4
C.G. Dos Santos et al. / Tetrahedron Letters xxx (2016) xxx–xxx
supernatant tested negative for gold in solution. The recyclability
of the AuNP@Nb catalysts was tested obtaining a maximum
Lyophilization of a fully deactivated catalyst for 8 h was per-
formed to establish if Lewis acid deactivation was due to water
adsorption, and the possibility to recover the lost catalytic activity
by removing adsorbed water. Further analysis of the DR spectra
(Fig. 7) showed that the blue shift in the absorption maxima upon
complete catalyst deactivation (by 5 week storage) was reversed
by lyophilization; the DR spectra of the lyophilized sample
matches the one corresponding to the freshly prepared catalyst.
Additionally, the catalytic activity of the lyophilized sample was
tested resulting in 18% conversion compared to the null percentage
before lyophilization indicating partial recovery of the catalytic
activity. Heating of the catalyst under vacuum was also considered
for water removal; however, complete catalyst deactivation was
maintained making lyophilization the best method for
reactivation.
2 5
O
conversion of approximately 11% upon a second use. A blue shift
in the diffuse reflectance spectrum was also observed when per-
forming the reaction using a conventional heat source (see
Fig. S9), which can also be interpreted as a possible decrease in
AuNP size and/or degree of aggregation, which may cause deacti-
vation of the catalytic sites.
Storage time dependence of the catalytic activity for plasmon-
mediated reactions was also evaluated. In the case of AuNP@Nb
2
-
O
5
HY340, the percent conversion of benzyl chloride decreased
from 100% for a freshly prepared catalyst to approximately 60%
and 13% for reactions performed utilizing the same catalyst batch
upon 1 week and 2 weeks aging, respectively (see Fig. 6). A slower
decay in activity was observed for AuNP@mesoNb
8% benzyl chloride conversion was still retained after five weeks
storage time (Fig. S8). The DR spectra of freshly prepared
AuNP@Nb catalyst and the same catalyst 2 weeks after prepara-
2 5
O for which
3
2
O
5
Discussion
tion remained virtually unchanged. However, for a fully deacti-
vated catalyst (5 week old) the DR spectrum showed significant
differences compared to its active counterpart.
Significant differences were observed between reactions per-
formed under irradiation, compared to those performed thermally
in the dark. In order for the reaction to overcome the activation
barrier and proceed thermally in the absence of irradiation the
temperature must reach 120 °C or higher. Nonetheless, high con-
versions were obtained under irradiation in spite that the bulk
temperature of the reaction mixture reached only 80 °C. Fasciani
and co-workers reported that the local temperature near the sur-
face of AuNPs reaches 500 °C when the surface plasmon band is
excited. These findings suggest that upon irradiation the localized
1
3
plasmon heating effect near the active sites of the AuNP@Nb
materials reach temperatures high enough to overcome the activa-
2 5
tion barrier. The use of light activated AuNP supported on Nb O
2 5
O
materials constitutes a major advantage due to its ability to per-
form transformations at lower temperatures and without the need
of an external heat source.
The proposed mechanism for these transformations is an aro-
matic electrophilic substitution were the Lewis acid sites would
facilitate the formation of a carbocation by reacting with benzyl
chloride. Then, the aromatic system of anisole –present in excess–
would attack the carbocation forming the final products.
Different strategies have been proposed to synthesize metal
nanoparticles supported on metal oxide materials and how stabi-
lizing the nanoparticles can control their corresponding catalytic
activities. Most of the literature is focused on size stabilization
and better thermal stability, which has been achieved by modify-
ing the environment of gold AuNPs, i.e., by strengthening metal–
support interactions and/or designing stabilizers that encapsulate
Fig. 6. % Conversion of benzyl chloride with AuNP@Nb
storage time and reaction under green LED irradiation: freshly made catalyst
AuNP) after 1 week (AuNP 1w) and 2 weeks (AuNP 2w).
2 5
O HY 340 catalyst with
(
1
9,31–33
them. Nevertheless, in the case of AuNPs supported on
2 5
Nb O materials, the efficiency of the plasmon heating effect was
achieved by deposition of unprotected AuNPs to avoid possible
secondary reactions.
In 2011, Nakashima et al., proposed the use of hydrated niobic
acid (Nb
2
O
5
ÁnH
2
O) as heterogeneous Lewis acid catalyst proving
-H
that in spite the presence of water molecules the NbO
4
2
O
1
0
adducts formed could still behave as active Lewis sites. However,
the changes observed in the DR spectra on Fig. 7 accompanied by
the decrease in catalytic activity with storage time, and the Lewis
acid affinity for water molecules proposed by Nakashima, suggest
that the deactivation of the catalysts with storage time could be
due to the adsorption of water molecules, making the catalyst less
active when used after long storage periods. This hypothesis would
2 5
also explain the fact that the AuNP@mesoNb O catalyst remained
active for longer storage times (>5 weeks) – compared to its amor-
phous analogue shown in Fig. S8 – since the acid sites located
inside the channels of the mesoporous materials are less accessible
to water.
Fig. 7. Diffuse reflectance spectra of AuNP@Nb
) 5 weeks old storage time previous lyophilization and ( ) 5 weeks storage time
after lyophilization.
2 5
O (HY340). ( ) Freshly prepared,
(