiv) and (S)-2ꢀ (2.78 mg, 6.25
mmol, 0.50 equiv). The mixture was dissolved in EtOH/
CH2Cl2 (85 mL/10 mL) and a solution of the corresponding
enantiopure organocatalyst in EtOH (5 mL, 0.025m,
1 mol%) was added. The mixture was stirred for 15 min at
room temperature. The reaction mixture was diluted with
CH3CN (1 mL) and subjected to ESI-MS analysis.
Acknowledgements
Financial support of this work by the Swiss National Science
Foundation is gratefully acknowledged. We thank Dr.
Markus Neuburger (Laboratory for Chemical Crystallogra-
phy, University of Basel) for the crystal structure analysis.
References
[10] a) M. Marigo, T. C. Wabnitz, D. Fielenbach, K. A. Jør-
gensen, Angew. Chem. 2005, 117, 804–807; Angew.
Chem. Int. Ed. 2005, 44, 794–797; b) Y. Hayashi, H.
Gotoh, T. Hayashi, M. Shoji, Angew. Chem. 2005, 117,
4284–4287; Angew. Chem. Int. Ed. 2005, 44, 4212–4215.
[11] a) F.-M. Gautier, S. Jones, S. J. Martin, Org. Biomol.
Chem. 2009, 7, 229–231; b) D. Enders, H. Kipphardt, P.
Gerdes, L. BreÇa-Valle, V. Bhushan, Bull. Soc. Chim.
Belg. 1988, 97, 691.
[1] a) D. W. C. MacMillan, Nature 2008, 455, 304–308;
b) B. List, Angew. Chem. 2010, 122, 1774–1779; Angew.
Chem. Int. Ed. 2010, 49, 1730–1734; c) Science of Syn-
thesis: Asymmetric Organocatalysis, Vol. 1 (Ed.: B.
List) and Vol. 2 (Ed. K. Maruoka), Thieme, Stuttgart,
2012; d) Comprehensive Enantioselective Organocataly-
sis, Vols. 1–3 (Ed.: P. I. Dalko), Wiley-VCH, Weinheim,
2013.
[2] a) A. Erkkilä, I. Majander, P. M. Pihko, Chem. Rev.
2007, 107, 5416–5470; b) S. Mukherjee, J. W. Jang, S.
Hoffmann, B. List, Chem. Rev. 2007, 107, 5471–5569;
c) B. M. Paz, H. Jiang, K. A. Jørgensen, Chem. Eur. J.
2015, 21, 1846–1853.
ˇ
[12] D. Seebach, U. Groselj, D. M. Badine, W. B. Schweizer,
A. K. Beck, Helv. Chim. Acta 2008, 91, 1999–2034.
[13] All depicted results are average values out of four in-
dependent measurements (see supporting information).
[14] The relative configuration of catalysts 7 was deter-
mined by 1H NMR comparison of the Boc-protected
amino alcohol intermediates with literature data: J. L.
Bilke, S. P. Moore, P. OꢁBrien, J. Gilday, Org. Lett.
2009, 11, 1935–1938.
[3] C. Ebner, C. A. Müller, C. Markert, A. Pfaltz, J. Am.
Chem. Soc. 2011, 133, 4710–4713.
[4] B. Dominguez, N. S. Hodnett, G. C. Lloyd-Jones,
Angew. Chem. 2001, 113, 4419–4421; Angew. Chem. Int.
Ed. 2001, 40, 4289–4291.
[5] For alternative approaches to racemic catalyst screen-
ing, see: a) F. Lagasse, M. Tsukamoto, C. J. Welch,
[15] L. J. Beeley, C. J. M. Rockell, Tetrahedron Lett. 1990,
31, 417–420.
2254
asc.wiley-vch.de
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2015, 357, 2247 – 2254