Pt-Carbon interaction-determined reaction pathway and selectivity for hydrogenation of 5-hydroxymethylfurfural over carbon supported Pt catalysts

Article abstract of DOI:10.1039/d0cy01920h

In this study, a series of Pt catalysts supported on carbon materials, including multi-walled carbon nanotubes (CNTs), reduced graphene oxide (rGO), graphitic carbon nitride (g-C₃N₄), and biochar (BC), were prepared by impregnation or atomic layer deposition (ALD) methods. The Pt content was around 2.5 wt% in all the Pt catalysts, and the average Pt particle size was 1.6-1.9 nm for the catalysts prepared by impregnation (IM), which was slightly larger than that for the Pt catalyst (～1.4 nm) synthesized by ALD (Pt/CNTs-ALD). All the catalysts were used in the HMF hydrogenation reaction to investigate the effects of different carbon materials and Pt-carbon interactions on the catalytic performance of catalysts. The main product was DHMF over all the Pt catalysts prepared by impregnation, and Pt/BC-IM showed the highest turnover frequency and the highest selectivity to DHMF (95.3%) with an HMF conversion of 94.6% after 10 h of reaction time, which could be due to the abundant functional groups on the surface of BC. Compared with the Pt/CNTs-IM catalyst, the main product over the Pt/CNTs-ALD catalyst was DMF with a yield of 47% due to the relatively strong Pt-CNT interaction. The interaction between Pt and CNTs is the dominant factor in determining the main product in the selective HMF hydrogenation reaction. After introducing Ni nanoparticles to Pt/CNTs-ALD, the yield of DMF reached 93.4% over the PtNi/CNTs-ALD bimetallic catalyst owing to the synergistic effect between Pt and Ni.

Full text of DOI:10.1039/d0cy01920h

View Article Online
View Journal
Catalysis
Science &
Technology
Accepted Manuscript
This article can be cited before page numbers have been issued, to do this please use: X. Wang, C.
Zhang, B. Jin, X. Liang, Q. Wang, Z. Zhao and Q. Li, Catal. Sci. Technol., 2021, DOI:
10.1039/D0CY01920H.
Volume
Number
21 January 2017
Pages 313-534
7
2
This is an Accepted Manuscript, which has been through the
Royal Society of Chemistry peer review process and has been
accepted for publication.
Catalysis
Science &
Technology
rsc.li/catalysis
Accepted Manuscripts are published online shortly after acceptance,
before technical editing, formatting and proof reading. Using this free
service, authors can make their results available to the community, in
citable form, before we publish the edited article. We will replace this
Accepted Manuscript with the edited and formatted Advance Article as
soon as it is available.
You can find more information about Accepted Manuscripts in the
Information for Authors.
Please note that technical editing may introduce minor changes to the
text and/or graphics, which may alter content. The journal’s standard
Terms & Conditions and the Ethical guidelines still apply. In no event
shall the Royal Society of Chemistry be held responsible for any errors
or omissions in this Accepted Manuscript or any consequences arising
from the use of any information it contains.
ISSN 2044-4761
PAPER
Chaoqiu Chen, Yong Qin et al.
Highly dispersed Pt nanoparticles supported on carbon
nanotubes produced by atomic layer deposition for hydrogen
generation from hydrolysis of ammonia borane
rsc.li/catalysis

Page 1 of 41
Catalysis Science & Technology
View Article Online
DOI: 10.1039/D0CY01920H
1
2
Pt-carbon interactions-determined reaction pathway and selectivity for
hydrogenation of 5-hydroxymethylfurfural over carbon supported Pt
catalysts
3
4
Xiaofeng Wang,a,* Chengcheng Zhang,a Baitang Jin,b Xinhua Liang,b Qingfa Wang,c Zhe
Zhao,a Qingbo Lia
5
a
6
College of Environmental Science and Engineering, Dalian Maritime University, Dalian
7
116026, China
8
b Department of Chemical and Biochemical Engineering, Missouri University of Science and
Technology, Rolla, MO 65409, United States
9
10
11
12
13
14
15
16
17
18
19
20
21
22
c Key Laboratory of Green Chemical Technology of Ministry of Education, School of Chemical
Engineering and Technology, Tianjin University, Tianjin 300072, China
* Corresponding author, email: Xiaofeng Wang, wangxf@dlmu.edu.cn
Abstract
In this study, a series of Pt catalysts supported on carbon materials, including multi-walled
carbon nanotubes (CNTs), reduced graphene oxide (rGO), graphitic carbon nitride (g-C3N4),
and biochars (BC), were prepared by impregnation or atomic layer deposition (ALD) methods.
The Pt content was around 2.5 wt.% in all Pt catalysts, and the average Pt particle size was 1.6-
1.9 nm for the catalysts prepared by impregnation (IM), which was slightly larger than that for
the Pt catalyst (~1.4 nm) synthesized by ALD (Pt/CNTs-ALD). All these catalysts were used
in the hydrogenation of HMF reaction to investigate the effects of different carbon materials
and Pt-carbon interactions on the catalytic performance of catalysts. The main product was
1

Catalysis Science & Technology
Page 2 of 41
View Article Online
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
DHMF over all Pt catalysts prepared by impregnation, and Pt/BC-IM showedDOthI:e10.h10i3g9h/De0sCtY01920H
turnover frequency and the highest selectivity to DHMF (95.3%) with an HMF conversion of
94.6% after 10 hr of reaction time, which could be due to the abundant functional groups on
the surface of BC. Compared with the Pt/CNTs-IM catalyst, the main product over the
Pt/CNTs-ALD catalyst was DMF with a yield of 47% due to the relatively strong Pt-CNTs
interaction. The interaction between Pt and CNTs is the dominant factor to determine the main
product in the selective hydrogenation of HMF reaction. After introducing Ni nanoparticles to
Pt/CNTs-ALD, the yield of DMF achieved 93.4% over the PtNi/CNTs-ALD bimetallic catalyst
owing to the synergistic effect between Pt and Ni.
Keywords: 5-Hydroxymethylfurfural; Hydrogenation; 2,5-Dimethylolfuran; Pt-based
catalysts; Interactions; 2,5-Dishydroxymethylfuran
1. Introduction
With the shrinking of fossil resources, efficient synthesis of renewable fuel and high value-
added chemicals from biomass has recently attracted global attention [1, 2]. 5-
hydroxymethylfurfural (HMF), which can be derived from biomass, is one of the most
important platform chemicals [3-5]. Since it contains several functional groups, including a
furan ring, an aldehyde group, and a hydroxymethyl group, HMF can be converted to valuable
chemicals and fuels, such as 2,5-bishydroxymethylfuran (DHMF) [6, 7], 5-methyl furfural (5-
MF) [8], 2,5-dimethylfuran (DMF) [9-12], and 5-bis(hydroxymethyl) tetrahydrofuran
(DMTHF) [13-15], as shown in Figure 1. Among these chemicals, DMF is a promising and
2

Page 3 of 41
Catalysis Science & Technology
View Article Online
DOI: 10.1039/D0CY01920H
45
46
47
48
49
50
51
potential biofuel due to its high energy density (30 MJ/L) and high research octane number
(RON=119) [16], and DHMF is an important chemical diol that can be used as a precursor for
functional polymers, resins, artificial fibers, drugs, and high-valued fine chemicals [17, 18].
Thus, the high yield of DHMF and DMF in the catalytic hydrogenation of HMF is pursued by
researchers, and it is the key to design and fabricate the catalysts with high activity and high
selectivity to target product.
52
53
Figure 1. Proposed reaction pathways for the conversion of HMF. Compounds: 5-
54
55
56
57
58
59
60
61
62
(hydroxymethyl)furfural (HMF), 2,5-bis(hydroxymethyl)furan (DHMF), 5-methyl furfural (5-
MF),
5-methyl
furfuryl
alcohol
(MFA),
2,5-dimethylfuran
(DMF),
5-
bis(hydroxymethyl)tetrahydrofuran (DHTHF), and 2,5-dimethyltetrahydrofuran (DMTHF).
As shown in Figure 1, DHMF is synthesized from hydrogenation of HMF, and DMF will
be produced after further hydrogenolysis over various supported metal catalysts, including Pt-,
Ru-, and Cu-based catalysts [19, 20]. Among them, some metal catalysts supported on carbon
materials with outstanding catalytic performance have been reported and high yield of target
product has been achieved [21-24]. For example, Yu et al. reported a robust carbon nanotubes
3

Catalysis Science & Technology
Page 4 of 41
View Article Online
DOI: 10.1039/D0CY01920H
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
(CNTs) supported Ni-Fe catalyst and the selectivity to DHMF was as high as 96% at 110 C
[21]. Yang and co-workers prepared Ni-Co/C bimetallic catalysts by a wetness impregnation
method, and up to 95% of DMF yield was obtained at 130 °C and 10 bar H2. In our previous
study, Pt-Co/CNTs bimetallic catalysts were prepared by atomic layer deposition (ALD), and
more than 90% of DMF yield was achieved with a complete conversion of HMF [25]. Thus,
carbon materials were reported as efficient supports for selective hydrogenation of HMF
reaction. However, to the best of our knowledge, no study investigates the effects of carbon
supports and the interaction between Pt and carbon supports on the selective hydrogenation of
HMF to target products, and the catalytic mechanism is not well revealed.
Herein, in this study, a series of carbon materials supported Pt catalysts with an ultra-small
Pt particle size were prepared by impregnation to investigate the effects of carbon supports,
including multi-walled carbon nanotubes (CNTs), reduced graphene oxide (rGO), graphitic
carbon nitride (g-C3N4), and biochars (BC), in the HMF hydronation reaction. DHMF was the
main product over all catalysts, and Pt/BC-IM showed the highest yield of DHMF (> 90%)
with a selectivity of >95%. At various reaction conditions, the main product was still DHMF,
though the HMF conversion and DHMF selectivity varied, and almost no further
hydrogenation/hydrogenolysis reaction occurred. But interestingly, when we tuned the Pt-
carbon interactions through preparing the Pt/CNTs catalyst by ALD, the main product changed
to DMF over the Pt/CNTs-ALD catalyst with similar Pt loading and particle size. Thus, the
interactions between Pt and carbon supports are the dominant factor to determine the main
product in the selective hydrogenation of HMF reaction for Pt catalysts.
84
4

Page 5 of 41
Catalysis Science & Technology
View Article Online
DOI: 10.1039/D0CY01920H
85
86
2. Materials and methods
2.1. Chemicals and materials
87
CNTs (>95%, OD 10-20 nm) and GO were purchased from Nanjing XFNANO Materials Tech
Co Ltd. HMF (98%), DHMF (97%), DMF (99%), 1-butanol (99.5%), n-Tridecane (99%) were
purchased from J&K Chemicals, Ltd, and methylcyclopentadienyl)trimethyl platinum
[(MeCp)PtMe3] (99%) and bis(cyclopentadienyl)nickel (NiCp2, 99%) was obtained from
Strem Chemicals, Inc. Biochar and g-C3N4 were prepared in the lab and introduced in details
in the following section.
88
89
90
91
92
93
94
2.2. Preparation of carbon supports
95
Biochar (BC) was prepared by slow pyrolysis of rice husks, as described in details previously
[26, 27]. Briefly, a certain amount of rice husks was added in a thermal cracking reactor, and
then the pyrolysis temperature increased from room temperature to 550 C at a rate of 10
C/min. Finally, the obtained BC samples were crushed to pass through a 50-mesh sieve before
further processing. g-C3N4 was synthesized by heating urea at 550 C for 2 hours in a semi-
closed quartz crucible placed inside muffle furnace.
96
97
98
99
100
101
102
103
104
105
106
2.3. Preparation of Pt-based catalysts
Pt catalysts supported on carbon materials were prepared by a simple impregnation method.
Briefly, 0.1 g of H2PtCl6 was firstly dissolved in 80 mL of deionized water and then 0.5 g of
supports were added. After stirring to dry at 70 C, the obtained samples were washed three
times with ethanol. Then, the samples were dried overnight at 60 C. At last, the samples were
5

Catalysis Science & Technology
Page 6 of 41
View Article Online
DOI: 10.1039/D0CY01920H
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
reduced by flowing a 5 vol.% H2/Ar mixture at 400 C with a flow rate of 120 mL/min for 2
hours. The catalysts are labeled as Pt/CNTs-IM, Pt/rGO-IM, Pt/g-C3N4-IM, and Pt/BC-IM with
different supports.
In order to adjust the Pt-CNTs interaction and study the effects of the interaction on the
catalytic performance, ALD was adopted to prepare another Pt/CNTs catalyst (labeled as
Pt/CNTs-ALD) with similar Pt loading and particle size, and the scheme of the ALD reactor is
shown in Figure S1. 1 cycle of Pt ALD was applied on CNTs at 300 C, and
(methylcyclopentadienyl)trimethyl platinum [(MeCp)PtMe3] and oxygen (O2) were used as the
precursors. During the ALD process, (MeCp)PtMe3 reacted with functional groups (e.g.,
hydroxyl groups) on the surface of the CNTs, and thus the chemical bonding formed between
Pt and CNTs with strong interactions [28]. Since the burning point of BC is relatively low and
O2 is a precursor in the ALD process, Pt ALD was not applied for the BC. In order to further
optimize the ALD-prepared catalyst and enhance the catalytic performance, 3 cycles of Pt ALD
and 1 cycle of Ni ALD were applied on CNTs to prepare a PtNi bimetallic catalyst (labeled as
PtNi/CNTs-ALD). For Ni ALD, NiCp2 and H2 were used as precursors and the reaction
temperature was 300 C. The details of Pt ALD and Ni ALD are described in the
Supplementary Information (SI).
2.4. Catalyst characterizations
The Brunauer-Emmett-Teller (BET) surface area and pore size distribution of carbon supports
were analyzed by Micromeritics ASAP 2020. The Pt and Ni contents in all catalysts before and
after HMF hydrogenation reactions were measured by inductively coupled plasma atomic
6

Page 7 of 41
Catalysis Science & Technology
View Article Online
DOI: 10.1039/D0CY01920H
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
emission spectroscopy (ICP-AES). The size and distribution of metal NPs on the supports were
directly observed by high-resolution transmission electron microscopy (HRTEM, FEI-Tecnai
G2 F20). The crystal structure and the functional groups of the catalysts were analyzed by X-
ray powder diffraction (XRD) and Fourier transform infrared spectrometer (FT-IR),
respectively. X-ray photoelectron spectroscopy (XPS) was carried out for all Pt catalysts. The
reduction properties of the catalysts were evaluated by H2 temperature programmed reduction
(H2-TPR) using Micromeritics Autochem II 2920. The details of BET analysis, HRTEM, XRD,
XPS, and H2-TPR are described in the SI.
2.5. Hydrogenation of HMF
The HMF hydrogenation reactions were performed in a stainless-steel reactor, which was
described in details previously [25]. Typically, first, a certain amount of catalyst was added
into the reactor. Then, 0.5 g HMF, 0.05 g n-Tridecane (internal standard), and 40 mL of 1-
butanol were introduced to the reactor. In order to remove air in the reactor, the reactor was
first pressurizedto 10 bar with H2 and then depressurized to atmospheric pressure for ten times.
Then, a certain H2 pressure (10-30 bar) was applied in the reactor and the temperature was
increased to a setpoint (160-180 C). After the temperature reached the setpoint, the reaction
started with a slurry stirring speed of 500 rpm for all experiments. At certain time intervals, 1-
1.5 mL sample was collected, filtered, and analyzed by a gas chromatograph with a flame
ionization detector (GC-FID, Agilent Technologies Inc., 6890). The individual components
were separated using a 30-meter DB-5 capillary column. The initial column temperature was
set at 70 ºC and programmed to 250 ºC at 20 ºC/min, then the temperature was maintained at
7

Catalysis Science & Technology
Page 8 of 41
View Article Online
DOI: 10.1039/D0CY01920H
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
250 °C for 4 min. Both the injector temperature and the FID temperature were set at 250 ºC.
After reaction, the used catalysts were collected, washed by 1-butanol for 3 times, and dried at
60 C in an oven over night for cycling tests.
Besides, in order to study the effect of carbon supports on the catalytic performance, we
evaluated the adsorption of HMF by the supports. Briefly, in each test, 0.1 g carbon supports
were added in a 1-butanol solution with 0.2 g HMF. At 10-60 min intervals, 1 mL sample was
collected, filtered, and analyzed by a GC-FID to get the adsorption rate of HMF.
3. Results and discussion
3.1. Characterizations of Pt catalysts
As shown in Figure 2 and Table 1, all four carbon supports had porous structure, and CNTs
had the highest surface area (197 m2 g-1) with a total pore volume of 0.79 cm3 g-1, followed by
g-C3N4 (105 m2 g-1, 0.62 cm3 g-1), rGO (65 m2 g-1, 0.32 cm3 g-1), and BC (9 m2 g-1, 0.27 cm3 g-
1). Compared with the surface areas of the rGO samples in previous reports [29, 30], the low
surface area of the rGO samples in this study could be due to the accumulation of graphite
layers. From the pore size distribution in Figure 2b, all samples showed a sharp peak centered
around 3.3-3.6 nm. In addition, both CNTs and g-C3N4 showed another broad peak around 5-
25 nm and 16-50 nm, respectively. Based on the ICP-AES analysis, the Pt contents in all
catalysts were between 2.2 wt.% and 2.6 wt.%. Though the supports had different surface areas
and pore volumes, the initial volume of platinum precursor solution (80 mL) was much larger
than the pore volume of the supports, so the surface properties of the supports had little effect
on the Pt loading in the impregnation method. The similar phenomenon was also reported by
8

Page 9 of 41
Catalysis Science & Technology
View Article Online
DOI: 10.1039/D0CY01920H
173
174
175
Zhang et al. [31]. The Pt and Ni loading of PtNi/CNTs-ALD were 8.6 wt.% and 1.5 wt.%,
respectively.
(a)
500
CNTs
rGO
400
g-C3N4
300
BC
200
100
0
0
0.2
0.4
0.6
P/P0
0.8
1
(b)
0.005
0.004
0.003
0.002
0.001
0
CNTs
rGO
g-C3N4
BC
20
200
Pore width, Å
2000
176
177
Figure 2. (a) N2 adsorption and desorption isotherms, and (b) mesopore size distributions of
178
179
180
carbon supports.
Table 1. Properties of carbon supports.
CNTs
197
rGO
65
g-C3N4
105
BC
9
BET surface area (m2 g-1)
Total pore volume (cm3 g-1) 0.79
Pt loading (wt.%)
2.6a/2.3b/8.6c
0.32
2.2
0.62
2.5
0.027
2.6
9

Catalysis Science & Technology
Page 10 of 41
View Article Online
DOI: 10.1039/D0CY01920H
Ni loading (wt.%)
Pt particle size (nm)
Pt dispersion (%)
1.5c
-
-
-
1.6a/1.4b/1.7c
71a/81b/67c
1.8
63
1.9
60
1.9
60
181
182
183
184
185
186
187
188
189
190
191
192
193
a Pt/CNTs-IM catalyst; b Pt/CNTs-ALD catalyst; c PtNi/CNTs-ALD catalyst.
HRTEM images of the catalysts prepared by the impregnation method are presented in
Figure 3. Ultra-small Pt NPs were highly dispersed on four supports, and the average Pt particle
size was similar (1.6-1.9 nm) for all four catalysts (Table 1). Figure 4 shows HRTEM images
of Pt/CNTs-ALD and PtNi/CNTs-ALD samples, and metal NPs were deposited on CNTs with
an average particle size of 1.4-1.7 nm, which was similar to that on CNTs prepared by the
impregnation method. The result is in agreement with our previous study [32]. Some “twinned”
NPs were observed and circled in Figure 4d, indicating the formation of Pt-Ni bimetallic NPs.
To verify the bimetallic structure of Pt-Ni, STEM-EDS analysis was performed for a single
nanoparticle (NP). As shown in Figure S2, the result of elemental composition indicates the
formation of the Pt-Ni bimetallic NP prepared by ALD.
10

Page 11 of 41
Catalysis Science & Technology
View Article Online
DOI: 10.1039/D0CY01920H
(d)
(b)
(a)
(c)
(f)
(g)
(h)
(e)
194
195
Figure 3. Low magnification and high magnification HRTEM images of Pt/CNTs-IM (a and
e), Pt/rGO-IM (b and f), Pt/g-C3N4-IM (c and g), and Pt/BC-IM (d and h) catalysts prepared
by the impregnation method, respectively.
196
197
198
11

Catalysis Science & Technology
Page 12 of 41
View Article Online
DOI: 10.1039/D0CY01920H
(b)
(a)
20 nm
(c)
(d)
5 nm
199
200
Figure 4. Low magnification and high magnification HRTEM images of Pt/CNTs-ALD (a and
c) and PtNi/CNTs-ALD (b and d) samples.
201
202
203
204
205
206
207
208
209
210
In order to obtain the turnover frequency (TOF) and evaluate the catalytic activity, Pt
dispersion on different catalysts is required, which is calculated using the following two
equations [33, 34]:
퐿 × 푓
푔) = 푑 × 푧
푚2
(1)
(2)
active surface area (ASA,
퐴푆퐴 × 푀
dispersion (%) =
푁 × 퐴 × 퐿
where L is Pt content (wt.%), f is the particle shape correction factor (f = 6 for a sphere), d is
the average diameter of Pt NPs measured by HRTEM, z is Pt density, M is the formula weight
of Pt, N is the Avogadro’s number, and A is metal atomic surface area (for Pt, A= 8 Å2/atom).
12

Page 13 of 41
Catalysis Science & Technology
View Article Online
DOI: 10.1039/D0CY01920H
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
As shown in Table 1, Pt dispersion of the catalysts prepared by impregnation was 60-70%, and
that of the Pt/CNTs-ALD catalyst reached 81%.
FT-IR spectra of all samples were shown in Figure 5a. The FT-IR spectra of Pt/g-C3N4-IM
exhibited several conventional peaks of g-C3N4 at about 3,000-3,600, 1,200-1,650, and 812 cm-
1. The broad band in the region of 3,000-3,600 cm-1 could be assigned to the stretching vibration
of N-H groups [35]. The absorption bands of 1,200-1,650 cm-1 contributed to the aromatic C-
N stretching vibrations [36, 37]. The peak at 812 cm-1 was considered to be the characteristic
breathing mode of the triazine units [35, 36]. For the Pt/CNTs-IM, Pt/rGO-IM, and Pt/BC-IM
catalysts, the peak of 3,423 cm-1 contributed from the hydroxyl groups (-OH) [38-40], besides
the obtained peaks at around 2,920 cm−1 that are linked to the stretching of a methylene group,
indicating the presence of methylene groups (CH) on the surface of Pt/CNTs-IM and Pt/rGO-
IM [41, 42]. The band assigned to the skeletal vibration of C=C and C=O is shown at around
1,633 cm−1 [43], and the peaks located at 1,080 and 807cm-1 are ascribed to C-O and C-H
stretching vibration [38, 44].
13

Catalysis Science & Technology
Page 14 of 41
View Article Online
DOI: 10.1039/D0CY01920H
(a)
(b)
226
227
Figure 5. (a) FT-IR and (b) XRD spectra of Pt catalysts supported on different carbon materials.
228
229
230
231
232
233
234
235
236
237
XRD profiles of the Pt catalysts are presented in Figure 5b. Pt/rGO-IM exhibits two
diffraction peaks at 30.9° and 64.2°, corresponding to the (0 0 2) and (0 0 4) crystal planes of
graphite-2H (JCPDS Card No.41-1487), respectively. For the Pt/CNTs-IM catalyst, there are
two peaks at 30.2° and 50.3°, corresponding to the (0 0 2) and (1 0 0) crystal planes of graphite,
respectively, according to JCPDS Card No.08-0415. The XRD pattern of Pt/g-C3N4-IM at 32.2°
corresponds to the (1 1 0) of carbon nitride according to JCPDS Card No.50-1512. These
characteristic peaks indicate that the supports kept typical structure in the preparation of
catalysts. In the XRD pattern of Pt/BC-IM, the diffraction peak at 47° corresponds to the (1 1
1) plane of Pt, owing to better crystallinity, while the diffraction peaks related to Pt phase were
14

Page 15 of 41
Catalysis Science & Technology
View Article Online
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
not detected on Pt/CNTs-IM, Pt/rGO-IM, and Pt/g-C3N4-IM. In addition, the Pt lattDicOeI: 1s0t.r10u3c9t/uD0rCeY01920H
can be observed in the HRTEM images, as shown in Figures 3e-3h.
XPS analysis was carried out to examine the composition and valence state of Pt-based
catalysts, as shown in Figures 6a and S3-S5. For the Pt/CNTs-IM, Pt/rGO-IM, Pt/g-C3N4-IM
and Pt/BC-IM catalysts, the Pt0 4f7/2 binding energies (BE) were around 71.4, 71.1, 71.5 and
71.4 eV, respectively, and the Pt0 4f5/2 BE were around 74.8, 74.5, 74.9 and 74.8 eV,
respectively. The rGO contains large amount of graphitic carbon structure, and the main reason
of the lower Pt0 BE for the Pt/rGO-IM catalyst could be ascribed to the π-electron transfer from
the graphitic carbon to Pt NPs [45]. The slightly higher Pt0 BE of the Pt/g-C3N4-IM catalyst
could be due to the fact that the outer electrons of Pt transferred from the Pt NPs to the g-C3N4
containing a large amount of N element with high electronegativity [46]. For these Pt catalysts,
the ratios of Pt0/PtII were around 3:1 (Table S1). The Pt XPS spectrum of Pt/CNTs-ALD
displayed two Pt0 characteristic peaks at 71.9 and 75.2 eV, corresponding to Pt 4f7/2 and 4f5/2,
respectively, which exhibited a higher binding energy owing to the chemical bonding between
Pt NPs and CNTs. In addition, for all Pt catalysts, the peaks of PtIIO were weak, implying that
most of Pt NPs remained metallic state. For PtNi/CNTs-ALD, the binding energy at around
71.5 eV and 75.1 eV are the peaks of Pt7/2 and Pt5/2 owing to the presence of zero-valent Pt0,
and the Pt7/2 and Pt5/2 position of PtIIO were around 72.8 eV and 76.0 eV, respectively.
Compared to that of the Pt/CNTs-ALD catalyst, the d-band center of the PtNi/CNTs-ALD
catalyst upshifted, and such a tendency implies the formation of Pt-Ni bimetallic structure and
an electron transferred from Ni to Pt [47, 48]. In addition, the Pt loading in the PtNi/CNTs-
ALD catalyst (8.7 wt%) was higher than that in the Pt/CNTs-ALD catalyst (2.3 wt%), and
15

Catalysis Science & Technology
Page 16 of 41
View Article Online
DOI: 10.1039/D0CY01920H
260
261
thereby more Pt NPs could be oxidized in the air, resulting in the increase of PtIIO, as Figure
6a exhibited.
(a)
(b)
262
263
Figure 6. (a) XPS spectra and (b) TPR analysis of Pt-based catalysts supported on different
264
265
266
267
268
269
270
carbon materials.
In addition, the XPS spectra of C 1s and O 1s for all catalysts presented the functional
groups on the surface of the supports (Figures S4-S5). The high resolution spectra of C 1s could
be deconvoluted into three peaks, corresponding to sp2 and sp3 (C=C and C-C 284.4 eV),
carbon-oxygen bond (C-O 285.5 eV), and carbonyl groups (C=O 286.8 eV), respectively [43],
and the ratios of these groups were presented in Table S1. Moreover, for the Pt/g-C3N4-IM
16

Page 17 of 41
Catalysis Science & Technology
View Article Online
271
272
273
274
275
276
277
278
279
280
281
282
283
catalyst, the N 1s peaks at 398.4, 399.7, 401.0 and 404.3 eV can be assigned toDOthI:e10N.10-3(9C/D)03C,Y01920H
C=N-C, N-H, and π excitations, respectively (Figure S5b) [43, 49]. The catalysts exhibited four
peaks of O 1s at 529.5, 531.1, 532.3, and 534.9 eV, which were assigned to PtIIO, C=O, C-O
and -OH, respectively [43, 44]. Notably, the O1s peak area for the Pt/BC-IM catalyst was much
larger than those of other supports, which indicated higher amounts of functional groups on the
BC. During the hydrogenation of HMF to DMF, these functional groups may assist the
activation C=C of HMF, thus enhance the selectivity of HMF to DHMF.
Figure 6b shows H2-TPR profiles of Pt-based catalysts. For Pt/BC-IM catalyst, there was
a broad and strong peak at around 700℃, attributed to the functional groups (C-O and C=O)
on the support [50]. The large number of functional groups was consistent with the XPS and
FT-IR results. A broad peak and a weak peak on 550 ℃ and 525 ℃ were observed for both
Pt/CNTs-IM and Pt/CNTs-ALD catalysts, indicating that there were only a few Pt NPs
oxidized in the air, which was consistent with the XPS results. The PtNi/CNTs-ALD presented
an apparent peak centered at 435 °C, which could be owing to the H2 consumption of the
oxidized Ni NPs, indicating that part of Ni NPs was oxidized in air after Ni ALD.
284
285
286
287
288
289
290
291
292
3.2. Catalysts performance
3.2.1. Role of different carbon supports
The catalytic performance of Pt catalysts supported on different carbon materials is presented
in Figures 7 and S6. The main product was DHMF over all four catalysts, and the conversion
of HMF reached 97.2% with a selectivity to DHMF of 70.4% over Pt/g-C3N4-IM. A similar
conversion (95.5%) and selectivity to DHMF (68.1%) over Pt/rGO-IM was observed. The
17

Catalysis Science & Technology
Page 18 of 41
View Article Online
DOI: 10.1039/D0CY01920H
293
294
295
296
297
298
299
300
301
302
303
304
Pt/BC-IM catalyst showed the highest selectivity to DHMF (95.3%), and the highest yield
(90.1%) of DHMF was obtained. Pt/CNTs-IM showed the lowest HMF conversion of 72.7%
with a DHMF selectivity of 49.4%. Figure S6 presents the product distribution over different
Pt based catalysts along with reaction time, and the reaction pathway (hydrogenation of HMF
to DHMF) was the same over all four Pt catalysts. The main byproduct of the reaction at 10
hrs was 5,5-(oxybis(methylene))bis(2-methylfuran) (OMBM), as we reported previously [25].
In addition, the TOF was calculated with a HMF conversion of 20%, and that of Pt/CNTs-IM,
Pt/rGO-IM, Pt/g-C3N4-IM, and Pt/BC-IM catalysts was 5.5, 9.4, 7.9, and 9.2 min-1, respectively.
The TOF of Pt/CNTs-IM was much lower than the other three catalysts, which could be related
to the inert surface of CNTs. As shown in Table 1, all of four catalysts had a similar Pt loading
and Pt particles size, however, they presented different conversions and selectivities to DHMF.
This indicated that supports may play an important role in the catalytic performance.
305
306
Figure 7. HMF conversion, selectivity to different products, and TOF over different Pt
catalysts. Reaction conditions: 0.5 g HMF, 0.05 g n-Tridecane, 40 mL 1-butanol, 50 mg
catalyst, 160 °C, 10 bar H2, 10 hrs, and 500 rpm. The TOF was calculated at a HMF conversion
307
308
309
of 20%.
18

Page 19 of 41
Catalysis Science & Technology
View Article Online
DOI: 10.1039/D0CY01920H
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
Among four catalysts, Pt/g-C3N4-IM and Pt/rGO-IM showed the highest conversion of
HMF, attributable to their high BET surface areas and strong HMF adsorption ability of the
supports. It is reported that the conversion of HMF could rely on the adsorption of it for
hydrogenation reaction [51]. To investigate the adsorption ability of supports for HMF,
adsorption experiments were performed, and the results are shown in Figure 8. For rGO and g-
C3N4, they exhibited higher adsorption capacity of HMF, which could be attributed to the fact
that their high BET surface areas (65 and 105 m2g-1) and pore volume (0.32 and 0.62 cm3g-1)
could increase the adsorption of HMF, leading to a high conversion of HMF. Similar
phenomena were also observed in other reactions [52-54]. Although CNTs had the highest BET
surface area (197 m2g-1) and pore volume (0.79 cm3g-1), the adsorption ability of HMF over
CNTs was the weakest owing to the relatively inert surface and fewer functional groups. In
contrast, though BC has the lowest BET surface area and pore volume, various functional
groups on BC could be beneficial to improve the adsorption ability for HMF and, thereby,
promote the conversion of HMF [55]. Moreover, the high selectivity to DHMF over Pt/BC-IM
could be owing to the functional groups on BC (e.g., hydroxyl (-OH), carbonyl (C=O), carboxyl
(-COOH), and alkoxy (C-O-C)).
19

Catalysis Science & Technology
Page 20 of 41
View Article Online
DOI: 10.1039/D0CY01920H
328
329
Figure 8. Adsorption of HMF over different carbon supports. Conditions: 0.2 g HMF, 0.05 g
n-Tridecane, 25 ℃, 20 mL 1-butanol, 0.1g support, and 500 rpm.
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
In addition, the 2D plane structure and surface functional group of g-C3N4 and rGO played
a vital role in hydrogenation of HMF to DHMF. As shown in Figure 5b, the XRD patterns of
Pt/g-C3N4-IM and Pt/rGO-IM exhibited diffraction peaks at 32.2° and 30.9°, attributed to (1 1
0) crystal plane of carbon nitride and (0 0 2) crystal graphite-2H, respectively. It is reported
that flat structure is more conducive to the adsorption of HMF [55]. Besides, the FT-IR spectra
of Pt/g-C3N4-IM and Pt/rGO-IM provided another evidence of the existence of graphite
structure and g-C3N4 structure. As shown in Figure 5a, the aromatic C-N (band of 1200-1650
cm-1) and triazine units (band of around 812 cm-1) over Pt/g-C3N4-IM were the characteristic
components of g-C3N4. The function group of C-H (band of around 2920cm-1) and aromatic
C=C (band of around 1,633 cm-1) over Pt/rGO-IM proved the existence of graphitic carbon. In
contrast, the one-dimensional shape of CNTs played a negative role in HMF adsorption on the
substrate. CNTs, that have the highest BET surface area and pore volume, showed a low
activity due to the 1D inert CNTs structure and a part of Pt NPs loading inside CNTs, as
20

Page 21 of 41
Catalysis Science & Technology
View Article Online
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
observed by HRTEM (Figure 3a). Moreover, the XRD spectra of Pt/CNTs-IMD(OFI:i1g0u.10r3e9/5Db0C)Y01920H
show a weaker diffraction peaks of (0 0 2) crystal graphite-2H than that of Pt/rGO-IM,
indicating that there was less graphite carbon over Pt/CNTs-IM, and this could affect the
catalytic performance of Pt/CNTs-IM.
As mentioned above, both the XPS analysis and FT-IR spectra of the Pt/BC-IM catalyst
verified the existence of functional groups, such as -OH, -COOH [56], C=O, and aromatic
C=C. These oxygen-containing functional groups could enhance the adsorption of HMF. In a
previous report, the transfer of π-electron from carbon support C=C to metal sites could
enhance the selectivity of C=O [25]. In addition, there was a large peak at 1,080 cm-1,
corresponding to alkoxyl C-O (1,080 cm-1) [56], and the alkoxyl group could be considered to
attract electronic, which is beneficial to adsorb HMF and enhance the activation of C=O in
HMF. In addition, the small amount of -COOH released H+ could enhance hydrogenation of
C=O in HMF. Thus, the functional groups played an important role in enhancing both the
conversion of HMF and the selectivity to DHMF, resulting in a high DHMF yield of 90.1% for
the Pt/BC-IM catalyst.
In order to figure out whether the reaction conditions dominated the produced main
product or not, a series of reactions under different reaction conditions, including temperature,
H2 pressure, reaction time, and catalysts dose, were performed and the results are presented in
the following section.
3.2.2. Effects of reaction conditions
21

Catalysis Science & Technology
Page 22 of 41
View Article Online
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
In this study, the Pt/BC-IM and Pt/CNTs-IM catalysts were used to study the effectDsOoI:f10r.e10a3c9t/iDo0nCY01920H
conditions on HMF conversion and product distribution, since the highest and lowest yields of
DHMF were obtained over these two catalysts. As shown in Figures 9 and S7-S8, the
conversion of HMF was 71.2 % and the selectivity of DHMF increased to 96.9% at 140 ℃
over the Pt/BC-IM catalyst. A lower temperature was beneficial to the formation of DHMF
from HMF but not conducive to activate HMF [57].When the temperature increased from 140 ℃
to 160 ℃, the conversion of HMF reached 94.6% and the selectivity of DHMF reached 95.3%.
As the reaction temperature further increased to 180 ℃, the conversion of HMF increased to
96.4%, but the selectivity decreased to 88.1%. For the Pt/CNTs-IM catalyst, the conversion of
HMF increased from 72.7% to 98.1% and the selectivity of DHMF decreased from 49.4% to
42.1%, when temperature increased from 160 ℃ to 180 ℃. The HMF conversion increased and
the selectivity to DHMF decreased with the increase of the temperature for both catalysts,
indicating that a high reaction temperature would accelerate the reaction process; however, the
main product was still DHMF. At 180 °C, the selectivity to both DMF and other byproducts
increased for Pt/CNTs-IM. It implies that more side reactions occurred at high temperature,
including ring-opening, dehydration, and hydrogenolysis reactions, which is consistent with a
previous report [25].
22

Page 23 of 41
Catalysis Science & Technology
View Article Online
DOI: 10.1039/D0CY01920H
384
385
Figure 9. Effects of reaction temperature on the conversion of HMF and selectivity to products.
Reaction conditions: 0.5 g HMF, 0.05 g n-Tridecane, 40 mL 1-butanol, 50 mg catalyst, 10 bar
H2, 10 hrs, and 500 rpm.
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
10, 20, and 30 bar of H2 pressure were applied for hydrogenation of HMF reaction, and
the results are presented in Figures 10 and S9-S10. For the Pt/BC-IM catalyst, the HMF
conversion increased along with the increase of H2 pressure, and it was as high as 98.1% at 30
bar of H2, while the selectivity to DHMF decreased to 79.6% due to the occurrence of more
side reactions. In contrast, both the HMF conversion and DHMF selectivity increased over
Pt/CNTs-IM with increasing reaction pressure from 10 bar to 30 bar. This could be explained
by the fact that a high pressure was beneficial to dissolve H2 in solvent, which could promote
hydrogenation of HMF and cause further hydrogenation to convert more byproducts. In
addition, when the reaction time was prolonged to 12 hrs, the selectivity to DHMF decreased
and the selectivity to other byproducts increased over the Pt/BC-IM catalyst (Figure S11). As
presented in Figures S12-S13, 50 mg and 100 mg of Pt/BC-IM catalysts were adopted, and the
byproducts increased greatly when 100 mg of catalyst was used. These results indicate that
23

Catalysis Science & Technology
Page 24 of 41
View Article Online
DOI: 10.1039/D0CY01920H
401
402
403
404
405
406
407
408
409
both the prolonging reaction time and increase of catalyst amount could not promote the
conversion of DHMF to DMF.
When the reaction condition got harsher (e.g., higher H2 pressure and temperature, more
catalyst dose, and longer reaction time), the main product was still DHMF and the
hydrogenolysis of DHMF to DMF reaction did not occur over these Pt catalysts prepared by
the impregnation method, implying that the reaction pathway did not change via adjusting the
reaction conditions, and the interaction between Pt NPs and supports is the dominant factor to
determine the selectivity of HMF hydrogenation products.
410
411
Figure 10. Effects of H2 pressure on the conversion of HMF and selectivity to products.
Reaction conditions: 0.5 g HMF, 0.05 g n-Tridecane, 40 mL 1-butanol, 50 mg catalyst, 160 ℃,
10 hrs, and 500 rpm.
412
413
414
415
416
417
3.2.3. Effects of metal-support interaction
As discussed above, the main product was DHMF for all carbon supported Pt catalysts prepared
by the impregnation method under different reaction conditions. Thus, the reaction conditions
24

Page 25 of 41
Catalysis Science & Technology
View Article Online
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
are not the dominant factor to determine the main product. In other words, the reactDioOnI: 1p0.a10th39w/Da0yCY01920H
and the formation of main product depend on the property of catalysts, especially the
interaction between Pt and supports. The main products were DHMF over these catalysts
prepared by the impregnation method, which indicated that the catalyst prepared by the
impregnation method had similar metal-support interaction. Herein, we prepared another
Pt/CNTs-ALD catalyst via ALD method to investigate the effect of the interaction between Pt
and CNTs on the hydrogenation of HMF reaction, since chemical bonding with strong binding
force between Pt and CNTs are expected for the catalyst prepared by ALD.
As shown in Figures 11a and S14, the main product changed to DMF when the Pt/CNTs-
ALD catalyst was applied, and a higher HMF conversion (100%) and selectivity to DMF
(46.1%) were achieved in comparison with the Pt/CNTs-IM catalyst, which could be due to the
different interactions between Pt and CNTs. As presented in Figure 6a, the XPS spectra of Pt
for Pt/CNTs-ALD exhibited 4f7/2 binding energy of Pt0 at around 71.9 eV, and that of the
catalysts prepared by the impregnation method had a lower binding energy, indicating that
there was a weak binding force between Pt NPs and carbon supports. Meanwhile, the H2-TPR
results were consistent with the XPS spectra. As shown in Figure 6b, the H2-TPR of Pt/CNTs-
IM and Pt/CNTs-ALD exhibited weak peaks at around 550 ℃ and 600 ℃, respectively,
implying that the binding force between Pt and CNTs was stronger for the Pt/CNTs-ALD
catalyst. Compared with that of Pt/CNTs-ALD, the Pt0 binding energy was lower for Pt/CNTs-
IM, which could enhance electron transfer between metal and supports. Instead, for the
Pt/CNTs-ALD catalyst, metal NPs and supports were linked by chemical bonding (Pt-O-
support), and the electrons could transfer from Pt to O, which was verified by XPS analysis
25

Catalysis Science & Technology
Page 26 of 41
View Article Online
DOI: 10.1039/D0CY01920H
440
441
442
443
444
and, thereby, leading to the decrease of electronic density of Pt NPs. It has been reported that
reducing the metal electron density helps weaken binding with the reactant [58]. Therefore, the
low binding energy is beneficial to enhance selectivity of C=O hydrogenation [45], but it could
reduce the adsorption of H2, which suppressed further hydrogenolysis of DHMF reaction [59].
(a)
(b)
445
446
Figure 11. (a) Selective hydrogenation of HMF and (b) cycling tests of HMF hydrogenation
over Pt-based catalysts prepared by different methods. Reaction conditions: 0.5 g HMF, 0.05
g n-Tridecane, 40 mL 1-butanol, 50 mg catalyst, 160 °C, 10 bar H2, and 500 rpm.
447
448
449
450
451
To better understand the factors affecting the yield of DMF, a PtNi/CNT-ALD catalyst
was prepared. A higher loading of Pt NPs on CNTs was achieved by ALD, since one of the
26

Page 27 of 41
Catalysis Science & Technology
View Article Online
DOI: 10.1039/D0CY01920H
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
main byproducts was DHMF (intermediate product) over the Pt/CNT-ALD catalyst and a high
loading of Pt could improve the activity and DMF selectivity for the Pt-Ni/CNT-ALD
bimetallic catalyst. The yield of DMF was as high as 93.4% over the PtNi/CNTs-ALD catalyst
due to the synergistic effect between Pt and Ni, which was reported in a previous paper [60].
The Pt XPS spectrum of PtNi/CNTs-ALD (Figure 6a) exhibits that the peaks of Pt0 had a lower
binding energy, compared to that of the Pt/CNTs-ALD catalyst, which implied that there was
an electron transfer from Ni to Pt and the Pt NPs had a higher charge density in the PtNi/CNTs-
ALD catalyst. The high charge density enhanced the adsorption of intermediate products and
thereby promoted the selectivity to DMF [25], as presented in Figure 12. Besides, the apparent
peak of PtNi/CNTs-ALD was located at 435 ℃ in Figure 6b, which was attributed to the
reduction of NiOx species. It is reported that NiOx species could accept electrons as the Lewis
acid center [61], which could enhance the activity of C-O hydrogenation and improve the
selectivity to DMF. Although PtNi/CNTs-ALD had a higher loading of Pt, the TOF was low
in comparison with Pt/CNTs-ALD. It indicates that Ni suppressed the high activity of Pt
catalysts, and the moderate activity could enhance the selectivity to DMF and avoid the
occurrence of side reactions. A similar phenomenon was observed for Pt-Co/CNTs catalysts
prepared by ALD [25].
(a)
(c)
(b)
470
27

Catalysis Science & Technology
Page 28 of 41
View Article Online
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
Figure 12. Proposed mechanism of selective hydrogenation of HMF over (a) Pt/CDNOTI: s10-.I1M039,/D(b0C)Y01920H
Pt/CNTs-ALD, and (c) PtNi/CNTs-ALD catalysts.
The cycling tests of CNTs supported Pt and PtNi catalysts were performed to evaluate their
stability. As shown in Figure 11, after 3 cycles of tests, the HMF conversion and selectivity to
DMF almost did not change over the Pt/CNTs-ALD catalyst, which were ~100% and ~47%,
respectively, indicating its good stability. It could be owing to the strong chemical bonding
between Pt and CNTs, as we proved previously [62]. In contrast, a noticeable decrease of HMF
conversion was observed over the Pt/CNTs-IM catalyst, implying the leaching of Pt NPs during
the reaction and the weak interaction between Pt and CNTs, which was verified by the ICP
analysis. In addition, after the second cycling test, the selectivity to DMF significantly
decreased over the PtNi/CNTs-ALD catalyst, though the HMF conversion was still 100%,
which could be attributed to the leaching of Ni NPs. Based on the ICP-AES analysis, the Ni
loading decreased from 1.5 wt.% to 0.9 wt.% after the cycling tests. As shown in Figure S15,
compared to that in the first cycling test, a complete HMF conversion reached in a shorter time
in the second cycling test. This implied that Ni suppressed the activity of Pt/CNTs-ALD
catalysts, and the leaching of Ni NPs enhanced the activity of PtNi/CNTs-ALD catalyst.
Enhancement of the stability of the PtNi/CNTs-ALD catalyst is being pursued by this group.
4. Conclusions
A series of Pt catalysts supported on various carbon materials were prepared by the
impregnation method or ALD with a similar Pt loading and particle size. DHMF was the main
28

Page 29 of 41
Catalysis Science & Technology
View Article Online
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
product over all Pt catalysts prepared by the impregnation method, and the Pt/BCD-OIMI: 10c.1a0t3a9/lDy0sCtY01920H
showed the highest yield of DHMF (90.1%) owing to the abundant surface functional groups
of BC. Compared with the Pt/CNTs-IM catalyst, the reaction pathway was altered and the main
product was DMF with a yield of 46.1% over Pt/CNTs-ALD due to the strong interaction
between Pt and CNTs. Thus, the interaction between Pt and carbon supports determined the
reaction pathway and main product in the hydrogenation of HMF reaction. After depositing Ni
NPs on the Pt/CNTs-ALD catalyst by ALD, the yield of DMF increased to 93.4% with a
complete HMF conversion over the PtNi/CNTs-ALD bimetallic catalyst attributed to the
synergistic effects between Pt and Ni NPs.
Conflict of interest
The authors declare no competing financial interest.
Acknowledgements
This work was supported by the National Natural Science Foundation of China (No. 22002010)
and Natural Science Foundation of Liaoning Province of China (Joint Foundation for Shipping:
No. 2020-HYLH-23).
References
[1]
Sang Yup Lee, Hyun Uk Kim, Tong Un Chae, Jae Sung Cho, Je Woong Kim, Jae Ho
Shin, Dong In Kim, Yoo-Sung Ko, Woo Dae Jang, and Yu-Sin Jang, A comprehensive
metabolic map for production of bio-based chemicals. Nature Catalysis, 2019. 2(1): p.
29