Angewandte
Communications
Chemie
cient, and surprisingly, the conversion was independent of the
yields of 32% (Table 1, entries 5 and 6 and Figure S24). In
loading of the two homogeneous catalysts. These results
indicate that guest encapsulation is essential for photoinduced
dimerization as it gives rise to improved coupling conversions
by prearranging the substrates close to the photo- and redox-
active Ru sites, simultaneously rendering the electron transfer
more efficient and stabilizing reactive radical species (see
below).
contrast, when the photoreactions were carried out with chiral
D/L-RuL metalloligand complexes (80 mol%), much lower
3
enantioselectivities and yields were observed (Table 1,
entries 11 and 12). The reduced coupling conversion of 3
compared with that of 1 may be related to the steric hindrance
imparted by the bromo substituent and faster guest exchange
in the presence of CH CN; however, the higher enantiose-
3
The stereoselectivity was explored by subjecting homo-
chiral D/L-MOC-16 to similar reaction conditions; the
similarity of the host–guest interactions in 1ꢀD-Ru-MOC
and 1ꢀL-Ru-MOC to those in 1ꢀrac-Ru-MOC was con-
lectivities of the D/L-MOC-16-catalyzed photoreactions
compared with the D/L-RuL photoreactions suggest that
3
the confined chiral space effectively renders the reaction
enantioselective. This enantioselectivity is due to host–guest
chiral recognition and stereocontrol of the dimerization
transition state (see below).
1
firmed by H NMR analysis (Figure S10). After irradiation of
1
ꢀD-Ru-MOC, product 2 was obtained, isolated, and ana-
lyzed by HPLC on a chiral stationary phase, which gave an
enantiomeric excess of 26% ee (Figures S11 and S12). How-
ever, when isolated 2 was kept in isopropanol solution for
The cage effect is also reflected in the fact that the
stereoselectivity depends on the host/guest ratio. The photo-
reaction of 3 with 5 mol% of D/L-MOC-16 (host/guest =
1:20) produced (S)/(R)-4 with enhanced enantioselectivities
of 54 and 58% ee, but at the expense of a lower yield of 9%
(Table 1, entries 7 and 8 and Figure S24). Decreasing the
D/L-MOC-16 loading to 2 mol% (1:50) caused a remarkable
decline of both the enantioselectivity and the yield (Table 1,
entries 9 and 10). According to the guest inclusion capacity
determined by NMR titration (Figures S17 and 18), one cage
may accommodate about nine guests, so a host/guest ratio
below 1:10 leaves more molecules of 3 outside the cage that
are subject to guest exchange between the free and encapsu-
lated states. Consequently, the photoreaction with a host/
guest ratio of 1:20 on the one hand ensures that a sufficient
number of substrates 3 are found in the cages for better
enantioselectivity, and on the other hand gives a lower yield
because there are many free molecules of 3. On the contrary,
the photoreaction with a host/guest ratio of 1:10 (10 mol%
D/L-MOC-16) gives higher yields, but cannot achieve the
optimum enantioselectivity owing to inevitable guest
exchange. For the reaction in much lower 1:50 ratio, there
are far more free molecules of 3 than captured ones, resulting
in both lower ee and yield. This speculation was supported by
2
4 h, the enantiopurity dropped drastically to 4% ee. This
finding confirms that the enantiomers of 2 can rapidly
racemize because of the low barrier of interconversion so
that the real enantioselectivity of the dimerization of 1 in the
presence of homochiral D-MOC-16 cannot be determined
with absolute certainty. Nevertheless, in the presence of
D-MOC-16 in water, the ee decreased much more slowly,
namely from 26 to 18% ee over 24 h (Figures S11–S13).
Therefore, homochiral D-MOC-16 can protect the stereo-
chemically unstable enantiomers of 2 from rapid racemiza-
tion. These results suggest that the enantioselectivity of the
dimerization is determined by the chirality of the confined
coordination space, probably by virtue of substrate prear-
rangement in the cage.
To further confirm the enantioselective photoinduction by
the cage, we used 3-bromo-2-naphthol (3) as the substrate to
examine its transformation into 4 in the presence of enantio-
pure D/L-MOC-16. Encapsulation of 3 in D/L-MOC-16 was
1
established by H NMR titration of 3 with the enantiomeric
1
1
cages and the H– H COSY and NOESY spectra of 3ꢀD-
MOC-16 at 298 K in [D ]DMSO/D O (1:4 v/v) solution
6
2
(
Figures S14–S16). Similar host–guest interactions were pres-
a control reaction with a 1:10 host/guest ratio in pure CH CN,
3
ent in CD CN/D O (1:4 v/v; Figures S17 and S18). Consider-
ing the insolubility of 3 in water and the poor NMR
where guest exchange becomes significant, and the ee (12%)
and yield (17%) are low as expected (Table 1, entries 13 and
14).
3
2
resonances of MOC-16 in CD CN (Figure S19), a CH CN/
3
3
H O (1:1 v/v) solution was chosen as the reaction medium to
guarantee effective guest encapsulation and prevent copreci-
pitation of 3 and MOC-16 (Figure S20). In the above NMR
A reaction pathway involving a radical mechanism under
aerobic conditions is proposed in Figure 3. The photoreaction
2
II
may proceed in three steps. The photoredox Ru centers are
2
+
studies, we observed that the CH CN solvent actually
irradiated by visible light to produce a *MOC-16 excited
+
3
3
diminished the hydrophobic effect of the cage and accelerated
guest exchange. Nevertheless, photoinduced coupling of 3 by
D/L-MOC-16 resulted in compound 4 as the sole product as
state, which is quenched by O to give MOC-16 . The O
2 2
molecules act as sacrificial oxidants and are reduced to afford
H O , which can easily produce hydroxyl radicals (see the
2
2
1
13
expected, which was confirmed by H/ C NMR spectroscopy
and ESI-MS (Figures S21 and 22). The R/S configuration of 4
was determined by matching the CD spectra of (R)/(S)-4 with
the calculated electronic circular dichroism (ECD) spectra
ESR experiments in the Supporting Information). Singlet
oxygen could also be generated in this process, but did not
3
+
damage 1 (Figures S25 and 26). MOC-16 may then oxidize
the naphthol substrate by single-electron transfer to give
radical species A. Radical A reacts with a hydroxyl radical to
give the intermediate naphthalene-1,2-dione (B). Afterwards,
A can react with B to yield 2 or two radical A species may
combine to form BINOL. In our case, regioselective 1,4-
coupling occurs exclusively to produce 2 without detectable
amounts of BINOL.
(
Figure S23). The absolute configuration of (+)-4 was corre-
lated to (R)-4, which is the product preferentially formed in
the presence of L-MOC-16.
As seen from Table 1, 10 mol% of D/L-MOC-16 enabled
the coupling of 3 into (R)/(S)-4 with moderate enantioselec-
tivities (32% ee for (S)-4 and 34% ee for (R)-4) and identical
Angew. Chem. Int. Ed. 2017, 56, 1 – 6
ꢀ 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
3
These are not the final page numbers!