J.I. García et al. / Journal of Catalysis 258 (2008) 378–385
379
enantioselection comes from the different favored reaction chan-
nels, as a function of the different steric interactions between the
ester group and the Box substituents.
One aim of the present work was to prove that specially de-
signed ligands can lead to results in the heterogeneous phase that
are rarely achieved in homogeneous conditions. Another aim was
to try to rationalize the role of the clay surface in the stereochemi-
cal course of cyclopropanation reactions, catalyzed by immobilized
Box–Cu complexes, by comparing the results obtained using C2-
and C1-symmetric Box ligands. The general structures of all the
bisoxazoline ligands used in this work are shown in Fig. 2.
2. Experimental
Fig. 1. Some structures of Box ligands lacking C2-symmetry.
2.1. Synthesis of Box ligands
oselectivities similar or worse than those in the homogeneous
phase [10,11].
Bis(oxazoline) ligands were synthesized as described previously
ꢀ
ꢀ
[28,32]. The novel Box (5R, 5 R)-2,2 -(propane-2,2-diyl)bis(5-phe-
nyl-4,5-dihydrooxazole) (1gs) was prepared from 2,2-dimethyl-
malononitrile and (R)-2-amino-1-phenylethanol to yield 83% of
the target compound in a reaction time of 72 h. The crude product
was purified by recrystallization from diethyl ether (Et2O) to afford
1gs as a white solid. [α]D = −153.22 (c = 1, CH2Cl2). 1H NMR:
7.24–7.18 (m, 10 H), 5.45 (dd, J = 10.26, 8.27 Hz, 2H), 4.24 (dd,
However, surface-mediated selectivity has been observed in
this reaction either by changing the reaction solvent [12] or by
using supported ionic liquid films [13]. The decrease in either
the dielectric permittivity of the solvent or the thickness of the
ionic liquid film induces the approach of the Box–Cu complex to
the clay sheet in the immobilized catalyst, leading to a reversal
of the diastereoselectivity (cis-cyclopropanes are the major prod-
ucts) and also to a reversal of the absolute configuration of the
major cis-cyclopropane. These surface effects also have been re-
ported for the same cyclopropanation reaction when immobilized
pyridinoxazoline–copper (Pyox–Cu) complexes were used as cata-
lysts [14], as well as for Diels–Alder reactions catalyzed by Box–Cu
complexes immobilized on silica [15,16]. In fact, given the fact that
chirality is easier to achieve in two dimensions than in three di-
mensions [17,18], heterogeneous catalysis should take advantage of
surface confinement effects more often if such effects were sought
out, instead of trying to avoid them.
In the case of cyclopropanation reactions, previous results
have demonstrated the crucial influence of the clay surface on
the stereochemical course of the reaction. In these cases, lig-
and C2-symmetry disappears through the effect of the surface.
These results demonstrate the need to design tailored Box lig-
ands for heterogeneous catalysis to enhance the surface effect
and to promote the preferential formation of cis-cyclopropanes,
some of which have interesting applications [19–21] and are dif-
ficult to obtain in other ways [19,22–27]. Following this idea, we
recently reported the synthesis of a new family of chiral Box lig-
ands that lack C2-symmetry and have different steric surroundings
(Fig. 1) [28]. Other authors also have recently reported the synthe-
sis of C1-symmetric Box ligands, but with a different purpose [29].
Experimental studies carried out in the homogeneous phase, to-
gether with theoretical calculations, show that C2-symmetry is not
a prerequisite to obtaining good enantioselectivity [30,31] and that
∗
J = 14.39, 10.26 Hz, 2H), 3.74 (dd, J = 14.39, 8.27 Hz, 2H), 1.60
(s, 6H). 13C NMR: 169.24, 140.84, 128.69, 128.18, 125.89, 81.63,
62.85, 38.92, 24.49. Elemental analysis. Experimental: C, 74.97;
H, 6.70; N, 8.28; O, 10.05. Calculated: C, 75.42; H, 6.63; N, 8.38;
O, 9.57.
2.2. Immobilization of Box–Cu(OTf)2 complexes on laponite
◦
Laponite was dried at 140 C overnight before the immobiliza-
tion process. The Box–Cu complexes were prepared by dissolving
the copper salt (CuOTf2, 0.18 mmol) and the ligand (0.20 mmol) in
anhydrous dichloromethane (1.5 ml). The mixture was stirred for
15 min, and the insoluble materials were removed by microfiltra-
tion. The solvent was evaporated under reduced pressure, and the
complex was redissolved in methanol (13 ml). Laponite (0.5 g) was
then added to the solution. The suspension was stirred at room
temperature for 24 h, after which the solid was filtered off and
thoroughly washed with methanol and dichloromethane. The cata-
lyst was dried for 24 h under vacuum at room temperature before
use.
2.3. Catalyst characterization
Copper analyses were carried out by plasma emission spec-
troscopy on a Perkin–Elmer Plasma 40 emission spectrometer.
Transmission FTIR spectra of self-supported wafers evacuated
−4
◦
Torr) at 120 C were recorded with a Mattson Genesis
(<10
R1=Ph, R2=R3=R4=R5=R6=H:
1a
R1=R2=Ph, R3=R4=R5=R6=H:
1as
1bs
1cs
1ds
1es
1fs
t
t
R1= Bu, R2=R3=R4=R5=R6=H:
1b
1c
1f
1h
1i
R1=R2= Bu, R3=R4=R5=R6=H:
R1, R5=Indanyl, R2=R3=R4=R6=H:
R1=Ph, R2=R3=R5=R6=H, R4=Me:
R1,R5=R2, R6=Indanyl, R3=R4=H:
R1=R2=Me, R3=R4=R5=R6=H:
t
i
R1= Bu, R2=Me, R3=R4=R5=R6=H:
R1=R2= Pr, R3=R4=R5=R6=H:
t
R1= Bu, R2=CH2Ph, R3=R4=R5=R6=H:
R1=R2=Ph, R3=R4=Me, R5=R6=H:
R1=R2=R3=R4=H, R5=R6=Ph:
t
R1= Bu, R2=Ph, R3=R4=R5=R6=H:
1j
1gs
Fig. 2. General structure of all the Box ligands used in this work.