Insights into the spontaneous emergence of enantioselectivity in an asymmetric Mannich reaction carried out without external catalyst

Article abstract of DOI:10.1016/j.tetasy.2012.10.021

ESI-MS, chiral HPLC, time-resolved ¹H NMR and optical rotation measurements were performed to gain insights into the nature of spontaneous mirror symmetry breaking in the catalyst-free Mannich reaction of PMP protected α-iminoethylglyoxylate wi

Full text of DOI:10.1016/j.tetasy.2012.10.021

Tetrahedron: Asymmetry 23 (2012) 1663–1669
Contents lists available at SciVerse ScienceDirect
Tetrahedron: Asymmetry
journal homepage: www.elsevier.com/locate/tetasy
Insights into the spontaneous emergence of enantioselectivity in an asymmetric
Mannich reaction carried out without external catalyst
⇑
Felix E. Held, Anja Fingerhut, Svetlana B. Tsogoeva
Department of Chemistry and Pharmacy, University of Erlangen-Nuremberg, Henkestraße 42, 91054 Erlangen, Germany
a r t i c l e i n f o
a b s t r a c t
ESI-MS, chiral HPLC, time-resolved 1H NMR and optical rotation measurements were performed to gain
insights into the nature of spontaneous mirror symmetry breaking in the catalyst-free Mannich reaction
Article history:
Received 4 October 2012
Accepted 29 October 2012
of PMP protected
a-iminoethylglyoxylate with hydroxyacetone. Spontaneous temporary generation of
enantiomeric excesses of up to 7.4% in the major syn diastereomer is reproducibly observed in aqueous
phosphate buffers, starting from achiral conditions. The syn-product ee values for both enantiomers
[(2S,3S) and (2R,3R)], although not with stochastic distribution, have been observed with no clear prefer-
ence for either enantiomer.
Ó 2012 Elsevier Ltd. All rights reserved.
1. Introduction
of a chiral surplus was implied in the theoretical study of the
product assisted hemiacetal formation of acetone with trihaloac-
The origin of biomolecular homochirality, that is that only one
enantiomeric form of amino acids dominates in Nature, still belongs
tooneof themostprofoundconundrumsinscience.1–8 Could the(R)/
(S) symmetry of biomolecules have become broken spontaneously?
One of the main dogmas of asymmetric synthesis has it that enantio-
selective reactions always give racemic products in the absence of
chiralinductorssuchaschiralcatalysts, auxiliaries, solventsorcircu-
larly polarized light.9–11 This partly explains the sensation caused by
Soai’s report in 1995 on the dramatic spontaneous asymmetric
amplification in the irreversible alkylation reaction of pyrimidine
carbaldehydes with diisopropyl zinc.12 At completion, the reaction
producesadominanceofeitherthe(R)-or (S)-productin thereaction
mixture, depending on an initial accidental and miniscule enantio-
meric imbalance. It has been disclosed that autocatalytically active
homochiral dimers of the product seem to be mainly responsible
for the symmetry breaking effect.13–16
etaldehydes.31
Self-replication processes of products have been studied earlier,
involving oligomeric nucleotide molecules and b-sheet pep-
tides.32,33 Even before the Soai reaction was discovered, it was sug-
gested that an enantioselective autocatalytic reaction in
conjunction with a positive non-linear effect of catalyst versus
product ee should lead to spontaneous asymmetric amplification.34
In 2007, it was reported that in the product assisted asymmetric
Mannich reaction35 of PMP protected
a-iminoethylglyoxylate with
acetone, a spontaneously generated product enantiomeric excess
of 9.4% can be observed with 31% yield after four days reaction
time – even when the initial reaction mixture is achiral or racemic
(Scheme 1a).36 This result has indicated that seemingly well-
known37 organic reactions might have much more complicated
mechanisms than was previously assumed. In order to study this
process further, we decided to use hydroxyacetone as the donor
The possibility of spontaneous autoamplification of ee in chemi-
cal reactions has already been the focus of several theoretical inves-
tigations.17–25 The first proposal of a mathematical mechanism for
the spontaneous emergence of enantiomeric excess through enan-
tioselective autocatalysis in conjunction with mutual inhibition of
enantiomers was made by Charles Frank in 1953.26 Symmetry break-
ing by spontaneous deracemisation of chiral conglomerates has also
been reported.27–29
OMe
(S) or (R)
(a)
O
HN
O
without catalyst
2-8 days, r.t.
MeO
CO2Et
11-36% yield
0.5-9.5% ee
In this process, autocatalysis was again proposed to be involved
N
in the crystallization.30 Very recently, the spontaneous generation
OMe
O
without catalyst
1-6 h, r.t.
H
CO2Et
(b)
O
HN
⇑
Corresponding author. Tel.: +49 (0) 9131 85 22541; fax: +49 (0) 9131 85 26865.
OH
CO2Et
E-mail address: tsogoeva@chemie.uni-erlangen.de (S.B. Tsogoeva).
Electronic supplementary information (ESI) available at http://dx.doi.org/
10.1016/j.tetasy.2012.10.021: Experimental procedures, spectroscopic, HPLC and
ESI-MS data.
OH
Scheme 1. Mannich reactions in the absence of external catalyst.
0957-4166/$ - see front matter Ó 2012 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.tetasy.2012.10.021

1664
F. E. Held et al. / Tetrahedron: Asymmetry 23 (2012) 1663–1669
(Scheme 1b), instead of acetone, because of the higher reactivity of
the former, and hence, the expectation of reduced reaction times.
Herein we report our results on the reaction kinetics of this
transformation using 1H NMR studies, which show an induction
period of product formation thus evidencing the involvement of
an autoinductive process. Based on ESI-MS measurements and 1H
NMR studies, we propose the in situ formation of an organic base,
a diamine side product, which could assist in the enolate formation
from the ketone. Furthermore, polarimetry was used to monitor
the variation of the optical rotations during the reaction. It is also
shown that optical rotation emerges spontaneously and temporar-
ily at the beginning of the reaction.
reaction of acetone with the same imine, as reported previously.36
Notably, we observed syn-product ee values for both enantiomers
[(2S,3S) and (2R,3R)), although not with stochastic distribution.
No clear preference for either enantiomer can be recognized. It
should be noted that the initial conditions differ in the experiments
and that the small number of runs forbids a statistical evaluation.
Several reaction mixtures remain exactly racemic when the ee is
measured even after only 1 h (entries 1, 3, 4 and 10) and after up
to 4 h (entries 6, 9 and 13). After two days reaction time, all mix-
tures were fully racemic as could be predicted from equilibrium
thermodynamics.
Monitoring of ee values in two independent experiments in a
phosphate buffer at pH 7 (entries 15 and 16) in short intervals, re-
vealed that in both cases, the ee value observed in the first mea-
surement decreased rapidly to a near-zero value after about 1.5 h
(Fig. 1). This shows that thermodynamic equilibrium is approached
quickly. Significant deviations from the equilibrium values occur
shortly after the reaction starts at low yields (Table 1).
2. Results and discussion
Firstly, the reaction of PMP protected
a-iminoethylglyoxylate
with hydroxyacetone (Scheme 1b)37 was studied under various
conditions (solvent, temperature, reactant concentration) and
without any external catalyst. The results are summarized in Ta-
ble 1. The diastereoselectivity is almost the same (84:16 to 89:11
dr, entries 1–14) under most of the studied conditions and for
the investigated reaction times, except for the results in aqueous
conditions (in phosphate buffer) measured after 1 h reaction time
(66:34 to 72:28 dr, entries 15–18), with the syn-diastereomer al-
ways being predominant. This implies that the syn-isomer is the
thermodynamically more stable one. The yield strongly depends
on the temperature; the higher temperatures gave better yields
after the same reaction time: while in DCM at room temperature
after 1 h, the yield was only 9.4% (Table 1, entry 1), it is 28.4% at
40°C (entry 10).
Table 1
Mannich reaction of imine with hydroxyacetone (from Scheme 1b) carried out
without an external catalyst
Figure 1. Monitoring of the ee-values with the reaction time under aqueous
conditions (phosphate buffer) for the syn-product. (Table 1, entries 15 and 16).
Entry Time, h Solvent T, °C Yielda (%) drb (syn:anti) eeb (%) (syn)
1c
1
1
1
1
1
3
6
6
4
1
6
6
3
6
DCM
DCM
DCM
DCM
DCM
DCM
DCM
DCM
DCM
DCM
THF
rt
rt
0
0
0
0
0
0
À20
40
rt
0
9.4 88:12
n.d. 86:14
n.d. 82:18
n.d. 89:11
n.d. 84:16
n.d. 88:12
n.d. 84:16
15.2 88:12
10.3 86:14
28.4 82:18
14.7 86:14
15.7 86:14
n.d. 82:18
15.3 86:14
n.d. 66:34
n.d. 70:30
n.d. 72:28
n.d. 67:33
rac
1.0 (2S,3S)
rac
rac
2.1 (2S,3S)
rac
2.0 (2S,3S)
2.4 (2R,3R)
rac
We also studied the time dependence of the imine conversion in
CDCl3, employing 1H NMR (Fig. 2). After 80 min, that is at about the
time when ee values begin to decrease in the phosphate buffer
(Fig. 1), the imine was fully consumed. We observed an S-shaped
curve with an induction period and significant rate of acceleration,
which indicates the involvement of an autoinductive process.
2d
3c
4d
5d
6c
7d
8e
9c
10c
11c
12c
13d
14c
15d
16d
17d
18d
rac
4.3 (2R,3R)
1.6 (2S,3S)
rac
1.4 (2R,3R)
4.4 (2S,3S)
6.3 (2R,3R)
7.4 (2S,3S)
5.1 (2S,3S)
THF
THF
0
Bufferf
rt
rt
rt
rt
rt
1.7 Bufferf
1
1
1
Bufferf
Bufferf
Bufferf
a
Yields of the isolated products.
Determined by chiral HPLC analysis (Daicel Chiralpak IA) in comparison with
b
authentic racemic material.
c
0.3 M imine concentration.
0.5 M imine concentration.
0.1 M imine concentration.
Phosphate buffer at pH 7.
d
e
f
Figure 2. Conversion time profile for the Mannich reaction carried out with imine
(c = 0.5 M) and hydroxyacetone (3 equiv) in CDCl3.
When the reaction was left to run in a buffer solution for 6 h, in-
stead of one, the dr and yield approach the values observed for the
reactions in organic solvents (entry 14).
Despite the achiral starting conditions, small enantiomeric ex-
cess values of up to 7.4% have reproducibly been observed in the
syn-product from one to a few hours after the start of the reaction
in some experiments in accordance with similar results for the
In order to verify the spontaneous emergence of the enantio-
meric excess in the Mannich reaction, we carried out optical rota-
tion measurements with a polarimeter in DCM (Fig. 3). In four
consecutive independent experiments under the same conditions
we observed significant optical rotation values increasing shortly

F. E. Held et al. / Tetrahedron: Asymmetry 23 (2012) 1663–1669
1665
30
20
10
0
0
20
40
60
80
100
SiCl4
(1 equiv)
Cl
OH
*
O
-10
-20
-30
*
toluene, rt
Figure 3. Monitoring of the optical rotation during the reaction of imine (c = 0.5 M) with hydroxyacetone (10 equiv) in DCM (4 independent runs).
OMe
OMe
HN
H
O
N
CO2Et
HO
H
EtO2C
syn anti
t
7.90 7.85 7.80 7.75 7.70 7.65 7.60
4.65 4.60 4.55 4.50 4.45 4.40 4.35
Figure 4. Fragments of time-resolved 1H NMR spectrum (400 MHz) recorded during the reaction of imine (c = 0.5 M) with hydroxyacetone (3 equiv) in CDCl3.

1666
F. E. Held et al. / Tetrahedron: Asymmetry 23 (2012) 1663–1669
after mixing the reactants. The rotation angle vividly swerves from
positive to negative values or vice versa during the first 40 min of
the reaction. A control experiment with pure solvent (DCM) re-
mained near the zero line throughout the whole observation time.
To determine whether the effect depends on the nature of the reac-
tion, we also performed a similar monitoring experiment with the
irreversible meso epoxide ring opening reaction shown in the in-
sert in Figure 3: the optical rotation was negligible during the
whole reaction time. This shows that not all stereoselective reac-
tions display the effect of spontaneous mirror symmetry breaking.
The epoxide ring opening differs from the Mannich reaction in its
irreversibility, which might indicate that reversibility is crucial
for the observation of the aforementioned effect. Furthermore,
the product of the epoxide ring opening (2-chlorocyclohexanol)
could hardly compete with the external promoter/reagent SiCl4
(Lewis acid) in activating the epoxide. Hence, product assistance
is unlikely in this reaction.
These observations motivated us to further elucidate the mech-
anism of the Mannich reaction with time-resolved 1H NMR (Fig. 4).
We were able to detect the formation of several unidentified
species in the course of the reaction (see also SI). Some of the spe-
cies appear to be intermediates (e.g. see signal at 4.6 ppm, which
almost disappears during the later stages of the observation time
window). We suspect that some of the species might be side prod-
ucts taking part in the Mannich reaction.
the time after which complete conversion of the imine was
achieved, we suspected that the imine conversion might provide
the thermodynamic driving force fueling the non-equilibrium pro-
cess necessary to deviate from the racemic state.
To gain an indication of the potential nature of these side prod-
ucts, we performed ESI-MS measurements for the reaction of
hydroxyacetone with imine, and also, for the previously studied
reaction of acetone with the same imine. In both cases we observed
(besides other peaks, which correspond, to the imine and product)
the formation of a species at m/z 352.2 (Fig. 5), from which it is
obvious that the ketone (acetone or hydroxyacetone) is not in-
volved in the formation of this species. In both reactions, we ob-
served the formation of anisidine traces (characteristic signal at
m/z 124.1 [M+H]+ in ESI-MS or LC–MS spectra).
Therefore, we hypothesized that the peak at m/z 352.2 might
correspond to the presumed diamine product formed in the
reaction of the imine with anisidine. To determine if our hypothe-
sis was plausible, we stirred the 1:1 mixture of the imine with
anisidine and measured 1H NMR and ESI-MS spectra at different
time intervals.
A peak at m/z 352.2 (corresponding to the
anticipated species) was detected in ESI-MS spectra recorded after
3 h and characteristic signals of the formed product in 1H NMR
spectra, measured after 10 and 60 min, respectively (see SI).
Injection of a sample taken from the above reaction mixture into
a chiral HPLC column, which we employed in the stereoselectivity
determinations (Table 1), showed that the diamine had the same
retention time as the (2R,3S)-enantiomer of the anti-diastereomer
Since the observation time of the highest ee values at about 1 h
after the beginning of the reaction approximately coincides with
Figure 5. ESI-MS spectra (positive mode) for the reactions of (a) hydroxyacetone with imine (60 min after the start of the reaction); and (b) acetone with imine (21 h after the
reaction start).

F. E. Held et al. / Tetrahedron: Asymmetry 23 (2012) 1663–1669
1667
(see SI). This means that the ee values in the minor anti-diastereo-
mer are not reliable due to peak overlapping and have therefore
not been reported in Table 1. This might also explain the lower dia-
stereomeric ratio (ca. 70:30 syn/anti) observed after 1 h in phos-
phate buffer (entries 15–18, Table 1).
Notably, the same side product observed in the reaction of the
imine with acetone has a pronouncedly different retention time
than either the (R)- or (S)-product (see SI).
The information thus gained led us to assume the following pro-
posal for the involvement of the formed base (side product) in the
Mannich reaction, relevant for both transformations with acetone
and hydroxyacetone respectively: via a reversible equilibrium,
the imine can revert to its constituents, anisidine and ethylglyoxy-
late, followed by the reaction of imine with the formed anisidine to
give the side product (a diamine species), which we assume will
act as an in situ formed organocatalyst and which might assist in
the formation of an enolate as a more active nucleophile, thus pro-
moting the first step of the Mannich reaction. This first step is
apparently rate determining, which could explain the conversion
time profile shown in Figure 2. The second step, the CC bond for-
mation, is configuration determining and could potentially be af-
fected by the involvement of the chiral Mannich product or by a
chiral side product, which might direct the product formation in
an enantioselective manner. Such product assistance through acti-
vation by hydrogen bridging in the configuration determining step
of the Mannich reaction has been previously reported.35 It should
be noted that this asymmetric ‘organoautocatalysis’, while affect-
ing the product ee value, does not necessarily result in reaction
rate acceleration. This mechanistic hypothesis, however, does not
explain why ee values can emerge spontaneously.
As proposed earlier,35a,36 the syn-product assisted nucleophilic
addition implies the initial formation of hydrogen bonded product
homo- (RRÁRR or SSÁSS) and heterodimer (RRÁSS) complexes. When
the reaction is enantioselective, homodimers must form faster than
heterodimers. The homodimers may be higher in energy with re-
spect to heterodimers and therefore may dissociate faster to the
monomeric chiral autocatalysts which could became engaged again
in further catalytic cycles, while the heterodimers could have a
much slower lifecycle.38 However, first order enantioselective
‘autocatalysis’ alone cannot give rise to asymmetric amplification.13
A further assumption is necessary. Due to a fluctuation, one of
the two homodimers may form accidentally with a tiny excess.
This small deviation can be amplified, as monomeric product enan-
tiomers (released through dissociation of the initially formed di-
meric products) may recombine again fast to more stable
heterodimers (before becoming engaged again in a further round
of enantioselective autocatalysis), resulting in the depletion of
the minor enantiomer. This is reminiscent of the well-known mu-
tual inhibition mechanism of Frank, where the initial opposite
enantiomer products of the enantioselective autocatalysis step
recombine rapidly to irreversibly give a heterodimer, thereby being
removed from the reaction.26 Such a combination of an autocata-
lytic reaction step of only linear order in the product concentration
and mutual inhibition allows spontaneous autoamplification of the
MeO
MeO
N
MeO
OMe
MeO
C18H20N2O4
M = 330.38 g/mol
ESI-MS: 352.2 [M+Na]+
O
NH2
H
NH
*
O
OH
*
NH
*
O
HN
*
NH
O
OH
*
OH OEt
O
OEt
EtO
O
*
O
O
*
O
EtO2C
*
HN
O
OH OEt
OH
OEt
OEt
C18H24NO8
C25H32N2O8
C7H12O5
OMe
M = 382.39 g/mol
M = 488.53 g/mol
M = 176.17 g/mol
ESI-MS: 405.1 [M+Na]+
ESI-MS: 511.4 [M+Na]+
ESI-MS: 199.1 [M+Na]+
OMe
OH
O
MeO
MeO
OMe
O
MeO
N
NH
O
HN
*
O
OH
+
H
O
O
*
HN
H
CO2Et
NH2
OEt
C4H6O3
M = 102.09 g/mol
C14H19NO5
OEt
HN
OH
C7H9NO
C11H13NO3
M = 281.30 g/mol
M = 123.15 g/mol
M = 207.23 g/mol
ESI-MS: 304.1 [M+Na]+
ESI-MS: 230.1 [M+Na]+
ESI-MS: 124.1 [M+H]+
OMe
OMe
C23H25N3O4
M = 407.46 g/mol
OMe
MeO
NH2
MeO
O
O
HO
OMe
O
HN
ESI-MS: 430.3 [M+Na]
N
N
O
O
O
*
*
OH
OH
H
O
*
N
*
H
OMe
OH
N
N
H
*
OEt
*
H
CO2Et
HO
OH
C16H16N2O3
C19H22N2O5
M = 358.39 g/mol
OMe
OMe
MeO
OMe
CO2Et
C17H25NO7
M = 284.31 g/mol
ESI-MS: 381.3 [M+Na]+
M = 355.38 g/mol
ESI-MS: 307.3 [M+Na]+
N
*
N
*
*
*
ESI-MS: 378.2 [M+Na]+
CO2Et
HO
MeO
OH
C25H32N2O8
MeO
N
M = 488.53 g/mol
O
O
ESI-MS: 511.4 [M+Na]+
O
*
OH
*
O
H
NH
*
O
OH
*
O
O
OEt
H
EtO
O
OH
*
N
OH
N
H
O
OH
N
H
H
C9H9NO3
C23H26N2O8
C12H14NO5
OMe
OMe
M = 179.17 g/mol
M = 458.46 g/mol
M = 252,24 g/mol
OMe
ESI-MS: 202.1 [M+Na]+
ESI-MS: 481.2 [M+Na]+
ESI-MS: 276.2 [M+Na]+
Figure 6. Proposed network of the Mannich reaction system.

1668
F. E. Held et al. / Tetrahedron: Asymmetry 23 (2012) 1663–1669
ee, even when the heterodimer formation is not irreversible.20,31 In
this case, the maximum achievable and spontaneously generated
overall ee value (taking into account the product molecules bound
in the form of heterodimers) would obviously be small, probably
only a few percent just as we observe here.
The ee values, once generated, soon fall back to their zero equi-
librium values (Fig. 1); this is indicative of the reversibility of the
whole process. Unfortunately, and in contrast to the famous second
order autocatalytic Soai reaction,13,16 the rate determining and
configuration determining reaction steps most likely do not
coincide for the Mannich reaction studied herein. As a conse-
quence, the background reaction (without product assistance),
which gives a racemic product, and the potentially product as-
sisted reaction pathway, both possess the same activation barrier.
This lowers the temporarily observable enantiomeric excess fur-
ther. It remains to find examples in which the background reaction
is suppressed.
4.2. Synthetic procedures
4.2.1. N-PMP-protected -imino ethyl glyoxylate
a
A solution of ethyl glyoxylate (0.198 mL, 1.00 mmol) (50% in
toluene) was slowly added to a stirring solution of p-anisidine
(100 mg, 0.810 mmol) in dry CH2Cl2 (1.40 mL) with a syringe pump
within 1 h with additional preactivated 4 Å molecular sieves
(1.50 g). After the reaction mixture was stirred at 40°C for 3.5 h,
the mixture was filtered and the solvent from the filtrate was evap-
orated under reduced pressure to afford an almost pure product.
Further purification by column chromatography on silica gel under
nitrogen using dry methylene chloride as an eluent gave the imine
as a viscous yellow oil in 93% yield. 1H NMR (400 MHz, CDCl3): d
[ppm]: 7.94 (s, 1H, N@CH), 7.37 (d, J = 8.99 Hz, 2H), 6.95 (d,
J = 8.98 Hz, 2H), 4.44 (q, 2H, J = 7.13 Hz, OCH2CH3), 3.86 (s, 3H,
CH3O), 1.43 (t, 3H, J = 7.13 Hz, OCH2CH3).
To obtain spontaneously even larger ee values closer to enantio-
meric purity, it is necessary to postulate a pathway for recycling
the minor enantiomer back to the reactant.36,39 Such a hypothetical
process can, however, only take place under explicit non-equilib-
rium conditions in order to avoid a violation of microscopic
reversibility.
4.2.2. General procedure for the spontaneous symmetry
breaking reaction between N-PMP protected a-imino ethyl
glyoxylate and hydroxyacetone
A
solution of N-PMP-protected
a
-imino ethyl glyoxylate
in different solvents
(102,8 mg, 0.496 mmol, 1 equiv)
(c (educt) = 0.1–0.5 mol/L) was stirred with hydroxyacetone at
room temperature for 1 to 6 h. The solvent was evaporated and
the residue was purified by chromatography on a SiO2-column
(CH2Cl2/EtOAc/Petrolether, 1:1:1) and additional preparative TLC
to afford the desired product (Table 1). Enantioselectivities
were determined by chiral HPLC analysis (Daicel Chiralpak IA,
i-Propanol/n-Hexane = 5:95; 4:96, flow rate 1.0 mL/min, k =
254 nm) in comparison with authentic racemic material. 1H NMR
(400 MHz, CDCl3,RT): d [ppm]: 6.78 (d, J = 8.97 Hz, 2H), 6.62 (d,
J = 8.99 Hz, 2H), 4.65 (bd, 1H, OH–CH), 4.48 (bd, 1H, NH–CH),
4.48 (bd, 1H, NH–CH), 4,22 (q, 2H, J = 7.11, OCH2CH3), 3.75 (s, 3H,
CH3O), 2.32 (s, 3H, CO–CH3), 1.27 (t, 3H, J = 7.14 Hz, OCH2CH3–).
The whole reaction network is definitely complex. A conjecture
on its nature is shown in Figure 6.
3. Conclusion
In conclusion, we have gained further insights into an asymmet-
ric Mannich reaction, carried out without external catalyst, and
which is capable of generating small ee values spontaneously in
the early stages of the reaction. Without resorting to the particular
physics of a non-equilibrium steady state,39,40 the spontaneous
asymmetric amplification in the Mannich reaction could be under-
stood from two basic assumptions in accord with the well-known
Frank mechanism of spontaneous asymmetric amplification: first,
the assistance of an asymmetric product in the configuration deter-
mining step of the reaction, and secondly, the greater stability of
the product heterodimer. Our earlier reports on spontaneous sym-
metry breaking in the Mannich reaction with acetone could be fur-
ther confirmed by polarimeter and HPLC measurements, showing
again that the counter-intuitive ee generation under formally achi-
ral conditions is indeed possible. It remains to be seen, whether
this previously overlooked phenomenon could be found in other
stereoselective reactions. It is conceivable that reaction protocols
might be developed in the future, which allows us to exploit this
novel and remarkable effect.
Acknowledgments
We gratefully acknowledge generous financial support from the
Deutsche Forschungsgemeinschaft and COST Action on Systems
Chemistry CM0703. We thank Dr. Michael Mauksch for stimulating
discussions. We further thank C. Placht, H. Maid, Professor Dr. W.
Bauer for time-resolved 1H NMR measurements; Dr. J. Einsiedel,
W. Donaubauer, M. Dzialach for ESI-MS and LC–MS spectra.
References
1. Japp, F. R. Nature 1898, 58, 452.
2. Special issue: Origin of Biochemical Homochirality, eds., Pizzarello, S.; Lahav,
M. Orig. Life 2010, 40, 1.
3. Orgel, L. E. J. Mol. Evol. 1989, 29, 465.
4. Cintas, P. Angew. Chem., Int. Ed. 2002, 41, 1139.
5. Calvin, M. Chemical Evolution; Oxford University Press: London, 1969.
6. Bonner, W. A. Orig. Life 1994, 24, 63.
4. Experimental
4.1. General
7. Decker, P. Orig. Life 1975, 6, 211.
8. Mauksch, M.; Tsogoeva, S. B. Life’s single chirality: origin of symmetry breaking
in biomolecules. In Biomimetic Organic Synthesis; Poupon, E., Nay, B., Eds.;
Wiley-VCH: Weinheim, 2011.
9. Eliel, E. In Asymmetric Reactions and Processes in Chemistry, ACS Symposium
Series; American Chemical Society: Washington, DC, 1982.
10. Mislow, K. Collect. Czech. Chem. Commun. 2003, 68, 849.
11. Feringa, B. L.; van Delden, R. A. Angew. Chem., Int. Ed. 1999, 38, 3418.
12. (a) Soai, K.; Shibata, T.; Morioka, H.; Choji, K. Nature 1995, 378, 767; (b) Soai, K.;
Kawasaki, T. Chirality 2006, 18, 469; (c) Kawasaki, T.; Matsamura, Y.; Tsutsumi,
T.; Suzuki, K.; Ito, M.; Soai, K. Science 2009, 324, 492.
13. Blackmond, D. G.; McMillan, C. R.; Ramdeehul, S.; Schorm, A.; Brown, J. M. J.
Am. Chem. Soc. 2001, 123, 10103.
14. Gridnev, I. D.; Serafimov, J. M.; Quiney, H.; Brown, J. M. Org. Biomol. Chem. 2003,
1, 3811.
15. Singleton, D. A.; Vo, L. K. J. Am. Chem. Soc. 2002, 124, 10010.
16. Schiaffino, L.; Ercolani, G. ChemPhysChem 2009, 10, 2508.
17. Kondepudi, D. K.; Nelson, G. W. Phys. Rev. Lett. 1983, 50, 1023.
Solvents were purified by standard procedures and distilled
prior to use. Reagents obtained from commercial sources were
used without further purification. TLC chromatography was per-
formed on precoated aluminium silica gel ALUGRAM SIL G/
UV254 plates (Macherey, Nagel & Co.). Flash chromatography
was performed using silica gel ACROS 60 Å, (particle size 0.035–
0.070 mm). 1H NMR spectra were recorded in CDCl3 with Bruker
Avance 300 or 400. Enantioselectivities were determined by chiral
HPLC analysis (Daicel Chiralpak IA, i-Propanol/n-Hexane = 5:95;
4:96, flow rate 1.0 mL/min, k = 254 nm) in comparison with
authentic racemic material. Optical rotations were determined on
a PerkinElmer polarimeter, model 341, k = 589 nm.

F. E. Held et al. / Tetrahedron: Asymmetry 23 (2012) 1663–1669
1669
18. Plasson, R.; Kondepudi, D. K.; Bersini, H.; Commeyras, A.; Asakura, K. Chirality
2007, 19, 589.
31. Azofra, L. M.; Alkorta, I.; Elguero, J. J. Phys. Org. Chem. 2012. http://dx.doi.org/
10.1002/poc.3017.
19. Saito, Y.; Hyuga, H. Top. Curr. Chem. 2008, 284, 97.
20. Crusats, J.; Hochberg, D.; Moyano, A.; Ribo, J. M. ChemPhysChem 2009, 10,
2123.
32. von Kiedrowski, G. Angew. Chem., Int. Ed. 1986, 25, 932.
33. Rubinov, B.; Wagner, N.; Rapaport, H.; Ashkenasy, G. Angew. Chem., Int. Ed.
2009, 83, 6683.
21. (a) Blanco, C.; Hochberg, D. Phys. Chem. Chem. Phys. 2011, 13, 839; (b) Blanco,
C.; Stich, M.; Hochberg, D. Chem. Phys. Lett. 2011, 505, 140; (c) Stich, M.; Blanco,
C.; Hochberg, D. Phys. Chem. Chem. Phys. 2012. http://dx.doi.org/10.1039/
C2CP42620J.
22. Iwamoto, K. Phys. Chem. Chem. Phys. 2003, 5, 3616.
23. Micheau, J. C.; Coudret, C.; Cruz, J. M.; Buhse, T. Phys. Chem. Chem. Phys. 2012.
http://dx.doi.org/10.1039/c2cp42041d.
24. Brandenburg, A.; Andersen, A. C.; Nilsson, M.; Höfner, S. Orig. Life 2005, 35, 507.
25. Sandars, P. G. H. Orig. Life 2003, 33, 575.
26. Frank, C. F. Biochim. Biophys. Acta 1953, 11, 459.
27. Viedma, C. Phys. Rev. Lett. 2005, 94, 065504.
34. Puchot, C.; Samuel, O.; Dunach, E.; Zhao, S.; Agami, C.; Kagan, H. J. Am. Chem.
Soc. 1986, 108, 2353.
35. (a) Mauksch, M.; Tsogoeva, S. B.; Martynova, I. M.; Wei, S.-W. Angew. Chem., Int.
Ed. 2007, 46, 393; (b) Amedjkouh, M.; Brandberg, M. Chem. Commun. 2008,
3043; (c) Wang, X.; Zhang, Y.; Tan, H.; Wang, Y.; Han, P.; Wang, D. Z. J. Org.
Chem. 2010, 75, 2403; (d) For recent reviews on organoautocatalytic reactions,
see:.; (e) Tsogoeva, S. B. J. Syst. Chem. 2010, 1–8; (f) Tsogoeva, S. B. Chem.
Commun. 2010, 46, 7662.
36. Mauksch, M.; Tsogoeva, S. B.; Wei, S.-W.; Martynova, I. Chirality 2007, 19, 816.
37. Córdova, A.; Notz, W.; Zhong, G.; Betancort, J. M.; Barbas, C. F., III J. Am. Chem.
Soc. 2002, 124, 1842.
28. Noorduin, W. L.; Vlieg, E.; Kellogg, R. M.; Kaptein, B. Angew. Chem., Int. Ed. 2009,
48, 9600.
38. Mauksch, M. S.; Wei, S.-W.; Freund, M.; Zamfir, A.; Tsogoeva, S. B. Orig. Life
2010, 40, 79.
29. Weissbuch, I.; Lahav, M. Chem. Rev. 2011, 111, 3236.
30. Uwaha, M. J. Phys. Soc. Jpn. 2004, 73, 2601.
39. Mauksch, M.; Tsogoeva, S. B. ChemPhysChem 2008, 9, 2359.
40. Nicolis, G.; Prigogine, I. Proc. Natl. Acad. Sci. 1980, 78, 659.