Communications
Table 3: Differential activation parameters.
in the catalytic system, increasing the catalyst loading
DDH° [kJmolꢀ1
]
DDS° [Jmolꢀ1 Kꢀ1
]
significantly enhances the reaction rate. In the presence of
5 mol% of 1a, the Mannich-type reaction of 2a with 3b is
completed in 2 hours and affords product in 97% yield with
87% ee (Table 2, entry 9). Importantly, when the reaction is
carried out in acetonitrile reversal of the enantioselectivity is
observed compared to that seen with m-xylene and toluene.
The Mannich-type reactions in acetonitrile give the corre-
sponding R products in 88–99% yield with 80–89% ee
(Table 2, entries 10–18).
Entry
Solvent
1
2
3
4
5
6
7
8
toluene
m-xylene
chlorobenzene
CH2Cl2
EtOAc
THF
+15.1
+19.9
+9.99
+16.8
ꢀ24.0
ꢀ29.4
ꢀ19.4
ꢀ14.6
+82.3
+99.6
+59.6
+79.7
ꢀ87.3
ꢀ106
EtCN
MeCN
ꢀ57.2
ꢀ37.2
Considering that Gibbs free energy is defined as DDG° =
DDH°ꢀTDDS°, the difference in temperature profiles
between the S-selective reaction and R-selective reaction in
the present organocatalytic system suggest that enthalpy-
entropy compensation may contribute to the solvent-depen-
dent enantioswitching. This speculation, along with the goal
of characterizing the chiral recognition processes that take
place in the bond-forming reaction,[21] led us to perform a
kinetic analysis of the reaction using Eyring plots[9,22–24] to
obtain the differential activation parameters.
stereodetermining step of solvent-dependent organocatalytic
reactions promoted by 1a. Positive values of DDH° and
S-R
DDS°
in S-selective reactions run in nonpolar solvent
S-R
systems are obtained (Table 3, entries 1–4). In these cases,
differential activation entropies (DDS°S-R) contribute to low-
ering the DDG° of reactions having unfavorable enthalpic
S-R
contributions. In contrast, negative values of DDH° and
R-S
DDS°
control the stereodiscrimination processes in
R-S
In the differential Eyring treatment,[22] the relative rates of
formation of (S)-(+)- and (R)-(ꢀ)-4aa in S-selective and
R-selective reactions are expressed by Equations (1) and (2),
respectively, where DDS° represents the differential activa-
tion entropy and DDH° represents the differential activation
enthalpy.
R-selective reactions in aprotic polar solvents (Table 3,
entries 5–8), in which the DDH°R-S term has a major influence
on lowering the DDG° in the R-selective reactions. Thus,
R-S
the observed solvent-dependent stereodiscrimination in
1a-catalyzed Mannich-type reaction resides in compensating
differences in the enthalpies and entropies of activation.[25]
Additional efforts designed to probe the link between kinetics
and molecular mechanism are underway.[26]
lnðkS=kRÞ ¼ ꢀDDH°S-R=RT þ DDS°S-R=R
lnðkR=kSÞ ¼ ꢀDDH°R-S=RT þ DDS°R-S=R
ð1Þ
ð2Þ
In conclusion, we have developed an enantiodivergent
catalytic Mannich-type reaction by utilizing conformationally
flexible organocatalysts. The simple methodology has a broad
aromatic N-Boc imine substrate scope and it enables selective
access to both enantiomers of the Mannich adducts using a
single chiral organocatalyst. Kinetic analyses uncovered that
the origin of solvent-dependent stereodiscrimination is con-
trolled by the enthalpy-entropy compensation. The stereose-
lectivities of S-selective Mannich-type reactions in nonpolar
solvents are governed by the differences in the entropies of
activation (DDS°S-R), whereas the stereodiscrimination pro-
cesses of R-selective reactions are governed by differences in
the enthalpies of activation (DDH°R-S). We believe that these
findings will serve as a foundation for the design of new
stereoswitchable asymmetric organocatalytic processes.
Additional efforts to apply the concepts described above to
other classes of asymmetric transformations, including dia-
stereoswitching and organocascade processes, are underway.
In accord with Equations (1) and (2), plots of natural
logarithms of the relative rates of formation of (S)-(+)- and
(R)-(ꢀ)-4aa versus reciprocal temperatures were fitted to
straight lines with good correlation coefficients (Figure 1).
These observations confirm that a single mechanism is
operable in the catalytic process occurring in each solvent in
the temperature range explored.[23d]
As seen by inspecting the data in Table 3, both enthalpy
(DDH°) and entropy (DDS°) compensation govern the
Received: August 16, 2010
Revised: September 7, 2010
Published online: October 26, 2010
Keywords: asymmetric synthesis · enantiodivergent catalysis ·
.
enthalpy · entropy · organocatalysis
Figure 1. Eyring plots of ln[(100+% ee)/(100ꢀ% ee)] vs. 1/T for
1a-catalyzed Mannich-type reactions in various solvents. a) S-Selective
reactions in toluene (closed circle; R2 =0.949), m-xylene (cross;
R2 =0.983), chlorobenzene (closed triangle; R2 =0.992), and CH2Cl2
(closed diamond; R2 =0.956). b) R-Selective reactions in EtOAc (open
circle; R2 =0.967), THF (open triangle; R2 =0.998), EtCN (open
square; R2 =0.955), and MeCN (open diamond; R2 =0.981).
[1] For selected recent reviews on asymmetric hydrogen-bond
donor catalysis, see: a) P. M. Pihko, Hydrogen Bonding in
Organic Synthesis, Wiley-VCH, Weinheim, 2009; b) Z. Zhang,
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2010, 49, 9254 –9257