Selectivity Control in Photocatalytic Valorization of Biomass-Derived Platform Compounds by Surface Engineering of Titanium Oxide

Full text of DOI:10.1016/j.chempr.2020.08.014

ll
Article
Selectivity Control in Photocatalytic
Valorization of Biomass-Derived Platform
Compounds by Surface Engineering of
Titanium Oxide
Xuejiao Wu, Jieqiong Li, Shunji
Xie, ..., Qinghong Zhang, Jun
Cheng, Ye Wang
zhangqh@xmu.edu.cn (Q.Z.)
chengjun@xmu.edu.cn (J.C.)
wangye@xmu.edu.cn (Y.W.)
HIGHLIGHTS
Selectivity for the reduction of
bioplatforms depends on
exposed facets of TiO2
Coupling products as fuel
precursors are produced in high
yields
Surface-oxygen vacancies play
pivotal roles in controlling the
selectivity
Density of oxygen vacancies
governs electronic structure of
adsorbed species
Wang and colleagues found that the selectivity for photocatalytic transformations
of lignocellulose-derived platform chemicals including furfural, methyl furfural,
and vanillin depended strongly on the exposed facet of TiO2. They demonstrated
that the facet-dependent density of oxygen vacancies governs the charge
distribution and adsorption strength of surface species, and thus determines
product selectivity. Hydrogenation products such as fine chemicals and coupling
products as biofuel precursors can be produced in high yields over oxygen-
vacancy-rich and oxygen-vacancy-free TiO2 surfaces, respectively.
Wu et al., Chem 6, 1–16
November 5, 2020 ª 2020 Elsevier Inc.
https://doi.org/10.1016/j.chempr.2020.08.014

Please cite this article in press as: Wu et al., Selectivity Control in Photocatalytic Valorization of Biomass-Derived Platform Compounds by Surface
Engineering of Titanium Oxide, Chem (2020), https://doi.org/10.1016/j.chempr.2020.08.014
ll
Article
Selectivity Control in Photocatalytic Valorization
of Biomass-Derived Platform Compounds
by Surface Engineering of Titanium Oxide
Xuejiao Wu,1,2 Jieqiong Li,1,2 Shunji Xie,1,2 Pengbo Duan,1 Haikun Zhang,1 Jun Feng,1
1,
1,3,
Qinghong Zhang,1, Jun Cheng, and Ye Wang
*
*
*
SUMMARY
The Bigger Picture
Photocatalysis has emerged as a
useful approach to the sustainable
production of value-added
Photocatalysis has offered a promising opportunity for selective
transformation of biomass to high-value chemicals or fuels under
mild conditions. Whereas titanium oxide has been widely used for
photocatalytic pollutant degradation, H2 evolution, and CO2 reduc-
tion, few studies have been devoted to TiO2-based photocatalytic
valorization of biomass or biomass-derived platform compounds.
Here, we report on surface-controlled photocatalysis of TiO2 for
selective valorization of furfurals and vanillin that are lignocellu-
lose-derived key platform compounds. The reaction can be switched
from hydrogenation of aldehyde group to C–C coupling by manipu-
lating exposed facets; furanic and aromatic alcohols or coupling
products, which are fine chemicals or jet-fuel precursors, could be
produced with high selectivity. Our studies elucidate that the
facet-dependent density of oxygen vacancies governs the charge
distribution and adsorption strength of surface species and thus
controls product selectivity. The present work offers an example
of selectivity control by engineering TiO2 surfaces for valorization
of biomass-derived feedstocks.
products from biomass, in which
selectivity is the key issue because
a number of transformations are
possible when multifunctional
biochemicals are the reactants. As
one of the most popular
photocatalysts, TiO2 has mainly
been applied to reactions like
pollutant degradation that
concern more about activity than
selectivity. The knowledge about
selectivity-controlling principles
for TiO2 is limited. Here, we report
the first illustration of TiO2-based
photocatalysis that succeeds in
modulating product selectivity in
bioplatform transformations. Fine
chemicals or jet-fuel precursors
have been produced in high yields
by controlling the density of
oxygen vacancies on TiO2, which
governs surface adsorption and
reaction and thus determines the
product selectivity. This work
offers useful insights into the
design of selective photocatalysts
for biomass valorization by surface
engineering.
INTRODUCTION
As one of the most popular photocatalysts, titanium dioxide (TiO2) has triggered
broad interest for decades and is still the focus of current intensive studies.1 Besides
strategies such as doping, sensitization, and heterojunction construction to facilitate
the light absorption and charge-carrier separation, much recent attention has been
paid to TiO2 surface engineering, because the interfacial charge-transfer process,
the substrate adsorption, activation, and reactions are dominated by surface prop-
erties.2–4 For examples, anatase TiO2 with high-energy {001} facet mainly exposed
was successfully fabricated,5 and the high-energy surface was found to favor the ac-
tivity in a number of photocatalytic reactions.3,6 However, so far only limited studies
have been devoted to surface-dependent selectivity control in TiO2 photocatalysis.7
This is partially because TiO2 has mainly been studied for pollutant degradation, H2
evolution, and CO2 reduction,1–4,6–8 and most of the studies in these fields have
been concerned more about activity than selectivity.
Recently, photocatalysis has shown great potential in the valorization of biomass or
biomass-derived platform compounds (bioplatforms).9,10 Thermocatalysis for
biomass transformations usually requires harsh reaction conditions such as elevated
temperatures and high pressures, which inevitably lead to undesirable side reactions
and low selectivity of target products. In contrast, photocatalysis can be performed
Chem 6, 1–16, November 5, 2020 ª 2020 Elsevier Inc.
1

Please cite this article in press as: Wu et al., Selectivity Control in Photocatalytic Valorization of Biomass-Derived Platform Compounds by Surface
Engineering of Titanium Oxide, Chem (2020), https://doi.org/10.1016/j.chempr.2020.08.014
ll
Article
under mild conditions and thus is capable of avoiding side reactions. Further, photo-
catalysis has the potential to accomplish reactions that are difficult to realize by ther-
mocatalysis owing to unique reaction mechanisms induced by photogenerated elec-
trons and holes.11 A few recent studies have utilized TiO2 for photocatalytic
oxidation of monosaccharides into the corresponding carboxylates or C2–C5 oxy-
genates.12,13 However, the studies on TiO2-catalyzed reductive transformations of
bioplatforms, which play crucial roles in biomass valorization, are very scarce.9,10
Therefore, the development of new photocatalytic systems for reductive valorization
of bioplatforms using the cheap, stable, and environmental friendly TiO2 would offer
promising routes for sustainable production of high-value biobased chemicals and
fuels under mild conditions.
Selectivity control is one of the most important issues in the valorization of bio-
platforms because of multiple functional groups in bioplatform molecules and
many possible reaction channels to different products. For example, furfural, methyl
furfural, and vanillin are key feedstocks easily produced from lignocellulosic biomass
(Scheme S1),14–16 and the reductive transformations of these bioplatform molecules
can offer both furanic and aromatic primary alcohols, which are important fine chem-
icals in the fragrance and resin industries, as well as C–C coupling products with 10–
18 carbons, which are potential precursors for high-quality diesel or jet fuel (Fig-
ure 1).17 In addition, reactions such as hydrogenation of furanic/aromatic rings or de-
carbonylation may also take place under thermocatalytic conditions. It is noteworthy
that the selective production of fuel precursors by homo-coupling or cross-coupling
of bioplatforms is particularly difficult by thermocatalysis and only very limited suc-
cess has been achieved in spite of the significance of this type of reactions.18,19
Reductive coupling of bioplatforms with high selectivity remains one of the most
challenging goals in biomass valorization.
Here, we report that TiO2 nanocrystals efficiently catalyze photoreductive transfor-
mations of furfurals and vanillin, two types of important bioplatforms, into corre-
sponding alcohols and coupling products under mild conditions. We discovered
that the selectivity toward hydrogenation of aldehyde group to alcohols or reductive
coupling to coupling products can be easily controlled by engineering exposed fac-
ets of TiO2 nanocrystals. We uncovered that the facet-controlled selectivity is deter-
mined by the density of surface-oxygen vacancies generated in situ. Different reac-
tion mechanisms in the presence and absence of surface-oxygen vacancies have
been elucidated on the molecular level. To the best of our knowledge, this is the first
example of success in modulating product selectivity in photocatalytic reductive
transformations of bioplatforms and in the achievement of high yields of coupling
products. The present work would inspire the design of photocatalytic systems for
biomass valorization with controlled reaction pathways and product selectivity
through surface engineering of semiconductor catalysts.
1State Key Laboratory of Physical Chemistry of
Solid Surfaces, Collaborative Innovation Center
of Chemistry for Energy Materials, National
RESULTS
Engineering Laboratory for Green Chemical
Productions of Alcohols, Ethers and Esters,
College of Chemistry and Chemical Engineering,
Furfural Conversion over TiO2 with Different Facets
We fabricated three types of TiO2 nanocrystals with different morphologies and
Xiamen University, Xiamen 361005, China
exposed facets (Figure S1), including bipyramid-shaped anatase with {101} facets
2These authors contributed equally
mainly exposed (denoted as A-bipyramid), anatase nanosheet dominated by {001}
3Lead Contact
facets (denoted as A-sheet), and rod-shaped rutile enclosed mainly by {110} facets
*Correspondence: zhangqh@xmu.edu.cn (Q.Z.),
chengjun@xmu.edu.cn (J.C.),
most available anatase TiO2 nanocrystal,3 has been investigated in detail. No prod-
wangye@xmu.edu.cn (Y.W.)
(denoted as R-rod). Photocatalytic conversion of furfural over the A-bipyramid, the
uct was observed without light irradiation or in the absence of catalyst. Furfural
https://doi.org/10.1016/j.chempr.2020.08.014
2
Chem 6, 1–16, November 5, 2020

Please cite this article in press as: Wu et al., Selectivity Control in Photocatalytic Valorization of Biomass-Derived Platform Compounds by Surface
Engineering of Titanium Oxide, Chem (2020), https://doi.org/10.1016/j.chempr.2020.08.014
ll
Article
Figure 1. Transformations of Lignocellulose-Derived Platforms
Furfural, 5-methyl furfural, and vanillin can be derived from hemicellulose, cellulose, and lignin,
which are three major components of lignocellulosic biomass, respectively.
alcohol, a product by hydrogenation of the aldehyde group, as well as hydrofuroin
and furoin, which could be formed by reductive coupling of furfural (Figures 2A and
À1
S2), were formed over the A-bipyramid (Figure 2B). We adopted mmol
g
furanic ring
as the unit of the yield of each product, and it is noteworthy that one mole of hydro-
furoin or furoin contains two moles of furanic ring. Formaldehyde was also formed in
a considerable amount (Table S1; Figure S2F). The analysis of the reactions induced
by photogenerated electrons and holes suggests that furfural undergoes both alde-
hyde-group hydrogenation to furfural alcohol and reductive coupling to hydrofuroin
and furoin by photogenerated electrons on the A-bipyramid, whereas formaldehyde
is the product from methanol, the solvent, by photogenerated holes (Table S1;
Scheme S2). We found that the reaction completely stopped when the solvent
was changed from methanol to a proton-free solvent, e.g., acetonitrile. The inten-
tional addition of Na2S/Na2SO3 (4 mM Na2S + 2 mM Na2SO3) as sacrificial agents
to help the consumption of holes could not cause the formation of reduction prod-
ucts with acetonitrile solvent. These findings indicate that methanol also functions as
the proton source in combination with photogenerated electrons for the reductive
transformation of furfural (Scheme S2). Using methanol, a cheap solvent that can
be derived from biomass,20 as the hydrogen donor can avoid the use of high-pres-
sure H2 or corrosive formic acid, rendering the current process safe and environmen-
tally friendly.21,22
The selectivity of products strongly depended on the exposed facet. Unlike the A-
bipyramid, the A-sheet was highly selective toward the hydrogenation of aldehyde
group, offering furfural alcohol as the dominant product with a selectivity reaching
90% after a 4-h reaction (Figure 2C). On the other hand, the R-rod was very specific
for the formation of coupling products, i.e., hydrofuroin and furoin, and the forma-
tion rate of coupling products reached 32 mmolfuranic
gÀ1 hÀ1 with nearly
ring
100% selectivity (Figure 2D).This rate is more than twice that reported over a Ru-
doped ZnIn2S4 photocatalyst, which showed a formation rate of 13 mmolfuranic ring
g
À1hÀ1 for coupling products in the photocatalytic conversion of methylfurans that
could be produced from furfurals by further hydrodeoxygenation.23 Further, as
compared with some typical thermocatalytic systems for furfural reduction (Table
Chem 6, 1–16, November 5, 2020
3

Please cite this article in press as: Wu et al., Selectivity Control in Photocatalytic Valorization of Biomass-Derived Platform Compounds by Surface
Engineering of Titanium Oxide, Chem (2020), https://doi.org/10.1016/j.chempr.2020.08.014
ll
Article
Figure 2. Photocatalytic Conversion of Furfural over TiO2 Nanocrystals
(A) Structures of products from hydrogenation of aldehyde group and reductive coupling of furfural.
(B) Changes of product yield and selectivity with time over A-bipyramid.
(C) Changes of product yield and selectivity with time over A-sheet.
(D) Changes of product yield and selectivity with time over R-rod. The yields of products is presented in mmolfuranic ring gcatalystÀ1. Note that 1 mole of
hydrofuroin or furoin contains 2 moles of furanic ring. The inset models show the morphologies of TiO2 with the mainly exposed facet.
S2), the present TiO2-based photocatalytic system showed better activity under mild
conditions. Moreover, the by-products that are usually formed in thermocatalysis
such as tetrahydrofurfuryl alcohol by hydrogenation of furanic rings and furan by
decomposition of aldehyde groups have not been observed in our system. The
mild reaction conditions may contribute to the perseveration of valuable functional
groups. Therefore, the present simple TiO2-based photocatalytic system is very
promising for valorization of furfural, and in particular, not only furfural alcohol but
also reductive coupling products can be formed with high selectivity and a high for-
mation rate.
It is noteworthy that while the yield of coupling products increased almost linearly
with reaction time from the beginning (Figure 2D), the formation of furfural alcohol
proceeded quite slowly in the initial stage (< 0.5 h) and then accelerated after an in-
duction period over both the A-bipyramid and A-sheet (Figures 2B and 2C). The se-
lectivities of furfural alcohol were 4.6% and 63% at a reaction time of 0.5 h for A-
bipyramid and A-sheet, respectively, and they increased significantly to 35% and
87% after 3 h of reaction (Figures 2B and 2C). The existence of induction period
for the formation of furfural alcohol suggests the evolution of surface structures of
the A-bipyramid and A-sheet during the reaction.
Characterizations and Structure-Performance Relationship
The band-gap energies of A-sheet, A-bipyramid, and R-rod before reaction esti-
mated from UV-vis diffuse reflectance spectra (Figure 3A) were 3.24, 3.22, and
3.03 eV, respectively. These values are consistent with the band-gap energies of
anatase (~3.2 eV) and rutile (~3.0 eV) TiO2.1 The UV-vis spectra for the A-sheet
and A-bipyramid catalysts after photocatalytic reactions showed shifts of absorption
edges to longer wavelengths corresponding to band-gap narrowing and additional
absorption tails extending to the visible-light region (Figure 3A). This observation
corresponded well to the change in color from white to grayish blue for these two
catalysts after reaction.24,25 The band-gap narrowing and UV-vis spectrum-tailing
phenomena may be contributed by the generation of oxygen vacancies (VO)26 or
4
Chem 6, 1–16, November 5, 2020

Please cite this article in press as: Wu et al., Selectivity Control in Photocatalytic Valorization of Biomass-Derived Platform Compounds by Surface
Engineering of Titanium Oxide, Chem (2020), https://doi.org/10.1016/j.chempr.2020.08.014
ll
Article
Figure 3. Characterizations of TiO2 Nanocrystals
(A) UV-visible diffuse reflectance spectra. The inset pictures show the color of TiO2 after
photocatalytic reaction.
(B) Low-temperature EPR spectra measured at 77 K.
(C) Ti 2p XPS spectra for A-sheet.
(D) Ti 2p XPS spectra for A-bipyramid.
(E) Ti 2p XPS spectra for R-rod.
(F) Consumption of thionine versus photo-irradiation time.
(G) Density of oxygen vacancies.
structural disorders27 on the A-sheet and A-bipyramid. The irradiation of TiO2 in
methanol, the solvent, may create surface VO sites (Scheme S3).28 In contrast, the
UV-vis spectrum and color of the R-rod did not undergo significant changes after
Chem 6, 1–16, November 5, 2020
5

Please cite this article in press as: Wu et al., Selectivity Control in Photocatalytic Valorization of Biomass-Derived Platform Compounds by Surface
Engineering of Titanium Oxide, Chem (2020), https://doi.org/10.1016/j.chempr.2020.08.014
ll
Article
the photocatalytic reaction. Electron paramagnetic resonance (EPR) and X-ray
photoelectron spectroscopy (XPS) measurements provided further evidence for
the presence of VO on the A-sheet and A-bipyramid and the absence of VO on the
R-rod after reaction. EPR measurements performed at 77 K showed a signal at g
value of 1.98, which could be ascribed to Ti3+ in TiO2,29 for the A-sheet and A-bipyr-
amid after photocatalytic reaction, whereas this signal was insignificant for the used
R-rod catalyst (Figure 3B). It is widely accepted that Ti3+ is a charge-compensation
species accompanying the appearance of VO sites on TiO2.30 While not distinct, a
weak signal at g = 2.01 also appeared. EPR measurements conducted at 298 K for
the catalysts after photocatalytic reactions showed a decrease in the intensity of
signal at g = 1.98, which is assignable to Ti3+, whereas the signal at g = 2.01 became
significantly stronger for the A-sheet and A-bipyramid (Figure S3). The change in EPR
spectra may result from the interaction of O2 with the surface Ti3+ and VO site, form-
ing OÀ on the surface at the expense of Ti3+ 31
The Ti 2p core-level spectra obtained
.
by XPS for the A-sheet or A-bipyramid before reaction were dominated by peaks at
458.6 and 464.4 eV, assignable to Ti 2p3/2 and Ti 2p1/2 of Ti4+, whereas two shoul-
ders at 457.4 and 463.1 eV, attributable to those of Ti3+ accompanying the surface
VO sites, appeared after reactions (Figures 3C and 3D).32,33 On the other hand, no
significant differences in Ti 2p spectra for the R-rod catalyst before and after reac-
tions were observed (Figure 3E). The O 1s XPS results for all the samples exhibited
two peaks; the peak at 529.7 eV could be ascribed to the lattice oxygen, while that at
531.5 eV was assignable to the Ti-OH species or defective oxygen species associ-
ated with VO sites.32,34 No significant differences were observed in the O 1s spectra
before and after reactions (Figure S4). This is possibly because the VO density is low
and the peak at 531.5 eV may mainly be contributed by the Ti-OH species. There-
fore, these characterization results demonstrate the generation of VO sites on the
A-sheet and A-bipyramid but not on the R-rod during the photocatalysis.
We have measured the surface VO sites quantitatively by electron titration with thi-
onine, the consumption of which corresponds to the amount of electrons trapped
in VO sites.35 While the consumption of thionine over the R-rod irradiated with
different times was negligible, considerable amounts of thionine were consumed
over the A-sheet and A-bipyramid (Figure 3F). The consumption of thionine was
faster over the A-sheet than over the A-bipyramid in the initial stage, suggesting
that the VO site was easier to form on the A-sheet under light irradiation. The density
of VO sites evaluated from the consumption of thionine decreased in the order of A-
sheet (145 mmol gÀ1) > A-bipyramid (61.5 mmol gÀ1) > R-rod (2.52 mmol gÀ1
)
(Figure 3G).
In short, our characterizations suggest that the VO can be generated on catalyst sur-
faces during photocatalytic conversion of furfural in methanol and the density of VO
sites depends on the exposed facets, decreasing in the order of anatase {001} >
anatase {101} > rutile {110}. This trend is in agreement with the surface energy of
these facts.2–4 As displayed in Scheme S3, we propose that the VO site is generated
through the stoichiometric reaction between methanol and TiO2. It should be noted
that the density of VO sites and the amount of methanol consumed for the genera-
tion of VO sites are remarkably lower than the amounts of furfural alcohol formed
in the hydrogenation of aldehyde group of furfurals. The yields of furfural alcohol
were 72 and 152 mmol gÀ1 over the A-bipyramid and A-sheet with VO densities of
61.5 and 145 mmol gÀ1 during the photocatalytic conversion of furfural for 3 and 4
h, respectively (Figures 2B and 2C). The turnover numbers for furfural alcohol forma-
tion per VO site are both >1,000 over the A-bipyramid and A-sheet. Therefore, the
reduction of furfural to furfural alcohol proceeds catalytically over the VO site. We
6
Chem 6, 1–16, November 5, 2020

Please cite this article in press as: Wu et al., Selectivity Control in Photocatalytic Valorization of Biomass-Derived Platform Compounds by Surface
Engineering of Titanium Oxide, Chem (2020), https://doi.org/10.1016/j.chempr.2020.08.014
ll
Article
further confirmed that the catalyst could be used repeatedly. Recycling tests of the
A-bipyramid for photocatalytic conversion of furfural showed only a slight change in
performance in the second cycle, and the performance could almost be sustained in
the third and fourth cycles (Figure S5).
The structure-performance correlation suggests that the surface VO site plays a
crucial role in aldehyde-group hydrogenation, whereas the VO-free TiO2 surface is
responsible for the coupling reaction. It is noteworthy that the density of VO sites
mainly determines the selectivity, and the activity may be influenced by many factors
such as the light absorption, the ability of photogenerated electron-hole separation
and the surface reaction. For the series of A-bipyramid, A-sheet, and R-rod catalysts,
the light absorption (Figure 3A) and the ability of electron-hole separation as re-
flected by the photocurrent density (Figure S6) do not have big differences. We
found that the specific surface areas of this series of catalysts, which may influence
the catalytic performance, are quite different (Table S3). We speculate that the larger
surface area of the A-bipyramid results in its higher yield of coupling products as
compared with the R-rod (Figure 2), although the selectivity of coupling products
is lower over the A-bipyramid than that over the R-rod.
We further studied the commercial TiO2 (Degussa P25) for the photocatalytic con-
version of furfural. Furfural alcohol and coupling products (hydrofuroin and furoin)
were formed with selectivities of 19% and 81%, respectively (Figure S7). The surface
V
O density measured by electron titration was 34 mmol gÀ1 for P25, and thus the VO
density increased in the sequence of R-rod < P25 < A-bipyramid < A-sheet. It is of
interest that both the selectivity and the yield of furfural alcohol change in the
same order (Figure S7; Table S3). On the other hand, the selectivity of coupling
products decreased in the sequence of R-rod > P25 > A-bipyramid > A-sheet, con-
firming that the VO-free TiO2 surface plays a crucial role in the formation of coupling
products. However, the yield of coupling products based on specific surface area
over P25 was relatively lower than that over the A-bipyramid. This is probably
because of the lower ability of P25 in electron-hole separation (Figure S6). We further
prepared P25-based TiO2 samples with increased VO densities by treatment with
NaBH4, a strong reductant that is known to be capable of creating surface VO sites
on TiO2 under mild conditions.36 The catalysts denoted as P25-0.5, P25-0.75, and
P25-1, which had almost the same specific surface area (51–55 m2 À1), were ob-
g
tained by treating P25 with different concentrations of NaBH4 aqueous solutions.
It is of interest that both the furfural alcohol selectivity (Figure S7B) and the surface
VO density (Figure S7C) increased with the concentration of NaBH4. This result pro-
vides further evidence that the VO site favors the hydrogenation of aldehyde group
of furfural to furfural alcohol. However, the yield of total products per either catalyst
weight or surface area for the P25-1 catalyst with the highest VO density rather
decreased (Figure S7D). This is possibly because of the decreased electron-hole
separation ability of this deeply reduced catalyst (Figure S7E).
DFT Calculations and Roles of Oxygen Vacancies
To understand the role of VO sites in determining the reaction channel, we per-
formed density functional theory (DFT) calculations for the conversion of furfural
on catalyst surfaces with and without VO. The optimized geometries indicate a tilted
h1-(O) configuration of furfural adsorbed on both VO-free and VO-rich surfaces (Fig-
ures S8 and S9). As compared with other adsorption modes, the h1-(O) adsorption
mode, with the carbonyl oxygen bound to Ti cation and the furan ring repulsed
away from the surface, is beneficial to the selective activation of aldehyde group.
Chem 6, 1–16, November 5, 2020
7

Please cite this article in press as: Wu et al., Selectivity Control in Photocatalytic Valorization of Biomass-Derived Platform Compounds by Surface
Engineering of Titanium Oxide, Chem (2020), https://doi.org/10.1016/j.chempr.2020.08.014
ll
Article
Figure 4. DFT Calculations for Furfural Conversion on TiO2
(A) Reaction pathways of hydrogenation of aldehyde group and reductive coupling.
(B) Total-energy-change profiles for furfural reduction on VO-free anatase {101} surface.
(C) Total-energy-change profiles for furfural reduction on VO-rich anatase {101} surface.
For example, the h2-(C,O) configuration would cause the decarboxylation or hydro-
genation of the furan ring.37
Generally, for photocatalytic reduction, electron and proton transfer may occur not
only through stepwise proton transfer followed by electron transfer (PT-ET)38 or elec-
tron transfer followed by proton transfer (ET-PT)39 but also through a concerted pro-
ton-electron transfer (CPET)40–42 manner (Figure S10). We have studied the reaction
mechanism with a total-energy-change approach (the energy difference between
product and reactant), regardless of the detailed proton-electron transfer process.
Therefore, the above transfer mechanisms are uniformly expressed as (eÀ + H+)
transfer hereafter. Two possible reduction pathways exist for the hydrogenation of
adsorbed furfural in the first step (Figure 4A): (1) the addition of (eÀ + H+) to the O
atom of the carbonyl group, resulting in the formation of a C radical ($C–OH) (I/
II); and (2) the addition of (eÀ + H+) to the C atom of the carbonyl group, leading
to an O radical (CH–O$) (I/IV). In the second step, the $C–OH intermediate may
either be coupled with a second C radical to produce hydrofuroin (II/III) or be
further reduced by the addition of (eÀ + H+) to form furfural alcohol (II/V). The
8
Chem 6, 1–16, November 5, 2020

Please cite this article in press as: Wu et al., Selectivity Control in Photocatalytic Valorization of Biomass-Derived Platform Compounds by Surface
Engineering of Titanium Oxide, Chem (2020), https://doi.org/10.1016/j.chempr.2020.08.014
ll
Article
CH–O$ intermediate can only result in furfural alcohol by addition of another (eÀ
H+) (IV/V).
+
The DFT calculation reveals that on the VO-free anatase {101} surface, the hydroge-
nation energy for the attack of (eÀ + H+) to O atoms of carbonyl to form the $C-OH
intermediate (I/II) is 0.61 eV, which is more favorable by 0.54 eV than the hydroge-
nation through the attack of (eÀ + H+) to the C atom of carbonyl (Figure 4B). The $C–
OH subsequently undergoes coupling with an energy decrease of 2.06 eV (II/III),
which is more favorable than the hydrogenation by 0.8 eV. In contrast, on the VO-
rich anatase {101} surface, the formation of CH–O$ (I/IV) is an exothermic reaction,
whereas the formation of $C–OH (I/II) is endothermic (Figure 4C), and thus, furfural
prefers to undergo a two-step hydrogenation mechanism to furfural alcohol. Similar
conclusions have been obtained on the VO-free and VO-rich anatase {001} (Fig-
ure S11) as well as VO-free and VO-rich rutile {110} surfaces (Figure S12). Therefore,
our DFT calculations on the energetics of reaction pathways have confirmed the
experimental findings that the presence of VO on TiO2 surfaces favors the hydroge-
nation of aldehyde group in furfural, whereas coupling products are formed prefer-
entially on VO-free surfaces.
Further DFT calculations were performed to gain in-depth insights into the differ-
ence in selectivity on the VO-free and VO-rich TiO2 surfaces. For the VO-free surface,
the charge-difference analysis reveals that there is no significant charge transfer be-
tween TiO2 and furfural (green panel in Figure 5A). Instead, a small amount of charge
rearrangement occurs inside furfural molecules (Figure S13A). Because of the higher
electronegativity of O, the O and C atoms of carbonyl are negatively and positively
charged, respectively (green panel in Figure 5B). Thus, the O atom of carbonyl is
more favorably hydrogenated, leading to the $C–OH intermediate. On the other
hand, when furfural is adsorbed on the VO-rich TiO2 surface via the filling of the
VO site by the O atom of carbonyl, a significant charge transfer occurs from the
two Ti3+ ions neighboring to the VO site to furfural (orange panel in Figure 5A; Fig-
ure S13B). The charge redistribution may induce negative charges around the C
atom of carbonyl group (orange panel in Figure 5B). Moreover, the embedded
carbonyl O atom is coordinatively saturated by Ti, and consequently the (eÀ + H+)
is preferably bound to the C atom of carbonyl group to form the CH–O$ intermedi-
ate (orange panel in Figure 5B; Table S4). Furthermore, the calculation reveals that
the C radical formed on the VO-free surface has a much smaller adsorption energy
than the O radical formed on the VO-rich surface (Figure 5C; Table S5). Thus, the
C radical may readily desorb from the VO-free TiO2 surface and dimerizes to form
the coupling products, whereas the strongly bound O radical on the VO-rich surface
preferably undergoes consecutive (eÀ + H+) addition to yield furfural alcohol.
Therefore, our calculations indicate that the product selectivity in furfural conversion
on the VO-free and VO-rich surfaces is determined by the interaction between the
reactant and the surface. The electron transfer between the TiO2 surface and furfural
molecule controls the preferential attack of (eÀ + H+) onto C or O atom in a carbonyl
group, leading to the formation of O or C radical intermediate and different types of
products. The strength of adsorption of reaction intermediate on TiO2 surfaces also
affects the product selectivity.
Experiments were performed to probe the interaction of the reactant with the VO-
rich and VO-free surfaces. Because rich VO and almost no VO sites were generated
in situ on the A-sheet and R-rod, respectively, these two catalysts were chosen to
represent the VO-rich and VO-free surfaces in our experiments. The transient
Chem 6, 1–16, November 5, 2020
9

Please cite this article in press as: Wu et al., Selectivity Control in Photocatalytic Valorization of Biomass-Derived Platform Compounds by Surface
Engineering of Titanium Oxide, Chem (2020), https://doi.org/10.1016/j.chempr.2020.08.014
ll
Article
Figure 5. Interactions between Furfural and TiO2 Surfaces
(A) Electron density changes (Dre) on TiO2 before and after furfural adsorption.
(B) Hirshfeld charge distributions in carbonyl C and O atoms. Right panel: preferential attack behavior of H atom onto carbonyl C or O atom.
(C) Adsorption energies (Ead) of C radical on VO-free surfaces and O radical on VO-rich surfaces.
(D) Transient photocurrent responses of R-rod and A-sheet before and after furfural (FF) injection.
(E) Schematic setup for ATR-FTIR spectroscopy. N2 is the protecting gas to prevent ATR-FTIR measurements from the influence of air.
(F) Time-resolved ATR-IR spectra of furfural on R-rod.
(G) Time-resolved ATR-IR spectra of furfural on A-sheet. The VO-free and VO-rich surfaces in (A), (B), and (C) include anatase {101}, anatase {001} and
rutile {110} surfaces. The chemical structures and bonding length of carbonyl group on VO-free surface of rutile {110} and VO-rich surface of anatase
{001} are displayed on the right panels of (F) and (G).
photocurrent response study was performed in methanol, i.e., the solvent, and the
electrode with photocatalyst was illuminated under UV light for 1 h before measure-
ment. This pretreatment generated VO sites with saturated concentrations on the A-
sheet and R-rod surfaces. The result for the R-rod showed a slight difference in the
presence and absence of furfural, suggesting a weak interaction between furfural
and the VO-free surface (Figure 5D). In contrast, a big decrease in the photocurrent
density was observed after the injection of furfural onto the A-sheet, indicating that
the interfacial electron transfer induced a strong interaction between the VO-rich
10
Chem 6, 1–16, November 5, 2020

Please cite this article in press as: Wu et al., Selectivity Control in Photocatalytic Valorization of Biomass-Derived Platform Compounds by Surface
Engineering of Titanium Oxide, Chem (2020), https://doi.org/10.1016/j.chempr.2020.08.014
ll
Article
surface and furfural.43 Transient photocurrent response measurements were also
performed by using CH3CN as the electrolyte without pretreatment of the photoca-
talyst, and thus the VO density on catalyst surfaces would be very low. The result
showed that the decrease in photocurrent density after the injection of furfural
was not big not only for the R-rod but also for the A-sheet catalyst (Figure S14).
The phenomenon that the photocurrent density only decreases slightly after furfural
injection in the case of CH3CN is similar to that for the R-rod with catalyst pretreat-
ment in methanol (Figure 5D). This agrees with the finding that the density of VO sites
was low on the R-rod catalyst under irradiation in methanol.
Attenuated total reflection Fourier-transform infrared (ATR-FTIR) spectroscopy (Fig-
ure 5E), which could provide liquid-phase IR spectra of molecules on solid catalyst
surfaces,44 typically provides IR bands of furfural in a wavenumber region of
1,500–1,200 cmÀ1 (Table S6).45,46 The thermogravimetric-differential thermal anal-
ysis (TG-DTA) showed that no organic adsorbates were present on catalyst surfaces
before adsorption (Figure S15). Our time-resolved ATR-FTIR spectra recorded for
the furfural-adsorbed R-rod and A-sheet both exhibited characteristic signals assign-
able to furfural (Figures 5F and 5G). Furthermore, an additional band centered at
around 1,241 cmÀ1 was observed for the A-sheet, which could not be observed
for the R-rod (Figures 5F and 5G). This new IR band could be assigned to the red-shift
of carbonyl group caused by the strong interaction with the VO-rich surface.47 Our
DFT calculations further revealed that for furfural on the VO-free rutile {110} surface,
˚
the bond length of C=O (1.26 A) had negligible change with respect to that in the
˚
gas phase (1.23 A). However, on the VO-rich anatase {001} surface, the C=O bond
˚
length for adsorbed furfural increased to 1.32 A (Table S7), indicative of the change
of C=O to C–O bond. The new IR band of adsorbed furfural on the VO-rich surface
may arise from the C–O stretching vibration for furfuryl-oxy adsorption state (Fig-
ure 5G). The appearance of this band further confirms the strong interaction and
electron transfer between furfural and the VO-rich surface.
The addition of a small amount of H2O2, a radical scavenger, into the system signif-
icantly suppressed the conversion of furfural over the A-sheet and R-rod (Table S8),
confirming that both the hydrogenation and coupling reactions proceeded via
radical intermediates. The formation of $C–OH radical over the R-rod was further
verified by the addition of 1,1-diphenylethylene as a radical trap.48 A new compound
with a mass-to-charge ratio (m/z) of 276, which could be assigned to the adduct of
$C–OH radical and 1,1-diphenylethylene, was detected in the product mixture (Fig-
ure S16). The adduct of CH–O$ radical and 1,1-diphenylethylene, which should have
fingerprint peaks of m/z at 97, 167, 181, and 278, could not observed for the A-
sheet. This is probably because of the strong adsorption of CH–O$ on VO-rich sur-
faces (Figure 5C).
Valorization of Other Bioplatforms
Besides furfural, which can be easily obtained from hemicellulose, we have per-
formed photocatalytic conversions of other bioplatforms derived from lignocellu-
losic biomass, including 5-methyl furfural from cellulose and vanillin from lignin, us-
ing the A-sheet and R-rod, which represent VO-rich and VO-free TiO2 catalysts,
respectively. The A-sheet was highly selective toward the hydrogenation of alde-
hyde group in these bioplatforms to form the corresponding alcohols (Figure 6),
which are versatile fine chemicals. On the other hand, the R-rod exhibited unique
selectivity toward the formation of coupling products and the yield of the corre-
sponding coupling products reached R 80%. These results demonstrate the
Chem 6, 1–16, November 5, 2020
11

Please cite this article in press as: Wu et al., Selectivity Control in Photocatalytic Valorization of Biomass-Derived Platform Compounds by Surface
Engineering of Titanium Oxide, Chem (2020), https://doi.org/10.1016/j.chempr.2020.08.014
ll
Article
Figure 6. Photocatalytic Conversions of Bioplatforms
Furfural, 5-methyl furfural, and vanillin can be derived from hemicellulose, cellulose, and lignin, the
three major components of lignocellulosic biomass, respectively. The experiments in each case
were performed at least for three times. The error bar represents the absolute deviation, which is
within 10%.
feasibility of TiO2 surface engineering in tuning the selectivity in photocatalytic
transformation of lignocellulosic biomass-derived platforms.
Surface defects like VO sites, even in a low density, have the potential to regulate the
surface electronic structure of semiconductors and thus can tune adsorbed chemi-
cals and catalytic performance.49,50 The interaction between the functional groups
in reactants or intermediates and the catalyst surface is known to play a key role in
determining the selectivity in thermocatalytic transformations of multifunctional
chemicals.14 However, how the surface defect functions in photocatalytic valoriza-
tion of biomass is less understood. The present work contributes to presenting a first
example to illustrate that the surface engineering by tuning the density of oxygen
vacancies can successfully control the selectivity in photocatalytic transformation
of multifunctional bioplatforms.
Conclusion
We discovered that TiO2 is an efficient photocatalyst for the reductive transforma-
tion of lignocellulose-derived furfurals and vanillin and that surface-oxygen va-
cancies play a decisive role in determining the product selectivity. The catalyst
without surface-oxygen vacancies favors the reductive coupling reaction to form
12
Chem 6, 1–16, November 5, 2020

Please cite this article in press as: Wu et al., Selectivity Control in Photocatalytic Valorization of Biomass-Derived Platform Compounds by Surface
Engineering of Titanium Oxide, Chem (2020), https://doi.org/10.1016/j.chempr.2020.08.014
ll
Article
C10–C18 products, which are high-quality fuel precursors, whereas the catalyst with
rich surface-oxygen vacancies preferentially catalyzes the hydrogenation of alde-
hyde group to form aromatic and furanic alcohols as versatile fine chemicals. High
yields of the coupling products or aromatic and furanic alcohols have been achieved.
Our mechanistic studies revealed that the density of oxygen vacancies affects the
reactant-surface interaction and the charge transfer. The strong interaction between
the oxygen-vacancy-rich surface and furfural leads to the preferential formation of
CH–O$ intermediate, and the strong adsorption of intermediate facilitates the sub-
sequent addition of a second H atom to form furfural alcohol. On the other hand, the
$C–OH intermediate is formed on the oxygen-vacancy-free surface and the weak
adsorption of the radical intermediate enables easy desorption and the subsequent
C–C coupling. The present work not only demonstrates that surface-controlled TiO2
is a promising photocatalyst for biomass valorization but also offers an opportunity
to control the reaction pattern and product selectivity in transformation of multifunc-
tional chemicals.
EXPERIMENTAL PROCEDURES
Resource Availability
Lead Contact
Further information and requests for resources and reagents should be directed to
and will be fulfilled by the Lead Contact, Professor Ye Wang (wangye@xmu.edu.cn).
Materials Availability
This study did not generate new unique reagents.
Data and Code Availability
The published article includes all datasets generated or analyzed during this study.
Synthesis of TiO2 Nanocrystals with Controlled Facets
To synthesize anatase TiO2 with bipyramid morphology and {101} facet mainly
exposed, titanium isopropoxide (20 mL) was mixed with distilled water (2 mL) in a
Teflon autoclave. The mixture was then placed in an electric oven and held at
180ꢀC for 36 h. After cooling to room temperature, the sample was separated by
centrifugation, washed with distilled water and ethanol, dried overnight at 60ꢀC,
and calcined in air at 500ꢀC for 3 h.51 For the synthesis of anatase TiO2 sheets domi-
nated by {001} facet, tetrabutyl titanate (25 mL) was mixed with 47% hydrofluoric acid
solution (3 mL) and the mixture was kept at 180ꢀC for 24 h. After cooling to room
temperature, the sample was separated by centrifugation, washed with 1 M
NaOH, distilled water and ethanol, and dried overnight at 60ꢀC.52 For the synthesis
of rutile TiO2 rod enclosed mainly by {110} facet, tetrabutyl titanate (10 mL) was
mixed with distilled water (10 mL) and 38% hydrochloric acid (10 mL), and the
mixture was kept at 180ꢀC for 24 h. After cooling to room temperature, the sample
was separated by centrifugation, washed with 1 M NaOH, distilled water and
ethanol, dried overnight at 60ꢀC, and calcined in air at 500ꢀC for 3 h.53
Evaluation of Photocatalytic Performance
For the conversion of furfural, TiO2 (5 mg), furfural (100 mL, 1.2 mmol) and solvent
(CH3OH, 5 mL) were added into a quartz reactor (10 mL). The reactor was evacuated
and purged with N2 for 5 min. The reaction mixture was stirred at 650 rpm and irra-
diated under 300 W Xe lamp (l = 320–780 nm). During the reaction, water was circu-
lated to cool the reactor, keeping the temperature of the reactor at 40ꢀC. To ensure
the complete conversion of furfural in Figure 6, 50 mg of catalyst and 4 h of reaction
time were adopted. For the conversions of 5-methlyfurfual and villain, the reaction
Chem 6, 1–16, November 5, 2020
13

Please cite this article in press as: Wu et al., Selectivity Control in Photocatalytic Valorization of Biomass-Derived Platform Compounds by Surface
Engineering of Titanium Oxide, Chem (2020), https://doi.org/10.1016/j.chempr.2020.08.014
ll
Article
conditions were the same as those for the conversion of furfural in Figure 6 except for
using reactant of 0.5 mmol and reaction time of 12 h. The light intensity was 600 mW
cmÀ2 and the illumination area was 3.7 cm2. After reaction, the solid catalyst was
filtered and the products in the filtrate were identified by GC-MS and quantified
by GC-FID and HPLC.
DFT Calculation
All the geometries were optimized using DFT as implemented in the freely available
package CP2K/Quickstep.54 For the reduced TiO2 with excess electrons localizing at
the surface Ti site next to bridge OH group, the redox potentials of Ti4+/Ti3+ of the
localized trapped electronic state obtained by using Perdew-Burke-Ernzerhof (PBE)
density functional55 and hybrid functional HSE0656 are close.57 Therefore, to alle-
viate computation time, the PBE functional with the Grimme’s dispersion correc-
tion58 was adopted. The Goedecker-Teter-Hutter (GTH) pseudopotentials59 were
used to represent the core electrons, and the double-z basis functions with one
set of polarization functions (DZVP)60 were employed to valence electrons. The
atomic charges were calculated based on the Hirshfeld method.61 Hirshfeld charge
is considered to be the chemically meaningful charge62 and is widely used in the
charge-transfer analysis for adsorption system.63,64 Further information on the
computational setup in this work can be found in the Supplemental Information.
SUPPLEMENTAL INFORMATION
Supplemental Information can be found online at https://doi.org/10.1016/j.chempr.
2020.08.014.
ACKNOWLEDGMENTS
This work was supported by the National Natural Science Foundation of China (nos.
21690082 and 21972115).
AUTHOR CONTRIBUTIONS
X.W. and S.X. performed most of the experiments and analyzed the experimental
data; J.L. performed DFT computations and analyzed the computational data;
P.D. and H.Z. synthesized TiO2 semiconductors with different facets; J.F. performed
a part of characterizations; Q.Z. analyzed all the data and co-wrote the paper; J.C.
guided the computational work, analyzed all the data, and co-wrote the paper;
Y.W. designed and guided the study, and co-wrote the paper; all of the authors dis-
cussed the results and reviewed the manuscript.
DECLARATION OF INTERESTS
The authors declare no competing interests.
Received: February 18, 2020
Revised: April 10, 2020
Accepted: August 18, 2020
Published: September 25, 2020
REFERENCES
1. Schneider, J., Matsuoka, M., Takeuchi,
M., Zhang, J., Horiuchi, Y., Anpo, M.,
and Bahnemann, D.W. (2014).
2. Henderson, M.A.A. (2011). A surface science
perspective on TiO2 photocatalysis. Surf. Sci.
Rep. 66, 185–297.
semiconductor photocatalysts: motivations,
advances and unique properties. Chem.
Commun. (Camb.) 47, 6763–6783.
Understanding TiO2 photocatalysis:
mechanisms and materials. Chem. Rev.
114, 9919–9986.
3. Liu, G., Yu, J.C., Lu, G.Q., and Cheng, H.M.
(2011). Crystal facet engineering of
4. Jing, L., Zhou, W., Tian, G., and Fu, H. (2013).
Surface tuning for oxide-based nanomaterials
14
Chem 6, 1–16, November 5, 2020

Please cite this article in press as: Wu et al., Selectivity Control in Photocatalytic Valorization of Biomass-Derived Platform Compounds by Surface
Engineering of Titanium Oxide, Chem (2020), https://doi.org/10.1016/j.chempr.2020.08.014
ll
Article
as efficient photocatalysts. Chem. Soc. Rev. 42,
9509–9549.
catalytic conversion of g-valerolactone to liquid
alkenes for transportation fuels. Science 327,
1110–1114.
32. Zhang, X., Tian, H., Wang, X., Xue, G., Tian, Z.,
Zhang, J., Yuan, S., Yu, T., and Zou, Z. (2013).
The role of oxygen vacancy-Ti3+ states on TiO2
nanotubes’ surface in dye-sensitized solar cells.
Mater. Lett. 100, 51–53.
5. Yang, H.G., Sun, C.H., Qiao, S.Z., Zou, J., Liu,
G., Smith, S.C., Cheng, H.M., and Lu, G.Q.
(2008). Anatase TiO2 single crystals with a large
percentage of reactive facets. Nature 453,
638–641.
19. Gumidyala, A., Wang, B., and Crossley, S.
(2016). Direct carbon-carbon coupling of
furanics with acetic acid over brønsted zeolites.
Sci. Adv. 2, e1601072.
33. Fan, Y., Li, S., Bao, J., Shi, L., Yang, Y., Yu, F.,
Gao, P., Wang, H., Zhong, L., and Sun, Y. (2018).
Hydrofunctionalization of olefins to value-
added chemicals via photocatalytic coupling.
Green Chem. 20, 3450–3456.
6. Ong, W.J., Tan, L.L., Chai, S.P., Yong, S.T., and
Mohamed, A.R. (2014). Facet-dependent
photocatalytic properties of TiO2 -based
composites for energy conversion and
environmental remediation. ChemSusChem 7,
690–719.
20. Wang, M., Liu, M., Lu, J., and Wang, F. (2020).
Photo splitting of bio-polyols and sugars to
methanol and syngas. Nat. Commun. 11, 1083.
34. Razzaq, A., Sinhamahapatra, A., Kang, T.H.,
Grimes, C.A., Yu, J.S., and In, S.I. (2017).
Efficient solar light photoreduction of CO2 to
hydrocarbon fuels via magnesiothermally
reduced TiO2 photocatalyst. Appl. Catal. B 215,
28–35.
21. Zhang, J., and Chen, J. (2017). Selective
transfer hydrogenation of biomass-based
furfural and 5-hydroxymethylfurfural over
hydrotalcite-derived copper catalysts using
methanol as a hydrogen donor. ACS Sustain.
Chem. Eng. 5, 5982–5993.
7. Kou, J., Lu, C., Wang, J., Chen, Y., Xu, Z., and
Varma, R.S. (2017). Selectivity enhancement in
heterogeneous photocatalytic
transformations. Chem. Rev. 117, 1445–1514.
35. Keulemans, M., Verbruggen, S.W.,
Hauchecorne, B., Martens, J.A., and Lenaerts,
S. (2016). Activity versus selectivity in
photocatalysis: morphological or electronic
properties tipping the scale. J. Catal. 344,
221–228.
22. Zhang, Z., Wang, C., Gou, X., Chen, H., Chen,
K., Lu, X., Ouyang, P., and Fu, J. (2019).
Catalytic in-situ hydrogenation of 5-
8. Tu, W., Zhou, Y., and Zou, Z. (2014).
Photocatalytic conversion of CO2 into
renewable hydrocarbon fuels: state-of-the-art
accomplishment, challenges, and prospects.
Adv. Mater. 26, 4607–4626.
hydroxymethylfurfural to 2,5-dimethylfuran
over Cu-based catalysts with methanol as a
hydrogen donor. Appl. Catal. A 570, 245–250.
36. Tian, J., Hu, X., Yang, H., Zhou, Y., Cui, H., and
Liu, H. (2016). High yield production of reduced
TiO2 with enhanced photocatalytic activity.
Appl. Surf. Sci. 360, 738–743.
9. Colmenares, J.C., and Luque, R. (2014).
Heterogeneous photocatalytic nanomaterials:
prospects and challenges in selective
transformations of biomass-derived
23. Luo, N., Montini, T., Zhang, J., Fornasiero, P.,
Fonda, E., Hou, T., Nie, W., Lu, J., Liu, J.,
Heggen, M., et al. (2019). Visible-light-driven
coproduction of diesel precursors and
hydrogen from lignocellulose-derived
methylfurans. Nat. Energy 4, 575–584.
compounds. Chem. Soc. Rev. 43, 765–778.
37. Sitthisa, S., Sooknoi, T., Ma, Y., Balbuena, P.B.,
and Resasco, D.E. (2011). Kinetics and
10. Granone, L.I., Sieland, F., Zheng, N., Dillert, R.,
and Bahnemann, D.W. (2018). Photocatalytic
conversion of biomass into valuable products:
a meaningful approach? Green Chem. 20,
1169–1192.
mechanism of hydrogenation of furfural on Cu/
SiO2 catalysts. J. Catal. 277, 1–13.
24. Zhang, Y., Ding, Z., Foster, C.W., Banks, C.E.,
Qiu, X., and Ji, X. (2017). Oxygen vacancies
evoked blue TiO2(B) nanobelts with efficiency
enhancement in sodium storage behaviors.
Adv. Funct. Mater. 27, 1700856.
38. Lin, Y., Yan, Y., Peng, W., Qiao, X., Huang, D., Ji,
H., Chen, C., Ma, W., and Zhao, J. (2020).
Crucial effect of TiÀH species generated in the
visible-light-driven transformations: slowed-
down proton-coupled electron transfer.
J. Phys. Chem. Lett. 11, 3941–3946.
11. Wu, X., Fan, X., Xie, S., Lin, J., Cheng, J., Zhang,
Q., Chen, L., and Wang, Y. (2018). Solar energy-
driven lignin-first approach to full utilization of
lignocellulosic biomass under mild conditions.
Nat. Catal. 1, 772–780.
25. Liu, L., Zhao, H., Andino, J.M., and Li, Y. (2012).
Photocatalytic CO2 reduction with H2O on TiO2
nanocrystals: comparison of anatase, rutile,
and brookite polymorphs and exploration of
surface chemistry. ACS Catal. 2, 1817–1828.
39. Xie, X.-Y., Xiao, P., Fang, W.-H., Cui, G., and
Thiel, W. (2019). Probing photocatalytic
nitrogen reduction to ammonia with water on
the rutile TiO2 (110) surface by first-principles
calculations. ACS Catal. 9, 9178–9187.
12. Zhou, B., Song, J., Zhang, Z., Jiang, Z., Zhang,
P., and Han, B. (2017). Highly selective
photocatalytic oxidation of biomass-derived
chemicals to carboxyl compounds over Au/
TiO2. Green Chem. 19, 1075–1081.
26. Yang, Y., Yin, L.C., Gong, Y., Niu, P., Wang,
J.Q., Gu, L., Chen, X., Liu, G., Wang, L., and
Cheng, H.M. (2018). An unusual strong visible-
light absorption band in red anatase TiO2
photocatalyst induced by atomic
40. Meng, X., Ouyang, S., Kako, T., Li, P., Yu, Q.,
Wang, T., and Ye, J. (2014). Photocatalytic CO2
conversion over alkali modified TiO2 without
loading noble metal cocatalyst. Chem.
13. Da Via`, L., Recchi, C., Gonzalez-Yanez, E.O.,
˜
Davies, T.E., and Lopez-Sanchez, J.A. (2017).
Visible light selective photocatalytic conversion
of glucose by TiO2. Appl. Catal. B 202, 281–288.
hydrogen-occupied oxygen vacancies. Adv.
Mater. 30, 1704479.
Commun. (Camb.) 50, 11517–11519.
27. Chen, X., Liu, L., Yu, P.Y., and Mao, S.S. (2011).
Increasing solar absorption for photocatalysis
with black hydrogenated titanium dioxide
nanocrystals. Science 331, 746–750.
41. Schrauben, J.N., Hayoun, R., Valdez, C.N.,
Braten, M., Fridley, L., and Mayer, J.M. (2012).
Titanium and zinc oxide nanoparticles are
proton-coupled electron transfer agents.
Science 336, 1298–1301.
14. Mariscal, R., Maireles-Torres, P., Ojeda, M.,
Sa´daba, I., and Lo´ pez Granados, M. (2016).
Furfural: a renewable and versatile platform
molecule for the synthesis of chemicals and
fuels. Energy Environ. Sci. 9, 1144–1189.
28. Guo, Q., Zhou, C., Ma, Z., Ren, Z., Fan, H., and
Yang, X. (2016). Elementary photocatalytic
chemistry on TiO2 surfaces. Chem. Soc. Rev. 45,
3701–3730.
42. Xie, S., Shen, Z., Deng, J., Guo, P., Zhang, Q.,
Zhang, H., Ma, C., Jiang, Z., Cheng, J., Deng,
D., and Wang, Y. (2018). Visible light-driven C-
H activation and C-C coupling of methanol into
ethylene glycol. Nat. Commun. 9, 1181.
15. Chen, S., Wojcieszak, R., Dumeignil, F.,
Marceau, E., and Royer, S. (2018). How catalysts
and experimental conditions determine the
selective hydroconversion of furfural and 5-
hydroxymethylfurfural. Chem. Rev. 118, 11023–
11117.
29. Fenoglio, I., Greco, G., Livraghi, S., and Fubini,
B. (2009). Non-UV-induced radical reactions at
the surface of TiO2 nanoparticles that may
trigger toxic responses. Chem. Eur. J. 15, 4614–
4621.
43. Li, Q.H., Gao, T., Wang, Y.G., and Wang, T.H.
(2005). Adsorption and desorption of oxygen
probed from ZnO nanowire films by
photocurrent measurements. Appl. Phys. Lett.
86, 123117.
16. Fache, M., Boutevin, B., and Caillol, S. (2016).
Vanillin production from lignin and its use as a
renewable chemical. ACS Sustainable Chem.
Eng. 4, 35–46.
30. Sarkar, A., and Khan, G.G. (2019). The
formation and detection techniques of oxygen
vacancies in titanium oxide-based
44. Dolamic, I., and Burgi, T. (2007). Photocatalysis
¨
17. Zang, H., Wang, K., Zhang, M., Xie, R., Wang,
L., and Chen, E.Y.-X. (2018). Catalytic coupling
of biomass-derived aldehydes into
nanostructures. Nanoscale 11, 3414–3444.
of dicarboxylic acids over TiO2: an in situ ATR-
IR study. J. Catal. 248, 268–276.
31. Yu, X., Kim, B., and Kim, Y.K. (2013). Highly
enhanced photoactivity of anatase TiO2
nanocrystals by controlled hydrogenation-
induced surface defects. ACS Catal. 3, 2479–
2486.
intermediates for biofuels and materials. Catal.
Sci. Technol. 8, 1777–1798.
45. Babu, N.R., Subashchandrabose, S., Ali
ꢀ
Padusha, M.S., Saleem, H., and Erdogdu, Y.
18. Bond, J.Q., Alonso, D.M., Wang, D., West,
R.M., and Dumesic, J.A. (2010). Integrated
(2014). Synthesis and spectral characterization
of hydrazone derivative of furfural using
Chem 6, 1–16, November 5, 2020
15

Please cite this article in press as: Wu et al., Selectivity Control in Photocatalytic Valorization of Biomass-Derived Platform Compounds by Surface
Engineering of Titanium Oxide, Chem (2020), https://doi.org/10.1016/j.chempr.2020.08.014
ll
Article
experimental and DFT methods. Spectrochim.
Acta A Mol. Biomol. Spectrosc. 120, 314–322.
photoreduction under visible light. ACS Catal.
6, 1097–1108.
58. Grimme, S., Antony, J., Ehrlich, S., and Krieg, H.
(2010). A consistent and accurate ab initio
parametrization of density functional
dispersion correction (DFT-D) for the 94
elements H-Pu. J. Chem. Phys. 132, 154104.
46. Xiong, K., Wan, W., and Chen, J.G. (2016).
Reaction pathways of furfural, furfuryl alcohol
and 2-methylfuran on Cu(111) and NiCu
bimetallic surfaces. Surf. Sci. 652, 91–97.
52. Han, X., Kuang, Q., Jin, M., Xie, Z., and Zheng,
L. (2009). Synthesis of titania nanosheets with a
high percentage of exposed (001) facets and
related photocatalytic properties. J. Am.
Chem. Soc. 131, 3152–3153.
59. Hartwigsen, C., Goedecker, S., and Hutter, J.
(1998). Relativistic separable dual-space
Gaussian pseudopotentials from H to Rn. Phys.
Rev. B 58, 3641–3662.
47. Thang, H.V., Tosoni, S., Fang, L., Bruijnincx, P.,
and Pacchioni, G. (2018). Nature of sintering-
resistant, single-atom Ru species dispersed on
zirconia-based catalysts: a DFT and FTIR Study
of CO adsorption. ChemCatChem 10, 2634–
2645.
53. Zhang, J., Liu, P., Lu, Z., Xu, G., Wang, X.,
Qian, L., Wang, H., Zhang, E., Xi, J., and Ji, Z.
(2015). One-step synthesis of rutile nano-TiO2
with exposed {111} facets for high
photocatalytic activity. J. Alloys Compd. 632,
133–139.
60. VandeVondele, J., and Hutter, J. (2007).
Gaussian basis sets for accurate calculations on
molecular systems in gas and condensed
phases. J. Chem. Phys. 127, 114105.
48. Luo, N., Hou, T., Liu, S., Zeng, B., Lu, J., Zhang,
J., Li, H., and Wang, F. (2020). Photocatalytic
coproduction of deoxybenzoin and H2 through
tandem redox reactions. ACS Catal. 10,
762–769.
61. Hirshfeld, F.L. (1977). Bonded-atom fragments
for describing molecular charge densities.
Theor. Chim. Acta 44, 129–138.
54. VandeVondele, J., Krack, M., Mohamed, F.,
Parrinello, M., Chassaing, T., and Hutter, J.
(2005). Quickstep: fast and accurate density
functional calculations using a mixed Gaussian
and plane waves approach. Comput. Phys.
Commun. 167, 103–128.
62. Fonseca Guerra, C., Handgraaf, J.W.,
Baerends, E.J., and Bickelhaupt, F.M. (2004).
Voronoi deformation density (VDD) charges:
assessment of the Mulliken, Bader, Hirshfeld,
Weinhold, and VDD methods for charge
analysis. J. Comput. Chem. 25, 189–210.
49. Nowotny, J., Alim, M.A., Bak, T., Idris, M.A.,
Ionescu, M., Prince, K., Sahdan, M.Z., Sopian,
K., Mat Teridi, M.A., and Sigmund, W. (2015).
Defect chemistry and defect engineering of
TiO2-based semiconductors for solar energy
conversion. Chem. Soc. Rev. 44, 8424–8442.
55. Perdew, J.P., Burke, K., and Ernzerhof, M.
(1996). Generalized gradient approximation
made simple. Phys. Rev. Lett. 77, 3865–3868.
63. Jiang, Y., Adams, J.B., and Sun, D. (2004).
Benzotriazole adsorption on Cu2O (111)
surfaces: a first-principles study. J. Phys. Chem.
B 108, 12851–12857.
56. Krukau, A.V., Vydrov, O.A., Izmaylov, A.F., and
Scuseria, G.E. (2006). Influence of the exchange
screening parameter on the performance of
screened hybrid functionals. J. Chem. Phys.
125, 224106.
50. Li, H., Li, J., Ai, Z., Jia, F., and Zhang, L. (2018).
Oxygen vacancy-mediated photocatalysis of
BiOCl: reactivity, selectivity, and perspectives.
Angew. Chem. Int. Ed. 57, 122–138.
64. Watanabe, T., Manz, T.A., and Sholl, D.S.
(2011). Accurate treatment of electrostatics
during molecular adsorption in nanoporous
crystals without assigning point charges to
framework atoms. J. Phys. Chem. C 115, 4824–
4836.
51. Liu, L., Jiang, Y., Zhao, H., Chen, J., Cheng, J.,
Yang, K., and Li, Y. (2016). Engineering
57. Cheng, J., Liu, X., VandeVondele, J., and Sprik,
M. (2015). Reductive hydrogenation of the
aqueous rutile TiO2 (110) surface. Electrochim.
Acta 179, 658–667.
coexposed {001} and {101} facets in oxygen-
deficient TiO2 nanocrystals for enhanced CO2
16
Chem 6, 1–16, November 5, 2020