One-pot photoreforming of cellulosic biomass waste to hydrogen by merging photocatalysis with acid hydrolysis

Article abstract of DOI:10.1016/j.apcata.2018.06.030

Environmental-friendly utilization and conversion of abundant lignocellulosic biomass waste into chemical fuels is considered as one of the efficient approaches developing renewable energy. Herein, we merge photocatalysis and acid hydrolysis to solve problems in biomass conversion and hydrogen production. Specifically, photocatalytic reforming cellulose to hydrogen has been accomplished by hydrolysis of cellulose in 0.6 M sulfuric acid solution at 403 K in the presence of a photocatalyst (e.g. platinized TiO₂) under UV-light irradiation. Carbohydrates as the sacrificial electron donors are in situ generated via acid hydrolysis of cellulose. An unexpected production of 5-hydroxyl methyl furfural (HMF) has also been found during the combined process. In addition, efficient and repeatable hydrogen production has been achieved from photoreforming of raw biomass waste (paper pulp) as continuous supply of electron donors. Compared to traditional thermochemical gasification or pyrolysis of cellulose in high temperature and pressure, this work provides an alternative approach for producing hydrogen and valuable chemicals from lignocellulosic biomass.

Full text of DOI:10.1016/j.apcata.2018.06.030

Accepted Manuscript
Title: One-pot photoreforming of cellulosic biomass waste to
hydrogen by merging photocatalysis with acid hydrolysis
Authors: Jing Zou, Guan Zhang, Xiaoxiang Xu
PII:
S0926-860X(18)30308-9
DOI:
Reference:
https://doi.org/10.1016/j.apcata.2018.06.030
APCATA 16718
To appear in:
Applied Catalysis A: General
Received date:
Revised date:
Accepted date:
20-3-2018
22-6-2018
23-6-2018
Please cite this article as: Zou J, Zhang G, Xu X, One-pot photoreforming of cellulosic
biomass waste to hydrogen by merging photocatalysis with acid hydrolysis, Applied
Catalysis A, General (2018), https://doi.org/10.1016/j.apcata.2018.06.030
This is a PDF file of an unedited manuscript that has been accepted for publication.
As a service to our customers we are providing this early version of the manuscript.
The manuscript will undergo copyediting, typesetting, and review of the resulting proof
before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that
apply to the journal pertain.

One-pot photoreforming of cellulosic biomass waste to hydrogen by merging
photocatalysis with acid hydrolysis
a,b
b,*
c,**
Jing Zou , Guan Zhang , Xiaoxiang Xu
aSchool of Humanities & Social Science, The Chinese University of Hong Kong, Shenzhen,
Shenzhen 518172, P.R. China.
bSchool of Civil and Environmental Engineering, Harbin Institute of Technology, Shenzhen,
Shenzhen 510855, P.R. China
cShanghai Key Laboratory of Chemical Assessment and Sustainability, School of Chemical
Science and Engineering, Tongji University, Shanghai 200092, P.R. China.
*
E-mail: zhangguan@hit.edu.cn (G. Zhang), xxxu@tongji.edu.cn (X. Xu)
Graphical Abstract:
Efficient photocatalytic reforming cellulosic biomass to hydrogen has been accomplished via
merging photocatalysis and acid hydrolysis.
Highlights:



Efficient photocatalytic reforming cellulose to hydrogen has been accomplished.
Carbohydrates as sacrificial electron donors were in situ generated via acid hydrolysis.
An unexpected increase of 5-hydroxyl methyl furfural (HMF) production was also
found.
1


Repeatable hydrogen production was achieved from photoreforming of raw biomass.
Abstract: Environmental-friendly utilization and conversion of abundant
lignocellulosic biomass waste into chemical fuels is considered as one of the
efficient approaches developing renewable energy. Herein, we merge
photocatalysis and acid hydrolysis to solve problems in biomass conversion and
hydrogen production. Specifically, photocatalytic reforming cellulose to hydrogen
has been accomplished by hydrolysis of cellulose in 0.6 M sulfuric acid solution at
4
03 K in the presence of a photocatalyst (e.g. platinized TiO ) under UV-light
2
irradiation. Carbohydrates as the sacrificial electron donors are in situ generated via
acid hydrolysis of cellulose. An unexpected production of 5-hydroxyl methyl
furfural (HMF) has also been found during the combined process. In addition,
efficient and repeatable hydrogen production has been achieved from
photoreforming of raw biomass waste (paper pulp) as continuous supply of electron
donors. Compared to traditional thermochemical gasification or pyrolysis of
cellulose in high temperature and pressure, this work provides an alternative
approach for producing hydrogen and valuable chemicals from lignocellulosic
biomass.
Keywords: Biomass conversion; Photocatalysis; Hydrogen production; Acid
hydrolysis; Photoreforming
1
. Introduction
2

Hydrogen fuel as a clean and promising energy storage medium has attracted
tremendous attention for industrial application. To date, H is still mainly produced by
2
reforming or gasification of fossil fuel products in industry. However, fossil fuel based
energy sources definitely can not meet the requirement of globally boosting economic
growth and sustainable development. Therefore, H production from renewable energy
2
sources including solar energy, biomass, water and wind etc. has being intensively
investigated to solve global problems of fossil fuel depletion and environmental
deteriation [1-3].
Cellulose as the most abundant renewable biomass source has been widely
investigated for H
conversion of cellulose to H
instance, pyrolysis and gasification processes operated at high temperature (773 – 1073
K) transform cellulose to bio-oils and H gas, along with other gases, such as CO, CO
and CH [7]. In respect of the required high energy input, solar light as an inexhaustible
energy source could be used to replace the thermal input. In addition, photoreforming
of cellulose or its derivatives to H , firstly reported at 1980s [8], is receiving more
attention nowadays [9-16]. In this process, cellulose serves as an electron donor (hole
scavenger) to promote aqueous protons reduction to H by photogenerated electrons.
However, due to the insoluble nature of cellulose in water, suspended cellulose in water
is not efficient electron donor for promoting H production [12]. Caravaca et al. claimed
that photocatalysis driven cellulose hydrolysis to glucose is the first step for producing
2
production under different conditions [4-7]. The main approach for
2
has been realized by thermochemical gasification. For
2
2
4
2
2
2
hydrogen on metal loaded TiO
2
[13], however, cellulose hydrolysis is hardly occurred
3

in pure water even under photocatalysis, which inevitably limits the conversion of
cellulose. Therefore, it is a challenging task to make cellulose efficiently consumed by
oxidative holes generated from photocatalysts. To address this issue, few examples on
photoreforming of cellulose to H
2
are recently reported by immobilizing cellulose
particles onto CdS [14] and platinized TiO
2
photocatalyst [15] or dissolving cellulose
-
in highly concentrated alkaline solution (5~10 M [OH ]) in the presence of CdS/CdO
x
quantum dots photocatalyst [16].
On the other hand, cellulosic biomass can be converted into other valuable liquid
platform molecules such as glucose, gluconic acid and 5-hydroxyl methyl furfural
(
HMF) through acid hydrolysis or enzyme hydrolysis, to compete with thermochemical
process [17-19]. Typically, hydrolysis with concentrated acid (> 5%) allows the use of
milder conditions (373 K - 423 K). However, subsequent separation and purification of
the valuable bio-products from the reaction medium is rather difficult, which is still an
issue to be resolved for practical application. Thus, considering the widespread
applications of both photocatalysis and acid catalysis as shown in Scheme 1, we
questioned whether it might be possible to merge these two catalytic processes in one-
pot, with the goal of solving long-standing, yet elusive problems in H production and
2
biomass conversion. Specifically, acid hydrolysis of lignocellulosic biomass (e.g.
cellulose) generates soluble organic carbohydrates, which could be in situ used as the
electron donors to promote photocatalytic H
2
production from water. The generated
carbohydrates are very effective electron donors due to their low oxidization potentials
[
20], and photocatalytic H
2
generation from the carbohydrates (e.g. glucose)
4

decomposition has also been demonstrated [21,22]. To our best knowledge, this is the
first example that employing combined photocatalysis and acid catalysis for cellulose
conversion to hydrogen. The merit of this photocatalytic H production strategy lies in
2
the unnecessary input of valuable chemicals (e.g. alcohols) as electron donors and the
avoidance of redundant separation of liquid products in cellulose hydrolysis, since
gaseous H can be easily recovered and purified via a membrane filtration process.
2
Scheme 1. Merging acid catalysis and photocatalysis to enable efficient hydrogen
production by photoreforming cellulose.
2
. Experimental
2
.1. Materials
Original microcrystalline α-cellulose powder (Sigma-Aldrich, size 10 ~ 1000 μm)
without any pretreatment was used for test, chloroplatinic acid (H
2
PtCl
6
.6H O, >99.9%
2
Sigma-Aldrich), standard carbohydrates include glucose (>99.5%), fructose (>99.5%),
arabinose (99%), erythose (>75%), 5-(hydroxylmethyl)furfural (HMF, >99% ), formic
acid (>95%) and gluconic acid (49~53 wt% in water) etc. were all purchased from
Sigma-Aldrich company.
5

2
.2. Photocatalyst preparation
Platinized TiO (Degussa P25, specific surface area 50 m /g) with a typical Pt
loading of 0.5 wt.% was prepared by a photo-deposition method. A 500 ml TiO
2
2
2
(1.0
-1
-1
gL ) suspension with 50 ml methanol as electron donor and 6.5 ml of 1.0 gL
chloroplatinic acid (H PtCl .6H
lamp) for 60 min. After 1 h irradiation, Pt-deposited TiO
washed with D.I. water, and dried in air at room temperature.
.3. Cellulose and photocatalyst characterization
Particle size of cellulose was measured on a Mastersizer 2000 Particle Size
2
6
2
O) was irradiated with UV light ( 250 W iron-halide
(p-TiO ) powder was filtered,
2
2
2
Analyser (Malvern Instruments). Microstructure of samples was observed on a JEOL
JSM-6700F Scanning Electron Microscope (SEM) at 5 kV. The SEM specimen was
prepared by placing small amounts of sample powder onto a carbon plate attached to
an aluminum sample holder and gold powder was allowed to evaporate under vacuum
at room temperature. X-ray diffraction analysis of powder was examined on a SToe
STADI/P powder diffractometer. Incident radiation was generated using a Cu k
λ=1.54056 Å). Transmission electron microscope (TEM) was performed using a JEOL
JEM-2011 electron microscope at 200 kV.
.4. Activity test
For hydrolysis experiment, a certain amount of cellulose powder (2.5~12.5 mg)
and TiO catalyst (0~25 mg) was suspended in 50 ml H SO solution (typically 0.6 M)
α
source
(
2
2
2
4
in a home-made autoclave reactor with a quartz window on top of the reactor (as shown
o
in Scheme 2). The autoclave was heated to ~130 C in an oil bath with monitoring the
6

pressure inside of the autoclave (~ 3.0 bar). After reaction, the autoclave was directly
removed from oil-bath and cooled down to room temperature within 2 h. Then, the
aqueous solution was analyzed by high performance liquid chromatography (HPLC)
equipped with RI detector and Agilent Hi-Plex column for sugar analysis. (HPLC
condition: DI water was used as mobile phase, flow rate of 0.6 ml/min and column
o
temperature at 79 C). For quantifying the generated carbohydrates, commercial
carbohydrates such as glucose, fructose, 5-(hydroxylmethyl)furfural, arabinose,
erythose, formic acid and gluconic acid were used to measure the standard calibration
curve of each carbohydrate according to the HPLC profile.
In case of combined photocatalysis and hydrolysis experiment, the autoclave
reactor was heated in an oil-bath and UV-Visible light irradiation (250 W, iron doped
halide lamp) from the top of the reactor was also employed. In the preliminary test, a
certain amount of cellulose powder (2.5~12.5 mg) and TiO
were suspended in 50 ml 0.6 M H SO solution in the autoclave reactor under UV-
visible light irradiation. Before photoirradiation, the sealed photoreactor was purged
with Ar for 30 min to remove the O and N . The amounts of photogenerated H and
CO (from cellulose decomposition) in the head-space (0.65 L) of photoreactor under
2
(Pt) powder (0~25 mg)
2
4
2
2
2
2
irradiation were examined using an online gas chromatograph (Agilent 3000 Micro Gas
chromatograph). The pressure inside autoclave was identical with the hydrolysis
experiment. Each photocatalysis run was performed for 10.0 h.
For evaluation of the potential application of reforming raw biomass waste by this
strategy, paper pulp (2.5 and 500 mg) as a representative raw biomass waste was
7

suspended into the reactor in the presence of 25 mg Pt-TiO photocatalyst. The reaction
2
was performed both at 323 K and 403 K. During the multicycle tests, the reactor was
purged with Ar for 30 min again prior to starting the next cycle to remove the
accumulated H .
2
Scheme 2. Illustration of experimental reactor used for photoreforming cellulose to
hydrogen via combined photocatalysis and acid hydrolysis.
3
. Results and discussion
3
.1. Structural characterization of cellulose and photocatalyst
Cellulose is a polysaccharide consisting of a linear chain of several hundred to
many thousands of β(1→4) linked D-glucose units, and it consists of crystalline and
amorphous regions. Firstly, crystal structure and particle size of commercial α-cellulose
powder (without pre-treatment) was analysed by X-ray diffraction (XRD), scanning
electron microscopy (SEM) and particle size analyzer. The crystallinity of α-cellulose
was estimated about 74% according to the equation of Crystallinity Index (CI) = (I002
-
8

I
AM)/I002 as shown in the XRD pattern (Fig.1a) [23]. The crystalline structure of
cellulose may be broken up upon treating it with strong acid or base. The particle size
of α-cellulose was estimated about from 10~1000 μm according to the SEM image and
particle size measurement as shown in Fig. 1b and 1c. Thus, such rigid microcrystalline
cellulose powder normally requires harsh reaction condition to be further convertible.
The representative TEM image shown in Fig. 1d clearly shows the P25 TiO
2
are
composed by 20-30 nm nanoparticles. Through the photoreduction process, the 0.5 wt.%
Pt nanoparticles (dark spots) with few nanometer size are randomly deposited onto TiO
nanoparticles.
2
Fig.1. (a) XRD pattern of commercial microcrystalline α-cellulose powder; (b) SEM
image of microcrystalline α-cellulose; (c) Particle size distribution of original
microcrystalline α-cellulose; (d) Representative TEM image of platinized TiO
2
nanoparticles.
9

3
.2. Acid hydrolysis of cellulose
Sulfuric acid catalysed cellulose hydrolysis (cleavage of the β-1,4-glycosidic bonds
and H-bonds) has been studied theoretically and experimentally, as this degradation
step paves the way for other following catalytic transformations. Cellulose hydrolysis
has been modelled as a pseudo-first order sequential reaction of cellulose hydrolysis
followed by glucose degradation [24].
Cellulose → khyd [Glucose] → kdeg [HMF + H
2
O] → LA + FA
[1]
where khyd = kinetic constant of cellulose hydrolysis, kdeg = kinetic constant of glucose
degradation, HMF = hydroxymethyl furfural, LA = levulinic acid, FA = formic acid.
Hydrolysis of α-cellulose powder was tested in diluted sulfuric acid in a closed
autoclave reactor under varied conditions (Scheme 2). As shown in the equation [1]
describing the decomposition pathway of cellulose, glucose was found as the main
hydrolysis product, while other hydrolysis products such as fructose (an isomer of
glucose), 5-(hydroxylmethyl)furfural (HMF, dehydration from fructose) and other
carbohydrates were also detected with much lower yields (Table 1, entry 1), which
could be neglected compared to the glucose yield.
Table 1 The detected carbohydrates in aqueous phase from decomposition of cellulose under different conditions.
Entry
Process
Condition
Catalyst
Product Yield (%) based on the C6 unit in cellulose
Glucose
Fructose
HMF
Erythrose
Arabinose
1
2
3
4
5
6
Hydrolysis
Hydrolysis
Hydrolysis
Hydrolysis
Hydrolysis
Photocatalysis
Aerobic
Aerobic
Aerobic
Anoxic
Anoxic
Anoxic
-
78.2(±3.0)
86.8(±5.0)
0.6 (±0.5)
1.6 (±0.3)
1.2(±0.4)
0.8(±0.5)
0.6(±0.3)
trace
1.5(±0.3)
1.2(±0.5)
2.0(±1.0)
3.0(±1.0)
3.3(±1.0)
12.8(±2.0)
0.3(±0.1)
0.5(±0.3)
0.4(±0.3)
0.5(±0.2)
0.6(±0.3)
5.4(±2.0)
0.2(±0.1)
0.3(±0.2)
0.4(±0.2)
0.3(±0.1)
0.5(±0.2)
3.3(±0.5)
10 mg TiO
2
10 mg TiO
10 mg TiO
25 mg TiO
25 mg TiO
2
(Pt) 85.6(±4.0)
(Pt) 86.3(±4.0)
(Pt) 85.2(±3.0)
(Pt) trace
2
2
2
+
Hydrolysis
10

Experimental Conditions: cellulose 2.5 mg, 50 mL 0.6 M H
2 4
SO , reaction temperature: 403 K, autogenous pressure: 3.0 bar; reaction
time: 10 h; UV light irradiation from an iron-doped halide lamp.
Note 1: The organic acids products from hydrolysis and degradation of cellulose could not be quantified by our HPLC analysis, because
the retention times of most organic acids（4~5 min）are overlapped with the retention time of highly concentrated sulfuric acid, which
shows a huge and broad peak in the HPLC profile.
Fig. 2a shows outstanding yields of glucose from cellulose hydrolysis at 403 K
(
autogenous pressure 3.0 bar) after 10.0 h reaction with varied concentrations of H
Increasing concentration of H SO from 0.1 M to 0.6 M significantly enhanced glucose
yields, while the glucose yield in 1.2 M H SO was slightly higher than that obtained
2
SO .
4
2
4
2
4
in 0.6 M H
2
SO
4
. It is found that the increase of glucose yields is linearly proportional
+
to the increase of acid concentration (e.g. [H ]) from 0.1 M to 0.6 M, which confirms
+
the dependence of hydrolysis rate upon the concentration of H in this tested scale.
Cellulose hydrolysis was not efficient at lower temperature and pressure (< 403 K, <
2
.0 bar) even in 0.6 M H
2
SO (data not shown).
4
1
00
(
a)
8
6
4
2
0
0
0
0
0
0.1
0.3
0.6
1.2
H SO concentration (M)
2
4
11

1
00
(
b)
8
6
4
2
0
0
0
0
0
0.0
2.5
5.0
10.0
25.0
TiO (mg)
2
1
00
10.0h
(c)
8
6
4
2
0
0
0
0
0
4.0h
2.0h
1.0h
Fig. 2. Glucose production from sulfuric acid catalyzed cellulose hydrolysis under
different conditions: (a) with varied concentration of sulfuric acid; (b) with varied
amount of TiO
different reaction periods.
Therefore, hydrolysis experiment was performed in 0.6 M H
following studies. As shown in Fig. 2b, glucose yields slightly increased with increasing
2
in 0.6 M sulfuric acid; (c) with 10.0 mg TiO in 0.6 M sulfuric acid at
2
2
SO at 403 K for the
4
the amount of TiO
2
nanoparticles, which reached to the optimum concentration of TiO
2
at 0.2 g/L (e.g. 10.0 mg in 50 ml H
2
SO ). Metal oxides (e.g. CaFe and HNbMoO
4
2
O
4
6
)
have been employed as brØnsted acid catalysts for cellulose hydrolysis [25,26].
Similarly, sulfonated TiO , as a solid acid catalyst [27], would promote cellulose
hydrolysis due to the increased brØnsted acid density in acidic condition, as the glucose
yields were enhanced from ca. 78.2(±3)% (without TiO ) to ca. 86.8 (±5)% (TiO 0.2
g/L). Further increase of TiO (0.5 g/L) reduced the glucose yield, since large excess of
2
2
2
2
12

TiO may cause the aggregation of nanoparticles which inevitably inhibits the contact
2
between cellulose and catalyst particles. The time profiles of glucose production from
cellulose hydrolysis within 10.0 h (Fig. 2c) indicates that glucose production rate
gradually decreased along with reaction time, as over 60% glucose yield was obtained
within initial 4.0 h reaction. It is proposed that the depletion of cellulose accounts for
the reduced glucose production rate.
3
.3. Combined acid hydrolysis and photocatalysis
In order to study the combined hydrolysis and photocatalysis, cellulose hydrolysis
experiment was performed under anoxic condition with simultaneous UV-light
irradiation to the mixed titania photocatalyst and cellulose suspension. Platinum (0.5
wt.%) co-catalyst was pre-loaded onto TiO
2
nanoparticles by a photodeposition method
to enhance H production. The few amounts of platinum did not exert significant
2
influences on the hydrolysis process, as there was no remarkable difference for the
glucose yields with or without platinum deposition (see Table 1, entry 2 and 3). In
addition, oxygen effect on the hydrolysis could be neglected as well (see Table 1, entry
3
and 4). In the absence of photocatalyst, there was no H
suggesting that H was produced from the photocatalytic process, rather than from the
thermal or hydrolytic process. Furthermore, in the absence of cellulose, the H evolution
under the tested condition (0.6 M H SO , 403 K) was also negligible, implying that
water reduction is not favored without efficient electron donors in strong acidic
condition at high temperature. The optimum concentration of TiO (Pt) loading for H
evolution was found about 0.5 g/L ( 25 mg TiO (Pt) in 50 ml solution). Further increase
2
evolved as shown in Fig. 3,
2
2
2
4
2
2
2
13

in the concentration of TiO
yield probably because of the aggregation of the TiO
inhibited UV light transmission in the mixed suspension [28]. The H
were gradually reduced along with reaction time. This is thought mainly due to the
depletion of organic intermediates produced from cellulose hydrolysis. Since H was
only produced in the presence of cellulose, cellulose conversion efficiency reached to
ca. 66% for H production within 10.0 h under the optimized condition, based on the
assumption that 1 mole glucose unit hydrolysed from cellulose would produce 12 mole
after oxidizing to carbon dioxide (C + 6H O = 12H + 6CO ) theoretically.
2
(Pt) to 1.0 g/L ( 50 mg TiO
2
(Pt) in 50 ml solution) reduced
(Pt) nanoparticles and the
evolution rates
H
2
2
2
2
2
H
2
6
H
12
O
6
2
2
2
1
44
80
60
40
20
0
5
2
1
0
0 mg TiO (Pt)
2
5 mg TiO (Pt)
2
0 mg TiO (Pt)
108
2
mg TiO (Pt)
2
no cellulose
7
2
3
6
0
0
2
4
6
8
10
Time (h)
Fig. 3. Photocatalytic H
catalyst with varied amounts of TiO
The acid effect on H production was further investigated as shown in Fig. 4. Trace
amount of H was produced in 0.1 M sulfuric acid condition at 403 K, and the cellulose
2
production assisted by cellulose hydrolysis with TiO
2
(Pt)
2
(Pt) loading.
2
2
conversion efficiency was ca. 8.3%, in consistence with the glucose yield (ca. 10%)
from the cellulose hydrolysis experiment. This indicates that nearly all the generated
glucose and its hydrolytic intermediates as sacrificial electron donors were photo-
14

oxidized to CO
2
, in accompany with H
.6 M sulfuric acid condition was markedly enhanced. The H
5 μmol and 123 μmol in 0.3 M and 0.6 M H SO respectively, which accounts for ca.
2
production. However, H
2
evolution in 0.3 M or
0
9
5
4
2
yields in 10.0 h were ca.
2
4
0% and 66% cellulose conversion efficiency. Considering that the glucose yield of ca.
0% in 0.3 M in 10.0 h, it implies that the other hydrolysis carbohydrates (e.g.
cellodextrins) in addition to glucose are also served as electron donors for H
In case of 0.6 M H SO condition, the H yield (66%) is significantly lower than that of
glucose yield (85%), which indicates that the hydrolysis rate in 0.6 M H SO is much
faster than the photocatalytic H evolution rate, and proton reduction is rate-limiting
2
production.
2
4
2
2
4
2
step in the overall process. Some potion of generated glucose was converted into HMF,
erythrose and arabinose as shown in Table 1 (entry 6), which reduces the cellulose to
H
2
conversion efficiency. Furthermore, H
with cellulose at 323 K, which also demonstrates that cellulose hydrolysis process is
essentially necessary for efficient photocatalytic H production.
2
evolution was negligible in 0.6 M H
2
SO
4
2
1
44
8
6
4
2
0
0
0
0
0
0.6 M(403K)
0.3 M(403K)
0.1 M(403K)
0.6 M(323K)
1
08
7
3
2
6
0
0
2
4
6
8
10
Time (h)
Fig. 4. Photocatalytic H
2
production assisted by cellulose hydrolysis with TiO (Pt)
2
catalyst in varied concentrations of sulfuric acid.
15

In addition, cellulose effect on H
production is influenced by both the photocatalysis and hydrolysis. Cellulose
hydrolysis provides electron donors for photocatalytic H evolution, thus the
photocatalysis mainly determines the H evolution rate when cellulose loading is
sufficient. The amount of loaded cellulose in 50 ml 0.6 M H SO solution was varied
from 0.0 mg to 12.5 mg. It is found that H yield gradually increases along with the
increase of cellulose (Fig. 5a). Without cellulose loading, no H was evolved in 0.6 M
SO at 403 K. In the presence of highly concentrated sulfuric acid, the positive TiO
surface might be covered by SO 2-
and the photogenerated charge transfer process was
inhibited, which might be one of the reasons preventing H evolution. However, H
2
production was also investigated. Hydrogen
2
2
2
4
2
2
H
2
4
2
4
2
2
production rate gradually reached to a maximum level when the amount of cellulose
was increased to 12.5 mg. It can be assumed that carbohydrates production from
hydrolysis and H
point. The time profiles of photocatalytic H
cellulose loading without concentrated acid (Fig. 5b), implying that cellulose was not
involved in the photocatalytic reaction in pure water. The total amount of produced H
2
production from water splitting achieve a maximum balance at this
2
production are similar regardless of
2
(
ca. 33 µmol) without cellulose hydrolysis is about 5-fold less than that generated with
cellulose hydrolysis (ca. 170 µmol). This confirms that more efficient H production
2
only occurs when the photocatalytic process and cellulose hydrolysis are combined.
16

(
a)
1
2.5 mg
5.0 mg
.5 mg
1.0 mg
1
60
2
1
20
0.0 mg
8
0
0
0
4
0
2
4
6
8
10
Time (h)
(b)40
with cellulose
no cellulose
3
2
1
0
0
0
0
0
2
4
6
8
10
Time (h)
Fig. 5. Photocatalytic H
2
production assisted by cellulose hydrolysis with TiO (Pt)
2
catalyst (a): with varied amounts of cellulose powder in sulfuric acid; (b): with or
without cellulose powder in pure water.
To further examine the applicability of the combined photocatalysis, it is necessary
to use the raw cellulosic biomass waste as the feedstock. Pulp is one of the most
abundant raw materials worldwide. We examined the photocatalytic H
assisted by direct conversion of dry pulp paper as a representative raw feedstock with
TiO (Pt) photocatalyst. The H yield based on the amount of C6 monosaccharide unit
2
production
2
2
contained in pulp paper (2.5 mg) reached to ca. 89% after 10.0 h reaction (Fig. 6a).
Photoreforming of pulp paper is much easier compared to photoreforming of cellulose.
In addition, practically the content of cellulosic feedstock should be much higher than
that of catalyst. Therefore, we fed 500 mg pulp paper into the reactor with 25 mg
17

photocatalyst for the long-term stability test. The repeated cycles of photocatalytic H
2
production from water with large excess of solid pulp paper under UV-light irradiation
at 323 K and 403 K were tested for about 70 h as shown in Fig. 6b. There was no
obvious decay for H
03 K is over 10-folds higher than that at 323 K, since acid hydrolysis of cellulosic pulp
paper is not sufficient at 323 K. The efficient and stable H production (ca. 1320
implies the stability of photocatalyst even in
such strong acidic condition. Thus, photoreforming of cellulosic biomass into H in
2
generation during the repeated tests. H
2
generation efficiency at
4
2
-1
μmol/h.g cat.) at 403 K in 0.6 M H
2
SO
4
2
large-scale is technically feasible, considering the utilization of low-cost photocatalyst
and feedstock.
(
a)100
8
6
4
2
0
0
0
0
0
0
2
4
6
8
10
Time (h)
(
b)3
50
00
50
00
50
00
403 K
3
2
2
1
1
50
323 K
0
0
10
20
30
40
50
60
70
Time (h)
Fig. 6. (a) Photocatalytic hydrogen production in the presence of 2.5 mg paper pulp in
18

0
.6 M H
2
SO
4
; (b) Repeated cycles of H production with 500 mg pulp paper under UV-
2
light irradiation at temperature of 323 K and 403 K.
3
.4. Mechanism analysis and discussion
The H
e.g. glucose) and subsequent oxidation of carbohydrates and proton reduction on
platinized TiO ( P25). Thus, the H evolution is dependent on both the hydrolytic and
2
production is mainly proceeded via the hydrolytic formation of carbohydrates
(
2
2
photocatalytic processes, which has been demonstrated by the investigation of
photocatalyst, acid and cellulose effects as shown from Fig. 3 to Fig. 5. When the acid
hydrolysis rate and the amount of cellulose loading are sufficient (operated condition:
0
.6 M H
2
SO
4
at 403 K), the overall H evolution is limited by the photocatalytic water
2
reduction process. Regarding to the oxidation species, it has been reported that both
hydroxyl radicals and peroxy species generated on noble metal loaded P25 account for
the glucose oxidation [22], and the glucose conversion process is through α–scission,
with the arabinose and erythrose as the main intermediates. However, in this work, the
arabinose and erythrose were less found (Table 1), probably because the longer
photocatalytic reaction time and higher reaction temperature led to further
decomposition of arabinose and erythrose into much smaller organic acids and CO .
2
According to the results on HPLC analysis of the liquid samples, the cellulose
conversion route during the hydrolysis or combined process was proposed as shown in
Scheme 3. Organic acids intermediates such as gluconic acid, levulinic acid or formic
acid, which might be produced during the photocatalytic process, were not able to be
determined due to the interference of sulfuric acid. (See Table 1 and Note 1). Acid
19

hydrolysis of cellulose to glucose (step I) or other minor carbohydrates derived from
the conversion of glucose such as fructose (step II, isomerization), 5-hydroxymethyl
furfural, HMF (step III, photo or photocatalytic conversion), and erythrose and
arabinose (step IV, photooxidation), continuously supply the electron donors to TiO (Pt)
2
photocatalyst for enhancing photocatalytic reduction of protons to hydrogen, while
these organic hydrolytic products would be further oxidized into small intermediates
and converted into CO
2
eventually (step V and VI), if the organic carbohydrates are not
excess.
More interestingly, we have found that conversion of glucose into HMF
preferentially occurs during the photocatalytic process, compared to the hydrolysis
process (Table 1, entry 5 and 6). HMF is thought more easily to be produced from the
dehydration of (cyclic) fructose with the assistance of sulfuric acid, however, HMF
yield is much lower without the UV-light irradiation. Even though the production of
HMF from UV-photocatalytic degradation of cellulose catalysed by TiO
2
in
concentrated ZnCl solution has been reported [29], the detailed mechanism for the
2
enhanced HMF formation during the UV-light irradiation process needs to be further
investigated. On the other hand, it can be speculated that HMF is more resistible to the
UV-light photocatalytic degradation than the other carbohydrates. Previous study has
also mentioned that photoreforming HMF proceeds at a much slower rate [16]. HMF is
a valuable platform chemical converted from cellulose decomposition, and has many
applications as an important reagent in organic synthesis. Thus, the selective HMF
production could be further optimized in this system, which deserves a further
20

systematic study.
Scheme 3. Proposed pathway of cellulose decomposition in combined processes of
hydrolysis and photocatalysis.
In contrast to photoreforming of biomass substance into hydrogen reported in
literature, in which more valuable chemicals such as alcohols (e.g. methanol or ethanol)
and sugars (e.g. glucose) are mostly used as the electron donors, this work directly
employs low-cost cellulose and raw paper pulp as the feedstock, which makes this H
2
production process more competitive. In order to scale-up this biomass conversion
approach, from the economic point of view, the noble-metal (e.g. Pd or Pt) based co-
catalyst should be replaced. Nowadays, various transitional metal based materials such
as Ni or MoS etc. have been reported as potential alternative co-catalysts of noble-
2
metals [30,31]. Practically, sulfuric acids could be continuously used without further
purification. In terms of more efficient utilization of solar energy, further improvements
can be expected through the development of visible light photocatalysts with strong
redox property [32-40], and utilization of a solar concentrator heating system for the
required energy during hydrolysis. On the other hand, the hydrolysis process and
21

photocatalysis process should be concordant for achieving the maximum efficiency, and
reactor engineering is also required to gain the gaseous H fuel.
2
4
. Conclusions
We have developed a simple and highly efficient one-pot process, based on merging
photocatalysis and acid hydrolysis, for H production from photoreforming cellulose in
water. The photocatalytic H production is promoted by in situ generation of electron
2
2
donors from direct conversion of lignocellulosic biomass materials. The combined
catalytic system works through the UV-light irradiation of mixed suspension of
platinized titania photocatalyst and cellulose particles in concentrated sulfuric acid
solution at 403 K. The reaction conditions are much milder and simpler compared to
other technologies for H production from biomass such as high temperature
2
gasification and steam reforming. The efficient hydrogen production system is also
applicable to utilization of other biomass and raw biomass materials without any pre-
treatment. Because of high cost and rigid reaction condition for the degradation of
cellulosic biomass in other enzymatic, chemocatalytic, thermocatalytic processes, the
results in this work are quite encouraging in terms of recovering hydrogen gas through
the decomposition of cellulose.
Acknowledgements
The authors thank the
KQJSCX20170327162043431, JSGG20170413152540284), and National Natural
Science Foundation of China (NSFC, 51708153).
Shenzhen Technology Innovation Support
(
22

Notes and references
[1] R. D. Cortright, R. R. Davda, J. A. Dumesic, Nature 418 (2002) 964−967.
[2] C. Ma, Y. Li, H. Zhang, Y. Chen, C. Lu, J. Wang, Chem. Eng. J. 273 (2009)
2
77−285.
[3] T. Udomchoke, S. Wongsakulphasatch, W. Kiatkittipong, A. Arpornwichanop, W.
Khaodee, J. Powell, J. Gong, S. Assabumrungrat, Chem. Eng. J. 303 (2016)
3
38−347.
[
[
[
[
[
[
[
[
[
4] G. W. Huber, S. Iborra, A. Corma, Chem. Rev. 106 (2006) 4044−4098.
5] G. Wen, Y. Xu, Z. Xu, Z. Tian, Catal. Commun. 11 (2010) 522−526.
6] S. S. Toor, L. Rosendahl, A. Rudolf, Energy 36 (2011) 2328−2342
7] D. A. Bulushev, J. R. H. Ross, Catal. Today 171 (2011) 1−13.
8] T. Kawai, T. Sakata, Nature 286 (1980) 474−476.
9] K. Shimura, H. Yoshida, Energy Environ. Sci. 4 (2011) 2467−2481.
10] X. Lu, S. Xie, H. Yang, Y. Tong, H. Ji, Chem. Soc. Rev. 43 (2014) 7581-7593.
11] A. V. Puga, Coord. Chem. Rev. 315 (2016) 1−66.
12] A. Speltini, M. Sturini, D. Dondi, E. Annovazzi, F. Maraschi, V. Caratto, A.
Profumo, A. Buttafava, Photochem. Photobiol. Sci. 13 (2014) 1410−1419.
13] A. Caravaca, W. Jones, C. Hardacre, M. Bowker, Proc. R. Soc. A472 (2016) 0054.
14] D. Ke, S. Liu, K. Dai, J. Zhou, L. Zhang, T. Peng, J. Phys. Chem. C 113 (2009)
[
[
1
6021−16026.
[
[
15] G. Zhang, C. Ni, X. Huang, A. Welgamage, L. Lawton, P. K. J. Robertson, J. T.
S. Irvine, Chem. Commun. 52 (2016) 1673−1676.
16] D. W. Wakerley, M. F. Kuehnel, K. L. Orchard, K. H. Ly, T. E. Rosser, E. Reisner,
Nature Energy 2 (2017) 17021.
[
[
[
[
[
[
17] J. Song, H. Fan, J. Ma, B. Han, Green Chem. 15 (2013) 2619−2635.
18] A. T. To, P-W. Chung, A. Katz, Angew. Chem. Int. Ed. 54 (2015) 11050−11053.
19] A. Wang, T. Zhang, Acc. Chem. Res. 46 (2013) 1377−1386.
20] J. Wang, Z. Taha, Anal. Chem. 62 (1990) 1413−1416.
21] M. R. S. John, A. J. Furgala, A. F. Sammells, J. Phys. Chem. 87 (1983) 801−805.
22] R. Chong, J. Li, Y. Ma, B. Zhang, H. Han, C. Li, J. Catal. 314 (2014) 101−108.
23

[23] K. H. Hong, G. Sun, J. Appl. Polym. Sci. 129 (2013) 3782−3786.
[24] N. S. Mosier, C. M. Ladisch, M. R. Ladisch, Biotechnol. BioEng. 79 (2002)
6
10−618.
[25] A. Takagaki, C. Tagusagawa, K. Domen, Chem. Commun. (2008) 5363−5365.
[26] T. Komanoya, H. Kobayashi, K. Hara, W. J. Chun, A. Fukuoka, Appl. Catal. A
4
07 (2011) 188−194.
[
[
[
27] M. G. Mazzotta, D. Gupta, B. Saha, A. K. Patra, A. Bhaumik, M. M. Abu-Omar,
ChemSusChem 7 (2014) 2342−2350.
28] D. Hufschmidt, D. Bahnemann, J. T. Testa, C. A. Emilio, M. I. Litter, J.
Photochem. Photobio. A: Chem. 148 (2002) 223−231.
29] H. Fan, G. Li, F. Yang, L. Yang, S. Zhang, J. Chem. Technol. Biotechnol. 86 (2011)
1107−1112.
[
[
[
[
[
[
[
[
30] H. Zhao, S. Sun, P. Jiang, Z. J. Xu, Chem. Eng. J. 315 (2017) 296−303.
31] Q. Xiang, J. Yu, M. Jaroniec, J. Am. Chem. Soc. 134 (2012) 6575−6578.
32] G. Zhang, G. Liu, L. Wang, J. T. S. Irvine, Chem. Soc. Rev. 45 (2016) 5951−5984.
33] X. Huang, Z. Wu, H. Zheng, W. Dong, G. Wang, Green Chem. 20 (2018) 664−670.
34] E. Huang, G. Liu, G. Zhang, X. Xu, Dalton Trans. 47 (2018) 4360−4367.
35] G. Zhang, J. Zou, X. Xu, Adv. Sustainable. Syst. 2 (2018) 1700139.
36] J. Hui, G. Zhang, C. Ni, J. Irvine, Chem. Commun. 53 (2017) 10038−10042.
37] G. Zhang, Y. Zhou, X. Fan, J. Zou, W. Dong, X. Xu, Int. Hydrogen Energy 42
(2017) 19877−19884.
[38] G. Zhang, C. Kim, W. Choi, Catal. Today, 281 (2017) 109−116.
[39] G. Zhang, C. Ni, L. Liu, G. Zhao, F. Fina, J. Irvine, J. Mater. Chem. A 3 (2015)
1
5413−15419.
[40] G. Zhang, G. Kim, W. Choi, Energy Environ. Sci. 7 (2014) 954−966.
24

25