Asymmetric Morita–Baylis–Hillman Reaction: Catalyst Development and Mechanistic Insights Based on Mass Spectrometric Back-Reaction Screening

Article abstract of DOI:10.1002/chem.201604616

An efficient protocol for the evaluation of catalysts for the asymmetric Morita–Baylis–Hillman (MBH) reaction was developed. By mass spectrometric back-reaction screening of quasi-enantiomeric MBH products, an efficient bifunctional phosphine catalyst was identified that outperforms literature-known catalysts in the MBH reaction of methyl acrylate with aldehydes. The close match between the selectivities measured for the forward and back reaction and kinetic measurements provided strong evidence that the aldol step and not the subsequent proton transfer is rate- and enantioselectivity-determining.

Full text of DOI:10.1002/chem.201604616

DOI: 10.1002/chem.201604616
Communication
&
Asymmetric Catalysis
Asymmetric Morita–Baylis–Hillman Reaction: Catalyst
Development and Mechanistic Insights Based on Mass
Spectrometric Back-Reaction Screening
Patrick G. Isenegger, Florian Bꢀchle, and Andreas Pfaltz*[a]
Abstract: An efficient protocol for the evaluation of cata-
lysts for the asymmetric Morita–Baylis–Hillman (MBH) reac-
tion was developed. By mass spectrometric back-reaction
screening of quasi-enantiomeric MBH products, an effi-
cient bifunctional phosphine catalyst was identified that
outperforms literature-known catalysts in the MBH reac-
tion of methyl acrylate with aldehydes. The close match
between the selectivities measured for the forward and
back reaction and kinetic measurements provided strong
evidence that the aldol step and not the subsequent
proton transfer is rate- and enantioselectivity-determining.
The Morita–Baylis–Hillman (MBH) reaction is a valuable syn-
thetic method,[1] as the products are densely functionalized
and can be easily modified in various ways.[2] In the last
decade substantial progress has been made in the develop-
ment of enantioselective variants.[3] However, although many
chiral catalysts have been reported, their scope is generally
limited. Especially for MBH reactions of simple acrylic esters
with aldehydes, more efficient catalysts with broader applica-
tion range are needed. Here we report a combinatorial ap-
proach to the development of chiral MBH catalysts based on
a mass spectrometric screening method devised in our labora-
tory,[4] which has led to improved bifunctional chiral phosphine
catalysts for MBH reactions of methyl acrylate with aldehydes.
In addition, mass spectrometric data together with kinetic
studies allowed us to identify the rate- and enantioselectivity-
Scheme 1. Principle of back-reaction screening.
determining step.
Our screening method works as follows (see Scheme 1):
Starting from a 1:1 mixture of mass-labeled quasi-enantiomeric
MBH products 1a and 1b, a catalyst-mediated back reaction
via intermediates 2a and 2b is induced, which then undergo
cleavage to the aromatic aldehyde (in the present study 4-ni-
trobenzaldehyde) and the mass-labeled catalyst-substrate ad-
ducts 4a and 4b. If we assume a fast equilibrium between 1a
and 1b and the corresponding catalyst adducts 2a and 2b,
followed by a slow rate-determining CÀC bond cleavage
(Curtin–Hammet conditions), the ratio 4a/4b will be identical
to the enantiomeric ratio produced in the MBH reaction in the
forward direction, according to the principle of microscopic re-
versibility. The ratio 4a/4b can be reliably determined even at
very low concentration by electrospray ionization mass spec-
trometry (ESI-MS). Under the conditions of ESI-MS analysis, 4a
and 4b are protonated and their ratio is determined from the
signal ratio of the corresponding cationic species 6a and 6b.
This screening protocol is fast and operationally simple, as it
does not require any work-up or purification steps. Moreover,
mixtures of catalysts having different molecular masses can be
screened simultaneously. By multicatalyst screening of combi-
natorial catalyst libraries, catalyst discovery and optimization
can be considerably accelerated.
[a] P. G. Isenegger, Dr. F. Bꢀchle, Prof. Dr. A. Pfaltz
Department of Chemistry, University of Basel
St. Johanns-Ring 19, 4056 Basel (Switzerland)
E-mail: andreas.pfaltz@unibas.ch
Supporting information and ORCID(s) from the author(s) for this article are
available on the WWW under http://dx.doi.org/10.1002/chem.201604616.
Chem. Eur. J. 2016, 22, 1 – 6
1
ꢁ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Communication
Initial studies showed that, upon treatment of MBH adducts
derived from acrylic esters and 4-nitrobenzaldehyde with tri-
phenylphosphine, signals of catalyst adducts resulting from
CÀC bond cleavage could be detected by ESI-MS. Next we had
to find a suitable pair of quasi-enantiomeric MBH products.
Alkyl esters differing in the number of C atoms, which we eval-
uated first, gave unsatisfactory results as they slightly differed
in reactivity, resulting in unequal ratios of catalyst-substrate ad-
ducts with an achiral catalyst. Finally, methyl and trideutero-
methyl esters 1a and 1b proved to be optimal as they pos-
sessed identical reactivity and gave rise to easily detectable
signal pairs.
ately injected into the spectrometer. The signals of the posi-
tively charged retro-MBH products 6a and 6b as well as the
cationic species 5a and 5b generated by protonation of inter-
mediates 2a/3a and 2b/3b were all clearly visible in high in-
tensity. To validate the screening results, the ratios 6a/6b mea-
sured by ESI-MS were compared to the enantiomeric ratios
(e.r.) determined for the forward reaction by HPLC on a chiral
stationary phase (Table 1). To our delight, the results from
Table 1. Screening of bifunctional organocatalysts.
Successful application of back-reaction screening requires
the CÀC bond-forming step to be rate- and enantioselectivity-
determining (see above). However, in several studies the
proton transfer after the aldol step was found to be, at least in
part, rate-determining.[6–8] We therefore conducted preliminary
experiments with chiral catalyst 9a (Figure 1). We were pleased
Entry Catalyst
ESI-MS screening 6a/6b Preparative reaction[a] e.r.
1
2
(R,S)-8
(S,S)-8
77:23
87:13
68:32
74:26
71:29
81:19
85:15
82:18
9:91
76:24
87:13
66:34
75:25
72:28
80:20
85:15
82:18
10:90
6:94
3
(S)-9a
4
5
(S)-9b
(S)-9c
6
7
8
9
(S)-10a
(S)-10b
(S)-10c
(S,R)-11 a
10
11
12
13
14
15
16
17
18
19
20
21
22
23
(R,R)-11 a 7:93
(S,R)-11 b 12:88
(S,R)-11 c
(S,R)-11 d 13:87
(S,R)-11 e 17:83
(S,R)-12a 6:94
(R,R)-12a 5:95
(R,R)-12b 7:93
(S,R)-12c
(S,R)-12d 4:96
(S,R)-12e 5:95
(S,R)-12 f
(R,R)-12 f
(S,R)-12g 20:80
13:87
13:87
13:87
20:80
6:94
11:89
5:95
6:94
6:94
7:93
5:95
5:95
21:79
11:89
21:79
15:85 (11:89)[b]
19:81
[a] Reaction conditions: 18a (1.0 equiv), 7a (1.5 equiv), catalyst
(10 mol%), CH2Cl2, RT, 18 h. Determined by HPLC on a chiral stationary
phase; [b] ratio in parentheses measured after 30 min.
Figure 1. Phosphine-based organocatalysts.
to find that the ratios 6a/6b produced in the back reaction
closely matched the enantiomeric ratios determined for the
preparative forward reaction, implying that our screening pro-
tocol should indeed be applicable. Encouraged by these re-
sults, we started a systematic evaluation of chiral MBH catalysts
based on the screening protocol shown in Scheme 1.
back-reaction screening and from the corresponding prepara-
tive MBH reactions closely matched each other. Only in one
case (entry 22) the e.r. was significantly lower than the selectiv-
ity of the back reaction. However, we found that this diver-
gence was caused by slow catalyst-induced racemization of
the MBH products under the reaction conditions.
The best enantioselectivities in the MBH reaction of simple
acrylic esters with aldehydes so far were achieved by Yixin Lu
and co-workers with threonine-derived bifunctional phos-
phine-thiourea catalysts such as (2R,3S)-8.[3c] Their modular
nature seemed ideal for systematic structural optimization. We
therefore synthesized an array of related phosphine-thiourea
and -squaramide derivatives from commercially available
amino alcohols (Figure 1; for additional catalysts and synthetic
procedures see the Supporting Information).
Initial experiments with analogues of Lu’s catalyst (2R,3S)-8
lacking the silyloxy substituent gave similar or lower enantiose-
lectivities (Table 1, entries 3–5). Replacement of the thiourea
group by a squaramide unit led to improved e.r. values of up
to 85:15 (entries 6–8). However, attempts to further optimize
this class of squaramide derivatives were unsuccessful.[9]
A
more pronounced increase of the e.r. resulted when the
methyl group of catalyst (2R,3S)-8 was replaced by a phenyl
group (cf. entries 1 and 9). Introduction of electron-donating
or electron-withdrawing substituents into the N-phenyl group
of (1S,2R)-11 a had only a marginal effect, whereas a N-cyclo-
For back-reaction screening, an equimolar mixture of quasi-
enantiomers 1a and 1b was reacted with 10 mol% of catalyst
in CH2Cl2 at room temperature. After 30 minutes the reaction
mixture was diluted ten-fold with CH2Cl2 and a sample immedi-
&
&
Chem. Eur. J. 2016, 22, 1 – 6
www.chemeurj.org
2
ꢁ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Communication
hexyl substituent lowered the e.r. from 90:10 to 80:20 (en-
tries 9–14). Notably, the like-diastereomers induced better se-
lectivity than the unlike-diastereomers (entries 1–2 and 9–10).
Finally, a further significant increase in enantioselectivity was
achieved by replacing the P-phenyl groups with more bulky
ortho-tolyl, ortho-iPr-phenyl, or 1-naphthyl groups (entries
15–19). Catalysts 12 f and 12g with 3,5-dialkyl-substituted
P-phenyl groups gave inferior results (entries 21 and 23).
To demonstrate the potential of our screening protocol for
combinatorial catalyst development,[5] we synthesized a library
of six different organocatalysts by a three-step procedure with-
out purification or separation of the intermediates and final
products (Scheme 2). The resulting crude catalyst library was
then directly subjected to back-reaction screening.
Figure 2. Multi-catalyst screening.
Table 2. Substrate scope of catalyst (1R,2R)-12a.[a]
Entry
R
Product
x
t [h] Yield[b] [%] ee[c] [%]
1
2
3
4
5
6
7
8
4-NO2-Ph (18a)
2-NO2-Ph (18b)
4-CN-Ph (18c)
19a
19b
19c
10
10
6
6
98
88
93
98
69
89
90
82 (73)[d]
95
83
42
36
94
80
92
92
90
90
92
90 (90)[d]
90
80
85
30
10 18
10 18
20 48
20 48
20 48
20 72
20 72
20 72
20 96
20 96
3,5-(CF3)2-Ph (18 f) 19d
4-F-Ph (18e)
4-Cl-Ph (18 f)
4-Br-Ph (18g)
Ph (18h)
2-pyridine (18i)
2- furfural (18j)
4-Me-Ph (18k)
C6H11 (18l)
19e
19 f
19g
19h
19i
19j
19k
19l
9
10
11
12
[a] 18 (0.1 mmol), 7a (0.15 mmol) in THF (0.1 mL); [b] after purification by
preparative TLC; [c] determined by HPLC on a chiral stationary phase;
[d] 1 mmol scale.
Scheme 2. One-batch catalyst library synthesis.
showed lower reactivity. Cyclohexane-carbaldehyde gave only
low ee and yield. Overall, catalyst (1R,2R)-12a clearly outper-
formed the best literature-known catalysts such as (2R,3S)-8 in
terms of enantioselectivity and yield.[3]
For all catalysts the corresponding signals of intermediates
6a and 6b were detected (Figure 2). Although the selectivities
obtained by multi-catalyst screening were somewhat lower
compared to single-catalyst screening,[10] the selectivity order
among the six catalysts was correctly displayed. So the most
selective catalysts in a combinatorial library can be readily
identified in this way.
The results from back-reaction screening also provided in-
sights into the catalytic cycle. Mechanistic studies of amine-cat-
alyzed MBH reactions have shown that CÀC bond formation
and the subsequent proton transfer have similar activation bar-
riers and, depending on the conditions, one or the other step
may become primarily rate-limiting.[6,7] For phosphine-based
catalysts only a computational study was reported, which pre-
dicted the proton-transfer step to be rate-determining.[8] How-
ever, Plata and Singleton have recently demonstrated that at
present computational methods are unable to cope with the
complex multi-step catalytic cycle of MBH reactions and, there-
fore, fail to produce reliable results.[7]
From our screening experiments with 30 catalysts, three effi-
cient catalysts (1R,2R)-12a, (1S,2R)-12d, and (1S,2R)-12e have
emerged, which all induce an e.r. of 95:5 in the reaction of
methyl acrylate with 4-nitrobenzaldehyde. For further studies
addressing the substrate scope, (1R,2R)-12a was chosen be-
cause of its higher reactivity (Table 2).
Under optimized conditions,[9] a broad selection of aromatic
and heteroaromatic aldehydes afforded high enantioselectivi-
ties and yields in the preparative reaction (Table 2, entries
1–12), except for 4-methylbenzaldehyde (entry 11), which
The close match between the selectivities obtained from
back-reaction screening and the preparative forward reaction
Chem. Eur. J. 2016, 22, 1 – 6
www.chemeurj.org
3
ꢁ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Communication
group of acrylate 7a (3.73 ppm) and MBH adduct 19a
(3.71 ppm). Data were taken from four reactions, in which the
aldehyde/acrylate ratio was varied from 0.5 to 3.0. The results
clearly showed the reaction to be first order in aldehyde, pro-
viding further evidence that the aldol step is rate-limiting (see
Figure 4 and the Supporting Information).
Figure 4. Kinetic analysis of the MBH reaction of methyl acrylate with 4-ni-
trobenzaldehyde. Double logarithmic plot of the dependence of the reaction
rate on the initial aldehyde concentration. The data points correspond to al-
dehyde concentrations of 41.7, 83.3, 166.7, and 250.0 mmolLÀ1 and a con-
stant acrylate concentration of 83.3 mmolLÀ1. Linear fit for 18a:
y=0.9959xÀ12.175 (R2 =0.99), consistent with first order in 18a.
In conclusion, we have demonstrated the potential of mass
spectrometric back-reaction screening for the evaluation of
catalysts for asymmetric MBH reactions. Screening of 30 bi-
functional phosphines has led to an efficient catalyst for the re-
action of methyl acrylate with aldehydes, showing improved
enantioselectivity and scope compared to previously reported
catalysts. In addition, the results from back-reaction screening
and additional kinetic studies have provided evidence that the
enantioselectivity is determined in the CÀC bond-forming step,
which is turnover-limiting.
Figure 3. Qualitative energy profiles.
indicates that CÀC bond formation is the rate- and enantiose-
lectivity-determining step. A qualitative reaction coordinate
consistent with the selectivity data is shown in Figure 3a. The
consensus between the product e.r. of the forward reaction
and the ratio 4a/4b determined from the ESI-MS signals of 6a
and 6b in the back reaction implies that both ratios depend
on the same transition state. This is the case if the aldol step is
the step with the highest energy barrier. If the proton transfer
was rate-determining (Figure 3b), equilibration between inter-
mediates 3a/3b and 4a/4b would occur, due to the now re-
versible aldol step. Thus, the ratio 4a/4b would depend on
both DDG° of the proton transfer and the equilibrium con-
stants K(3a/4a) and K(3b/4b), which differ because of the un-
equal relative energies of 3a and 3b, whereas the enantiose-
lectivity of the forward reaction would be determined solely
by DDG° of the proton transfer. Accordingly, the ratio 4a/4b
would be expected to deviate from the e.r of the forward reac-
tion, in contradiction to the observed match between the for-
ward and back reaction.[11]
Acknowledgements
We thank Prof. Donna Blackmond (The Scripps Research Insti-
tute) for insightful discussions and advice for the kinetic ex-
periments. Financial support by the Swiss National Science
Foundation is gratefully acknowledged.
Keywords: asymmetric catalysis · high-throughput screening ·
mass spectrometry · Morita–Baylis–Hillman · organocatalysis
In addition, kinetic measurements were carried out to deter-
1
mine the reaction order in aldehyde by in situ H NMR analysis
[1] Y. Wei, M. Shi, Chem. Rev. 2013, 113, 6659–6690.
of reaction process based on the signals of the methyl ester
[2] T.-Y. Liu, M. Xie, Y.-C. Chen, Chem. Soc. Rev. 2012, 41, 4101–4112.
&
&
Chem. Eur. J. 2016, 22, 1 – 6
www.chemeurj.org
4
ꢁ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Communication
[3] a) J.-J. Gong, K. Yuan, X.-Y. Wu, Tetrahedron: Asymmetry 2009, 20, 2117–
2120; b) Y. Wei, M. Shi, Acc. Chem. Res. 2010, 43, 1005–1018; c) X. Han,
Y. Wang, F. Zhong, Y. Lu, Org. Biomol. Chem. 2011, 9, 6734–6740.
[4] C. A. Mꢂller, C. Markert, A. M. Teichert, A. Pfaltz, Chem. Commun. 2009,
1607–1618.
[10] The deviation of the ratios 6a/6b from the actual e.r. values is likely
caused by the high catalyst loading (30 mol%) that had to be used in
order to obtain satisfactory signal to noise ratios. Under these condi-
tions the assumption of a rapid pre-equilibrium between 1a and 2a,
and 1b and 2b, is no longer valid.
[5] a) C. Markert, P. Rçsel, A. Pfaltz, J. Am. Chem. Soc. 2008, 130, 3234–
3235; b) I. Fleischer, A. Pfaltz, Chem. Eur. J. 2010, 16, 95–99.
[6] a) R. Robiette, V. K. Aggarwal, J. N. Harvey, J. Am. Chem. Soc. 2007, 129,
15513–15525; b) D. Cantillo, C. O. Kappe, J. Org. Chem. 2010, 75, 8615–
8626.
[11] For an analogous analysis of the conjugate addition of aldehydes to ni-
troolefins, see: F. Bꢀchle, J. Duschmalꢃ, C. Ebner, A. Pfaltz, H. Wennem-
ers, Angew. Chem. Int. Ed. 2013, 52, 12619–12623; Angew. Chem. 2013,
125, 12851–12855.
[7] R. E. Plata, D. A. Singleton, J. Am. Chem. Soc. 2015, 137, 3811–3826.
[8] J. Xu, J. Mol. Struct. 2006, 767, 61–66.
Received: September 29, 2016
[9] See Supporting Information.
Published online on && &&, 0000
Chem. Eur. J. 2016, 22, 1 – 6
www.chemeurj.org
5
ꢁ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Communication
COMMUNICATION
&
Asymmetric Catalysis
P. G. Isenegger, F. Bꢀchle, A. Pfaltz*
&& – &&
Asymmetric Morita–Baylis–Hillman
Reaction: Catalyst Development and
Mechanistic Insights Based on Mass
Spectrometric Back-Reaction
Screening
Mass spectrometric catalyst screening
for the asymmetric Morita–Baylis–Hill-
man (MBH) reaction of methyl acrylate
has led to the bifunctional phosphine
(1R,2R)-12a that outperformed previ-
ously reported catalysts in terms of
enantioselectivity and scope. In addi-
tion, mass spectrometric data and kinet-
ic studies provided evidence for CÀC
bond formation as the enantioselectivi-
ty- and rate-determining step.
&
&
Chem. Eur. J. 2016, 22, 1 – 6
www.chemeurj.org
6
ꢁ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!