Journal of the American Chemical Society
Article
slightly basic, which decreased the activity of the Rh-TsDPEN
catalyst, compared to Table 1, entry 3. Entry 6 (Figure S37) of
Table 1 shows the formation of multiple intermediates, the
desired product and side products, during tandem catalysis
confirmed the competitiveness between the two pathways. The
unsupported catalytic tests of the single and tandem trans-
formations demonstrated the incompatibility between two Rh
catalysts and the existence of two competitive tandem reaction
pathways. We reasoned that this tandem reaction presents a
perfect testbed to investigate whether our strategy can
circumvent catalyst incompatibility and photoregulate the
reaction pathway to improve the overall synthetic efficiency.
We investigated the catalytic activity of the SCM for each
single step transformation. SCM 5 catalyzed the asymmetric
1,4-addition of phenylboronic acids to trans-1-phenyl-2-buten-
1-one quantitatively and with excellent enantioselectivities
(96% ee) under the same conditions as the unsupported
rhodium catalyst, which shows the hydrophilic corona is a
suitable catalytic environment for this transformation (Table 2,
entry 1). The ATH mediated with SCM 4 can be suppressed
or accelerated according to the type of stimulus applied
(Figure 3). Under visible light, the ATH with SCM 4 yielded
Next, we investigated whether our smart SCM nanoreactors
SCM 5 can facilitate two incompatible catalytic steps and
photoregulate the reaction pathways. In the presence of KOH,
SCM 5 catalyzed the asymmetric 1,4-addition of phenyl-
boronic acids to trans-1-phenyl-2-buten-1-one to generate the
only intermediate (S)-1,3-diphenylbutan-1-one (Ketone-1)
(Figure 4). After 15 h, UV irradiation was applied resulting
Figure 4. Time course of the tandem catalysis (Table 2, entry 4):
Under visible light irradiation, SCM 5 catalyzed the asymmetric 1,4-
addition of phenylboronic acids to trans-1-phenyl-2-buten-1-one to
generate the only observed intermediate (S)-1,3-diphenylbutan-1-one
accompanied by the complete consumption of starting material trans-
1-phenyl-2-buten-1-one. After 15 h, UV irradiation was turned on to
initiate the ATH to yield the desired and only product (1R,3S)-1,3-
diphenyl-1-butanol (black squares, trans-1-phenyl-2-buten-1-one; red
solid circles, (S)-1,3-diphenylbutan-1-one; purple rhombus, (1R,3S)-
1,3-diphenyl-1-butanol; butyrophenone and (R)-1-phenylbutan-1-ol
were not observed). Reaction conditions: 0.02 mmol of substrate
scale in 3 mL of water at 40 °C with SCM 5 (9 mg, containing 1.3
μmol of Rh-Diene and 1 μmol of Rh-TsDPEN), phenyl boronic acid
(2.0 equiv), KOH (0.5 equiv), and HCOONa (10 equiv) under an
argon atmosphere and the UV irradiation was applied after 15 h.
in the activation of the ATH to yield the desired product
(1R,3S)-1,3-diphenyl-1-butanol in 92% conversion with
excellent enantioselectivities (99% ee, >96:4 dr) (Table 2,
entries 4 and 5). These results indicated that the SCM 5
circumvented the incompatibility of the two catalysts via
compartmentalization and that one reaction pathway can be
selected over competing ones based on the external light
stimulus (compare Table 2, entry 4, and Table 1, entries 5 and
6). The SCMs can be easily recovered via dialysis and recycled
without major loss of reactivity (Table 2, entry 6).
Figure 3. Conversion vs time for Table 1, entry 3, and Table 2, entries
2 and 3. Rh-TsDPEN-catalyzed ATH of (S)-1,3-diphenylbutan-1-one:
purple circles represent the reaction with SCM 4 under UV light
irradiation; black squares represent the reaction with unsupported Rh-
TsDPEN; pink triangles represent the reaction SCM 4 under visible
light. All reactions were performed on a 0.02 mmol substrate scale in
3 mL of water at 40 °C with 5 mol % Rh catalyst loading and
HCOONa (10 equiv).
A limited substrate screen was performed using SCM 5
(Table 3). Aromatic boronic acids, substituted with either
electron-withdrawing or -donating groups, can be converted in
high yields and ee and dr values (Table 3, entries 1−3).
Aliphatic enone also underwent micellar tandem catalysis with
high conversions (98%) and enantioselectivities (99% ee, 5:1
dr). In contrast, the unsupported one-pot tandem catalysis
only yielded 48% desired product under the same reaction
conditions (Table 3, entry 4).
trace amounts of product in 20 h (Table 2, entry 3). Under UV
irradiation, however, the ATH with SCM 4 was slightly
accelerated in comparison to the unsupported Rh−TsDPEN
catalyst with quantitative conversion and excellent enantiose-
lectivities (99% ee, >96:4 dr) (Figure 3 and Table 2, entry 3).
We hypothesize that under visible light the hydrophobic SP
cross-linking layer prevents the diffusion of the hydrogen
donor HCOONa into the core resulting in the suppression of
the ATH. Under UV-light irradiation, the SP isomerizes to MC
resulting in an increase in hydrophilicity of the cross-linking
layer and ultimately an increase in permeability for HCOONa.
The kinetic data (Figure 3) and the catalytic tests of the ATH
(Table 2, entries 2 and 3) demonstrate that the phototriggered
reversible spiropyran to merocyanine transition results in the
gated behavior to SCMs, which is key to discriminating
between substrates/reagents during catalysis; photomodulating
the diffusion of HCOONa and providing an On/Off control
over ATH catalysis.
CONCLUSION
■
In summary, we realized the concept of compartmentalization
and responsiveness with an artificial multicatalytic system that
can photoregulate an efficient synthetic pathway for non-
orthogonal tandem catalysis in water. The key design is a
responsive cross-linking layer between the two compartments
in a core−shell micellar nanostructure. The smart nanostruc-
ture responds to light irradiation in a wavelength-selective
4710
J. Am. Chem. Soc. 2021, 143, 4705−4713