Bimetallic Ru-Ni Catalyzed Aqueous-Phase Guaiacol Hydrogenolysis at Low H2 Pressures

Article abstract of DOI:10.1021/acscatal.7b02317

Aqueous-phase hydrogenolysis of renewable biomass at low H₂ pressures is an attractive route to selectively produce renewable fuels and valuable chemicals. Here, we show that Ru and Ni nanoparticles (NPs) dispersed on HZSM-5 with an optimum H^? radical transfer catalyzed a rapid rate (152 mmol g^-1 h^-1) of hydrogenolysis of C-O bonds in lignin-derived guaiacol at 240 °C and 2 bar H₂ pressure in water. The coimpregnated individual Ru and Ni nanoparticles (NPs) on HZSM-5 were highly dispersed and did not present an alloy structure, but the individual Ru and Ni NPs were in close proximity. The guaiacol hydrogenolysis rates were proportional to the amounts of the adjacent RuO₂ and NiO NPs on the calcined samples, suggesting that the closely contacted Ru and Ni NPs on HZSM-5 are the active sites. In the water phase at low H₂ pressures, Ru dissociated the hydrogen molecules to H^? radicals (H^?), and then such radicals were transferred to adjacent Ni atoms to activate the capability of inert Ni centers. The adjustment of the H^? transfer length between Ru and Ni NPs led to shorter H^? transfer lengths, which resulted in activities as high as 118 mmol g^-1 h^-1. The transferring and anchoring of H^? radicals was considered to be achieved by the Si-OH groups and their defects on HZSM-5, as demonstrated by a temperature-programmed desorption of hydrogen coupled with mass spectroscopy (TPD/H₂-MS) experiment. To further shorten the H^? transfer length over uniformly formed Ru and Ni nanoparticles, the isolated Ni islands were removed through the incorporation of a Ru precursor that initially occupied the Br?nsted acid sites on HZSM-5. By fully activating the two metals in the aqueous phase via an H^? transfer mechanism at low H₂ pressures, the rational design of bimetallic Ru-Ni catalysts provides a promising approach for achieving substantially high rates in selective hydrogenolysis steps.

Full text of DOI:10.1021/acscatal.7b02317

Subscriber access provided by READING UNIV
Article
Bimetallic Ru–Ni Catalyzed Aqueous-phase
Guaiacol Hydrogenolysis at Low H2 Pressures
Zhicheng Luo, Zhaoxia Zheng, Lei Li, Yitao Cui, and Chen Zhao
ACS Catal., Just Accepted Manuscript • DOI: 10.1021/acscatal.7b02317 • Publication Date (Web): 27 Oct 2017
Downloaded from http://pubs.acs.org on October 30, 2017
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
ACS Catalysis is published by the American Chemical Society. 1155 Sixteenth Street
N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 11
ACS Catalysis
1
2
3
4
5
6
7
8
9
Bimetallic Ru–Ni Catalyzed Aqueous-Phase Guaiacol Hydro-
genolysis at Low H2 Pressures
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Zhicheng Luo,a Zhaoxia Zheng,a Lei Li,b Yi-Tao Cuic and Chen Zhaoad*
a. Shanghai Key Laboratory of Green Chemistry and Chemical Processes, School of Chemistry and Molecular Engi-
neering, East China Normal University, Shanghai, 200062, China
E-mail: czhao@chem.ecnu.edu.cn (C. Zhao)
b. Synchrotron Radiation Nanotechnology Center, University of Hyogo 1-490-2 Kouto, Shingu-cho Tatsuno, Hyogo
679-5165, Japan
c. Synchrotron Radiation Laboratory, Laser and Synchrotron Research Center (LASOR), The Institute for Solid State
Physics, The University of Tokyo, 1-490-2 Kouto, Shingu-cho Tatsuno, Hyogo 679-5165, Japan
d. Key Laboratory of Low-Carbon Conversion Science & Engineering, Shanghai Advanced Research Institute, Chi-
nese Academy of Sciences, 201210, Shanghai, China
attracted the attention of biomass conversion researchers1.
ABSTRACT: Aqueous phase hydrogenolysis of renewable
Water is found in biomass with large amounts and is also
biomass at low H2 pressures is an attractive route to selec-
produced as a result of oxygen removal reactions. Thus,
selective hydrogenolysis of biomass is most likely carried
out in water.
tively produce renewable fuels and valuable chemicals. Here,
we showed that the dispersive Ru and Ni nanoparticles (NPs)
on HZSM-5 with an optimum H· radical transfer catalyzed a
rapid rate (152 mmol·g−1·h−1) in hydrogenolysis of C–O bonds
in lignin-derived guaiacol at 240 °C and 2 bar H2 pressure in
water. The co-impregnated individual Ru and Ni nanoparti-
cles (NPs) on HZSM-5 were highly dispersed and did not
present an alloy structure, but the individual Ru and Ni NPs
were in a close proximity. The guaiacol hydrogenolysis rates
were proportional to the amounts of the adjacent RuO2 and
NiO NPs on the calcined samples, suggesting the closely
contacted Ru and Ni NPs on HZSM-5 are the active sites. In
the water phase at low H2 pressures, Ru dissociated the hy-
drogen molecules to H· radicals (H·), and then such radicals
were transferred to adjacent Ni atoms for activating the ca-
pability of inert Ni centers. The adjustment of the H· transfer
length between Ru and Ni NPs led to shorter H· transfer
lengths, which resulted in activities as high as 118
mmol·g−1·h−1. The transferring and anchoring of H· radicals
was considered to be achieved by the Si-OH groups and their
defects on HZSM-5, as demonstrated by temperature pro-
grammed desorption of hydrogen coupled with mass spec-
troscopy (TPD/H2-MS) experiment. To further shorten the
H· transfer length over uniformly formed Ru and Ni nano-
particles, the isolated Ni islands were removed through the
incorporation of a Ru precursor that initially occupied the
Brönsted acid sites on HZSM-5. By fully activating the two
metals in the aqueous phase via an H· transfer mechanism at
low H2 pressures, the rational design of bimetallic Ru–Ni
catalysts provides a promising approach for achieving sub-
stantially high rates in selective hydrogenolysis steps.
With regard to the aqueous-phase lignin conversion,
Ru has been reported to be a proper hydrogenolysis metal
showing medium activity and high selectivity for convert-
ing of lignin and its model compounds2-4. Ru/C showed
higher hydrogenolysis selectivity than Pd/C and Pt/C for
hydrogenolysis of lignin dimers (β-O-4 linkage) to arenes
in the aqueous phase2. The Ru–WOx bifunctional catalyst
displayed high catalytic activities for the hydrogenolysis
of lignin derived phenols and phenyl ethers to arenes3. A
porous Ru/Nb2O5 catalyst enabled the selective produc-
tion of arenes via direct hydrodeoxygenation of organo-
solv lignin4. A number of mechanisms have been pro-
posed to explain the high activity of Ru in the hydrother-
mal biomass conversion processes. Akpa et al. showed
that the interaction of water molecules with hydroxyl-
butyl intermediates considerably decreased the energy
barriers for the hydrogenation of 2-butanone over a
Ru(0001) surface, thereby increasing the reaction rates in
water5. Additionally, Gallezot et al. demonstrated that
water molecules co-adsorbed on Ru(0001) could lower the
energy barriers, while the dissociation of water could re-
sult in larger hydrogen concentrations on the Ru(0001)
surface. Thus, both effects would increase the hydrogena-
tion rates of carbonyl groups6.
Low H2 pressures have been found to favor the selec-
tive aqueous-phase hydrogenolysis of biomass feedstock.
We have previously reported that the weaker adsorbed
H· species on the Ru surface surrounding the benzene
rings will be fast desorbed as the temperature increases,
while the H· species nearby the oxygen atoms are strongly
adsorbed due to the induction by the oxygen atom of the
β-O-4 model compound (contributing to 45–62% C–O
linkages in lignin), and thus are retained on the Ru sur-
Keywords: Bimetallic Ru–Ni, XAFS, H· Radical Transfer,
Aqueous-Phase Hydrogenolysis, Lignin.
INTRODUCTION
The hydrogenolysis of C–O bonds represents an im-
portant reaction for the conversion of biomass that has
ACS Paragon Plus Environment

ACS Catalysis
Page 2 of 11
face even with a rising temperature. This resulted in se-
Ru/HZSM-5, RuCo/HZSM-5 showed a slightly higher
hydrogenolysis rate (ca. 40 mmol·g−1·h−1), while
RuCu/HZSM-5 and RuFe/HZSM-5 exhibited lower rates
(18 and 24 mmol·g−1·h−1, respectively). The stability of the
active bimetallic RuNi/HZSM-5 catalyst was subsequently
examined by conducting four consecutive catalytic runs
(Fig. S1). The conversion and the phenol selectivity re-
mained at around 72 and 90%, respectively, for the four
consecutive runs, thereby indicating that the
RuNi/HZSM-5 catalyst was robust. As revealed by TEM
images (Fig. S2), the bimetallic Ru and Ni NPs were
slightly sintered after the four consecutive runs, while N2
adsorption measurements revealed a minimal loss of sur-
face areas in the recycled catalysts (Table S3). These data
demonstrated that RuNi/HZSM-5 was reusable. In a next
step, we focused on understanding the role of
RuNi/HZSM-5 in enhancing the hydrogenolysis rates
during the aqueous-phase reaction at low H2 pressures.
1
2
3
4
5
6
7
8
lective cleavage of C-O bonds and thus, achieves very
high aromatic hydrocarbons selectivity7. However, low H2
pressures will partially lower hydrogenolysis capability of
Ru in water.
Bimetallic Ru–Ni materials may modify the electron-
ic and structural properties of the surface, which may
highly accelerate the hydrogenolysis rates of C-O bond
cleavage. In some cases it has been reported that unsup-
ported core-shell RuNi nanoparticles (NPs) allowed the
formation of smaller particle sizes as compared to Ni8, or
the addition of Ru to Ni enhanced the interaction be-
tween both metals while weakening the interaction of Ni
with Al-SBA-159. Herein, in this work, we report that the
optimization of the H· radical transfer length and the full
activation of the two metals (Ru and Ni) can maximize
the selectivity and activity for the cleavage of C–O bond
of bio-molecule guaiacol in the aqueous phase and at low
hydrogen pressures. Extensive characterization was used
to explore the intrinsic Ru and Ni structure and interac-
tions between Ru and Ni, as well as the structure-activity
relationship. Based on these understandings, we eventual-
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
160
120
80
40
0
100
80
60
40
20
0
ly
synthesized
the
optimum
bimetallic
Ru-
Nimodified/HZSM-5 catalyst for the selective hydrogenolysis
reactions on guaiacol in water.
RESULTS AND DISCUSSION
Bimetallic RuNi/HZSM-5 Catalyzed Aqueous-phase
Guaiacol Hydrogenolysis at a Low H2 Pressure. Guaia-
col, a lignin-derived model compound, was selected as
the feedstock for the selective aqueous-phase hydrogen-
olysis reaction to phenol over Ru–M, Ru, and Ni based
HZSM-5 catalysts. Mild conditions (240 °C, 2 bar H2) were
selected to ensure high hydrogenolysis rates and to avoid
the hydrogenation of the aromatic benzene rings10. We
summarized the gained rates for gas-phase and liquid
phase hydrogenolysis of guaiacol to benzene in literature
and in our work (Tables S1 and S2), and by comparison
Ru/HZSM-5 was shown to be a particularly active hydro-
thermally-stable hydrogenolysis catalyst in the aqueous
phase. The Ru/HZSM-5 catalyst showed a hydrogenolysis
rate of 32 mmol·g−1·h−1 and a 90% phenol selectivity,
which represents the highest catalytic rate reported to
date for guaiacol hydrogenolysis.11
A series of 5 wt%Ru–5 wt%M (M = Pd, Pt, Co, Cu, Fe,
and Ni) bimetallic catalysts supported on HZSM-5 (pre-
pared by the co-impregnation method) were tested to-
wards the model reaction guaiacol hydrogenolysis in the
aqueous phase (Fig. 1). The carbon balance in the liquid
phase exceeded 92% in the tested cases. The addition of
noble metals (i.e., Pd and Pt) to Ru/HZSM-5 resulted in
slightly lower hydrogenolysis rates, although the phenol
selectivity remained nearly unchanged (85-93%). Transi-
tion base metals (i.e., Fe, Co, Cu, and Ni) were subse-
quently added to the Ru/HZSM-5 catalysts. RuNi/HZSM-
5 showed the highest activity (118 mmol·g−1·h−1, 4 times
higher than that of Ru/HZSM-5) among the catalysts test-
ed and a selectivity towards phenol similar to that of
Ru/HZSM-5 (89%) within 1 h, although Ni/HZSM-5 only
attained a low rate of 3.2 mmol·g−1·h−1. When compared to
Ni Ru-Cu Ru-Fe Ru-Pt Ru-Pd Ru Ru-Co Ru-Ni
Figure 1. Selective hydrogenolysis of guaiacol to phenol
over Ru–M, Ru-, and Ni- based HZSM-5 catalysts. The left
axis shows the hydrogenolysis rate on guaiacol, while the
right axis shows the selectivity towards phenol. General
conditions: guaiacol (1.0 g), 5 wt% Ru-5 wt% M/HZSM-5
or 5 wt% Ru/HZSM-5 or 5 wt% Ni/HZSM-5 (0.03 g), H2O
(100 mL), 240 °C, 2 bar H2, 1 h, and stirring at 650 rpm.
Determination of the distribution and structure of
bimetallic Ru–Ni nanoparticles. The RuNi/HZSM-5
catalyst (5%Ru-5%Ni/HZSM-5) was extensively character-
ized by X-ray diffraction (XRD), high-resolution transmis-
sion electron microscopy (HR-TEM), scanning transmis-
sion electron microscopy coupled with high-angle annu-
lar dark field (STEM-HAADF), TEM coupled with elec-
tron energy loss spectroscopy (TEM-EELS), X-ray absorp-
tion fine structure spectroscopy (XAFS), and density func-
tional theory (DFT) calculations. As shown in XRD pat-
terns (Fig. 2a), no diffraction peaks ascribed to Ru were
observed for 5%Ru/HZSM-5 because of the high disper-
sion of the metal, while diffraction peaks ascribed to face
centered cubic phases (fcc)-structured Ni (111) were ob-
served for 5%Ni/HZSM-5. Unlike the monometallic cata-
lysts, no Ru- or Ni-related diffraction peaks in XRD pat-
terns were identified for the RuNi/HZSM-5 catalysts. The
absence of diffraction peaks corresponding to Ru in the
bimetallic catalyst may indicate the presence of highly
dispersed Ru NPs. The absence of diffraction peaks as-
cribed to Ni (111) in the RuNi/HZSM-5 catalyst was result-
2
ACS Paragon Plus Environment

Page 3 of 11
ACS Catalysis
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 2. (a) XRD patterns of 5%Ru/HZSM-5, 5%Ni/HZSM-5, and 5%Ru-5%Ni/HZSM-5. A series of characterizations for
RuNi/HZSM-5 sample (Figs. 2b−h); (b) HAADF image and particle distributions, (c) HRTEM image, (d) HRTEM image, (e)
HAADF image, (f) HAADF-mapping, (g) HAADF-mapping with higher magnification, (h) EELS line scan. (i) FT-EXAFS
spectra at the Ni K-edge and Ru K-edge of Ni, Ru–Ni and Ru NPs on HZSM-5.
Table 1. Curve-fitting results of 5%Ni/HZSM-5, 5%Ru/HZSM-5, and 5%Ru-5%Ni/HZSM-5.
Sample
Edge
Ni K-edge
Ni K-edge
Ru K-edge
Ru K-edge
Shell
Ni-Ni
Ni-Ni
Ru-Ru
Ru-Ru
N
R/Å
Δσ2/Å2
0.00604
0.00523
0.00318
0.00325
Ni/HZSM-5
7.250
2.472
3.948
3.972
2.4858
2.4751
2.6574
2.6544
RuNi/HZSM-5
Ru/HZSM-5
*N: Coordination number, R: distance, Δσ2: Debye-Waller factor
Table 2．The formation energy of Ru-Ni alloy calculated from DFT.
Alloy (Ru : Ni, mass ratio)
5 : 0.5
0.032
5 : 1.5
0.057
5 : 2.5
0.054
5 : 3.5
0.044
5 : 4.5
0.042
5 : 5
5 : 12.5
0.028
Formation energy (ev/atom)
0.045
ed from the weak crystallization of Ni NPs, thereby imply
ing that the bimetallic NPs maybe well dispersed, with Ru
promoting the dispersion of Ni NPs.
and HRTEM results consistently confirmed the uniformly
distribution of individual Ru and Ni NPs on HZSM-5.
X-ray absorption fine structure (XAFS) measurement
was conducted on RuNi/HZSM-5 for the Ru and Ni K-
edges. Fig. 2i shows the first shell fitting results (with k3
weight, and the fitting results in k-space are presented in
Figs. S3a to S3d), and the gained parameters are provided
in Table 1. The coordination numbers for the Ni–Ni con-
tributions in the first shell of Ni/HZSM-5 and
RuNi/HZSM-5 were 7.250 and 2.472 (see Table 1). The
smaller Ni–Ni coordination number of RuNi/HZSM-5
reflected the smaller size of Ni particles compared to that
on Ni/HZSM-5. The nearly identical coordination num-
bers for the Ru–Ru contributions in the first shell of
Ru/HZSM-5 (3.972) and RuNi/HZSM-5 (3.948) suggested
comparable Ru particle sizes on these two samples. The
Ni–Ni interatomic distances were both 2.48 Å on
Ni/HZSM-5 and RuNi/HZSM-5, respectively, which is in
good agreement with that of fcc Ni metal (2.49 Å). The
same Ru–Ru bond distances (ca. 2.65 Å) were detected for
the Ru k-edge in the Ru based samples. It was noted that
the Ru-Ni bond (at around 2.54 Å) 12 was not observed.
According to these results, RuNi alloy NPs were not
formed, which was consistent with the HRTEM results.
Moreover, DFT calculations (Table 2) indicated that the
A
representative
HAADF-STEM
image
of
RuNi/HZSM-5 (Fig. 2b) revealed that the bimetallic NPs
were uniformly dispersed throughout the HZSM-5 sup-
port (average diameter: ca. 4.5 nm), in line with the XRD
results. HRTEM observations on RuNi/HZSM-5 revealed
the presence of separate Ru (101) and Ni (111) NPs over
HZSM-5, with interlayer spacings of 0.26 and 0.23 nm,
respectively (Figs. 2c-2d), thereby indicating that individ-
ual Ru and Ni NPs are present on the HZSM-5 surface. To
figure out the intrinsic structures of the bimetallic
RuNi/HZSM-5 catalyst in a larger region (300 nm and 100
nm), the STEM-EELS and STEM-mapping measurements
were performed (Figs. 2e-2h). The STEM-Mapping results
(Figs. 2e-2g) clearly showed that Ru and Ni NPs were uni-
formly and separately distributed in the tested region,
especially with the higher magnitude image (Fig. 2g). A
small domain with Ni-rich particles (green region, Fig. 2g)
may form the Ni islands, and this information is highly
agreed with the following catalyst characterization of
TPR-H2 profiles. The EELS line scan (Fig. 2h) demonstrat-
ed the uniform distribution of Ru and Ni elements in the
region of 30 nm. Thus, EELS mapping and line scan, XRD,
3
ACS Paragon Plus Environment

ACS Catalysis
Page 4 of 11
formation energies of RuNi alloys at varying atomic ratios
that the CO adsorption peak over the Ni0 species in the
1
2
3
4
5
6
7
8
(i.e., 5/0.5–5/12.5) were in the 0.028–0.057 eV/atom range.
These positive values confirmed that a RuNi alloy phase
was hardly formed. These results were in agreement with
previous analyses indicating that Ni and Ru are essentially
immiscible at equilibrium in virtue of a positive enthalpy
of formation13. Thus, the synthesis of RuNi alloys was hin-
dered at mild conditions. Based on these combined re-
sults, we concluded that the individual Ru and Ni NPs
were highly dispersed on HZSM-5, and RuNi alloy phases
were not formed.
RuNi/HZSM-5 sample shifted to lower wavenumbers
(2049 cm−1) as compared to Ni/HZSM-5 (2055 cm−1), while
the three peaks ascribed to Ru species in RuNi/HZSM-5
(2023, 2082, and 2145 cm−1) shifted to higher wave-
numbers as compared to Ru/HZSM-5 (2017, 2080 and 2134
cm−1). The red- and blue-shifts of the Ni-CO and Ru-CO
signals, respectively, suggested that electrons were trans-
ferred from Ru to Ni atoms, thereby further confirming a
close interaction between both metals, which were in line
with the TPR-H2 results.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
a.
Interaction of Ru and Ni NPs on HZSM-5. We subse-
quently explored the interaction between Ru and Ni NPs
on HZSM-5. Temperature programmed reduction of hy-
drogen (TPR-H2) was used to assess the interaction be-
tween the metals and the support as well as the interac-
tions between the two metals. The monometallic Ru cata-
lyst exhibited a strong reduction peak with a maximum at
143 °C (Fig. 3a) that was assigned to the reduction of
NiOꢀ1
NiOꢀ2
321 °C
450 °C
Ni/HZSMꢀ5
207 °C
NiOꢀ1
310 °C
RuO2ꢀNiO
NiOꢀ2
409 °C
RuꢀNi/HZSMꢀ5
Ru/HZSMꢀ5
143 °C
RuO2
0
100 200 300 400 500 600 700
14
RuO2 . The second minor peak observed at lower tem-
T (°C)
Ruδ+−(CO)
Ni0−CO
peratures (66°C) was likely ascribed to the reduction of
small amounts of the Ru precursor (i.e., RuCl3·3H2O)15. In
contrast, the monometallic Ni catalyst exhibited two re-
duction peaks at 321 and 450 °C that can be attributed to
the reduction of NiO species showing different interac-
tion strengths with HZSM-516. The bimetallic RuNi cata-
lyst exhibited three reduction peaks at 207, 310, and
409 °C, which corresponded to “RuO2-NiO”, “NiO-1”, and
“NiO-2” species, respectively. The“RuO2-NiO” species
referred to the individual RuO2 and NiO nanoparticles in
a close proximity, while “NiO-1” and “NiO-2” represented
NiO species interacting with HZSM-5. Remarkably, the
reduction temperature for “RuO2-NiO” species (207 °C)
lied between RuO2 (143 °C) and NiO (321 and 450 °C) spe-
cies in the monometallic catalysts. The lower reduction
temperature of the “RuO2-NiO” species than “NiO” spe-
cies can be explained by a mechanism involving an initial
H2 dissociation over Ru NPs to form H· radical species
that are subsequently transferred to adjacent Ni atoms to
facilitate the reduction of NiO17. The presence of new
“RuO2-NiO” species also suggested a strong interaction
between Ru and Ni NPs.
Infrared (IR) spectroscopy of adsorbed CO was used
to identify the states of bimetallic NPs. Ru/HZSM-5
showed three main bands at 2017, 2080, and 2134 cm−1 (Fig.
3b), assigned to linear-adsorbed CO on metallic Ru0, Ruδ+-
(CO), as well as Ruδ+-(CO)n species located inside the
HZSM-5 pores, respectively18. Ni/HZSM-5 showed a band
at 2055 cm−1 that was assigned to linearly adsorbed CO on
Ni0.18 The bimetallic RuNi/HZSM-5 catalyst showed four
bands at 2023, 2049, 2082, and 2145 cm−1 ascribed to line-
arly adsorbed CO on Ru0, Ni0, and Ruδ+–(CO), and Ruδ+–
(CO)n species, respectively19. No additional CO-
adsorption peaks were observed for RuNi/HZSM-5 as
compared to Ru/HZSM-5 and Ni/HZSM-5, probably rul-
ing out the possibility of Ru–Ni alloy phase. In addition,
even after incorporating of Ni atoms, the CO adsorption
intensity on RuNi/HZSM-5 was almost unchanged com-
pared to Ru/HZSM-5, suggesting well dispersed Ru and
Ni NPs on the surface of RuNi/HZSM-5. It is worth noting
b.
Ru0−CO
Ruδ+−(CO)n
2055
Ni/HZSMꢀ5
2082
2023
2049
2145
RuꢀNi/HZSMꢀ5
2080
2017
2134
Ru/HZSMꢀ5
2200 2150 2100 2050 2000 1950 1900
Wavenumber (cmꢀ1)
Ru 3d
Ru0
284.8
C1s
c.
Ru0
285.7
280.4
285.9
280.6
280 281 282 283 284 285 286 287
Binding Energy (eV)
Ni2+
Ni2+
861.3
Ni0
852.9
855.5
Ni 2p3/2
Ni0
d.
858.8
852.5
858.4
852
856
860
864
868
Binding Energy (eV)
Figure 3. The characterizations of Ru/HZSM-5,
Ni/HZSM-5, and RuNi/HZSM-5 catalysts by (a) TPR-H2
profiles and (b) FT-IR spectra of adsorbed CO at ambient
temperature. XPS spectra of (c) Ru 3d spectral region in
Ru/HZSM-5 and RuNi/HZSM-5, and (d) Ni 2p spectral
region in Ni/HZSM-5 and RuNi/HZSM-5. Black curve
(RuNi/HZSM-5), Red curve (Ru/HZSM-5), and Blue
curve (Ni/HZSM-5).
X-ray photoelectron spectroscopy (XPS) measure-
ments were carried out to assess the chemical states of Ru
and Ni in the Ru/HZSM-5, Ni/HZSM-5, and RuNi/HZSM-
4
ACS Paragon Plus Environment

Page 5 of 11
ACS Catalysis
temperature reduction process may change the resulted
5 samples (Figs. 3c–d). The main peak at 284.8 eV in the
1
2
3
4
5
6
7
8
Ru 3d spectra (Fig. 3c) was attributed to a strong C1s sig-
nal from carbon20. The Ru peaks in Ru/HZSM-5 (red curve)
at binding energies (BEs) of 280.4 and 285.7 eV were as-
signed to Ru(0) 3d5/2 and Ru(0) 3d3/2, respectively20. Ru
remained in the metallic Ru(0) state for RuNi/HZSM-5
(black curve) after Ni addition, although the BE of the
Ru(0) 3d5/2 and Ru(0) 3d3/2 bands increased by 0.2 eV. Fig.
3d shows the Ni 2p3/2 spectra of Ni/HZSM-5 and
RuNi/HZSM-5. In the case of Ni/HZSM-5 (blue curve),
the peak at 852.9 eV (satellite peak at 858.8 eV) was as-
cribed to Ni(0) species, while the peak at 855.5 eV was
assigned to Ni in an oxidized state (NiO) originated via
surface oxidation of Ni(0) NPs in air21. In the case of
RuNi/HZSM-5, the BEs of the Ni 2p3/2 spectra decreased
to 852.5 and 858.4 eV, while the band associated to NiO
species remained unchanged as compared to Ni/HZSM-5.
The increased binding energy of Ru(0) species and de-
creased binding energy of Ni(0) species, respectively, fur-
ther supported that electrons were transferred from Ru to
Ni atoms, in line with the results from IR spectra of CO
adsorption (Fig. 3b). This electron transfer between Ru
and Ni resulted in electron-deficient Ru and electron-
enriched Ni atoms, thereby increasing the strength of the
interaction between Ni and Ru NPs.
particle sizes of Ru and Ni NPs on HZSM-5, and then alter
the corresponding catalytic activities.
a.
240
350
180
435
435
800
5 Ruꢀ5Ni
120
80
40
0
240
b.
350
9
800
180
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
156 234 312 390 468 546 624 702 780
Reduction Temperature (°C)
covered Ni
c.
RuꢀNi
isolated Ni
5 Ruꢀ7.5 Ni
5 Ruꢀ5 Ni
5 Ruꢀ3 Ni
5 Ruꢀ2 Ni
5 Ruꢀ1 Ni
5 Ruꢀ0.5 Ni
5 Ru
Ru
200
0
400
600
800
Reduction Temperature (°C)
4.0
d. 160
140
120
100
80
Hydrogenolysis of Guaiacol
H2 Consumption (TPR)
3.5
3.0
2.5
2.0
1.5
1.0
5Ruꢀ7.5Ni
5Ruꢀ5Ni
Determination of the active site on RuNi/HZSM-5.
According to the TPR-H2 profile (Fig. 3a), the calcined
RuNi/HZSM-5 catalyst contained “RuO2-NiO”, “NiO-1”,
and “NiO-2” species corresponding to interacting RuO2–
NiO NPs, weakly interacting NiO–HZSM-5, and strongly
interacting NiO–HZSM-5 entities, respectively. To deter-
mine the actual active site of RuNi/HZSM-5, the hydro-
gen reduction temperatures and the Ni/Ru ratios were
varied to investigate the relationship between the nature
of the active sites and their catalytic performances.
5Ruꢀ3Ni
5Ruꢀ2Ni
5Ruꢀ1Ni
60
40
5Ruꢀ0.5Ni
5Ru
20
0
1
2
3
4
5
6
7
8
x%
5% Ruꢀx% Ni/HZSMꢀ5
Figure 4. (a) TPR-H2 profile of calcined 5%Ru-
5%Ni/HZSM-5. (b) Hydrogenolysis rates of guaiacol to
phenol over 5%Ru-5%Ni/HZSM-5 reduced at different
temperatures. (c) TPR-H2 profile of calcined 5%Ru-
x%Ni/HZSM-5 with varying Ni contents. (d) Guaiacol
hydrogenolysis rates to phenol and normalized TPR-H2
areas for 5%Ru-x%Ni/HZSM-5 samples with varying Ni
contents. General conditions: guaiacol (1.0 g),
RuNi/HZSM-5 catalyst (0.03 g), H2O (100 mL), 240 °C, 2
bar H2, 1 h, and stirring at 650 rpm.
To separately reduce the “RuO2-NiO”, “NiO-1”, and
“NiO-2” species, the calcined RuNi/HZSM-5 catalysts
were reduced at varying temperatures (i.e., 180, 240, 350,
435, and 800 °C, Fig. 4a). Then the as-received
RuNi/HZSM-5 samples were tested in the aqueous-phase
guaiacol hydrogenolysis reaction at 240 °C and 2 bar H2
(Fig. 4b). RuNi/HZSM-5 showed different guaiacol hydro-
genolysis rates (25, 102, 82, 53, and 36 mmol·g−1·h−1) de-
pending on the reduction temperatures (180, 240, 350, 435,
and 800 °C, respectively). The maximum hydrogenolysis
rate (102 mmol·g−1·h−1) was achieved after reduction of
RuNi/HZSM-5 at 240 °C, with exclusive interacted Ru and
Ni NPs after reduction of “RuO2-NiO” species. Above
240 °C, the hydrogenolysis rates decreased gradually with
temperature increase, probably due to the presence of
Ni/HZSM-5 species at higher reduction temperatures.
These results may indicate that closely contacted individ-
ual RuO2 and NiO particles (labeled as “RuO2-NiO” spe-
cies) on HZSM-5 rather than “RuO2” or “NiO” were the
main precursors for generating active sites of interacted
Ru and Ni NPs for the hydrogenolysis of guaiacol. In addi-
tion, the reduced Ni islands on HZSM-5 at higher reduc-
tion temperatures may impede the hydrogenolysis capa-
bility of interacted Ru and Ni NPs from “RuO2-NiO” spe-
cies. But this evidence is not sufficient, since the high-
Subsequently, the relation of guaiacol hydrogenolysis
and Ru/Ni ratios was explored. The Ni/Ru ratios were
varied by adjusting the Ni content (0, 0.5, 1, 2, 3, 5 and 7.5
wt%) at a constant Ru content of 5 wt%. As shown in the
TPR-H2 profiles (Fig. 4c), the temperatures and peak are-
as for reducing “RuO2-NiO” species gradually increased
with the increasing Ni contents. The higher reduction
temperatures from 140 to 220 °C with Ni contents of 0-3
wt% (at 5 wt% Ru loading) suggested stronger Ru and Ni
interaction, which may be resulted from closer distance of
RuO2 and NiO species. This is in line with the results
from XPS spectra and IR spectra of adsorbed CO. On the
5%Ru-5%Ni/HZSM-5 sample, isolated Ni island species
(reduction temperatures from 220 to 435 °C) appeared
and the reduction temperatures for “RuO2-NiO” species
kept stable, indicating that the atom numbers for inter-
acted Ru and Ni NPs reached an equilibrium value. Con-
tinuously increasing the Ni content to 7.5 wt%, the reduc-
5
ACS Paragon Plus Environment

ACS Catalysis
Page 6 of 11
tion peaks at 243 °C appeared between reduction peaks samples, probably due to the presence of isolated NiO
1
2
3
4
for interacted “RuO2-NiO” and isolated NiO species, sug-
gesting that additional NiO NPs may partially cover
RuO2-NiO species.
islands and covered RuO2-NiO species by NiO on their
calcined samples, as shown in TPR-H2 profiles at Fig. 4c.
Thus, based on the above understandings, we speculate
the formation process of x%Ru-5%Ni/HZSM-5 (x=0-7.5)
samples at Fig. 5.
5
6
7
8
9
The hydrogenolysis activities can be influenced by
the accessible Ru atoms on the tested reduced
RuNi/HZSM-5 samples, apart from the amounts of closely
interacted Ru and Ni species. To figure it out, we explored
the accessibility of Ru atoms by IR spectroscopy of ad-
sorbed CO with 5%Ru- x%Ni/HZSM-5 (x=0-7.5) samples
(Fig. 6a), and the deconvolution of the corresponding
species was provided in Fig. S4. The peak at 2134 cm-1 as-
signing to Ruδ+-(CO) species gradually decreased from
5%Ru/HZSM-5 to 5%Ru-7.5%Ni/HZSM-5 as Ni contents
increased. In line with it, the plotted integrated areas of
the Ruδ+-(CO) species band was displayed as a function
of Ru/Ni ratios in Fig. 6b, showing that the accessible Ru
atoms were decreasing with increased Ni contents. This
gained information concerning on the accessible Ru at-
oms (Fig. 6b) was not consistent with the activity changes
in Fig. 4d (black line), further confirming that the closely
contacted Ru and Ni NPs on HZSM-5 are the active sites.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 5. The speculated formation process of x%Ru-
5%Ni/HZSM-5 (x=0-7.5) catalysts with varying Ni con-
tents at a Ru loading of 5 wt%.
As shown in TPR-H2 profile at Fig. 4d, the consumed
H2 were proportional to the concentrations of the RuO2
and NiO precursors on HZSM-5. Below the Ni contents at
3 wt% on RuNi/HZSM-5 samples, the consumed H2 were
quantitatively resulted from the concentrations of “RuO2-
NiO” species. Within the higher Ni contents of 5-7 wt%
on RuNi/HZSM-5 samples, the consumed H2 included the
hydrogen-reduction of RuO2-NiO and surplus NiO spe-
cies.
Understanding the H transfer on Ru and Ni NPs on
·
HZSM-5 for aqueous phase guaiacol hydrogenolysis.
The RuO2, RuO2–NiO, and NiO surface species are pre-
sent on the calcined RuNi/HZSM-5 catalyst. We prepared
three RuNi/HZSM-5 samples (i.e., T-1, T-2, and T-3) hav-
ing different Ru–Ni distances in an attempt to investigate
the influence of the Ru–Ni distance on the reaction rates
(Fig. 7a). RuNi/HZSM-5 (T-1) was synthesized by sepa-
rately pressing the impregnated Ru/HZSM-5 and
Ni/HZSM-5 powders and subsequent physical mixture of
the pressed samples. RuNi/HZSM-5 (T-2) was prepared by
initially mixing the Ru/HZSM-5 and Ni/HZSM-5 followed
by pressing. RuNi/HZSM-5 (T-3) was prepared by press-
ing the co-impregnated RuNi/HZSM-5 catalyst.
Ruδ+−(CO)n
Ruδ+−(CO) Ni0−CO
a.
Ru0−CO
5Ruꢀ7.5Ni
5Ruꢀ5Ni
5Ruꢀ3Ni
5Ruꢀ2Ni
5Ruꢀ1Ni
5Ruꢀ0.5Ni
5Ru
2200 2150 2100 2050 2000 1950 1900
Wavenumber (cmꢀ1)
b.
Ruδ+ꢀ(CO)
5Ruꢀ5Ni
1.6
1.5
1.4
1.3
1.2
1.1
1.0
5Ru
5Ruꢀ0.5Ni
Subsequently, TPR-H2 technique was used to assess
the degree of interaction between Ru and Ni centers (Fig.
7b). The TPR profile of RuNi/HZSM-5 (T-1) revealed the
presence of separate RuO2 (reduction peaks at 140 °C) and
NiO (reduction peaks at 330 and 460 °C) particles on the
support, demonstrating no apparent interaction between
them. The reduction temperature at 197 °C on
RuNi/HZSM-5 (T-2) was relatively lower than 207 °C for
the interacting RuO2–NiO species on RuNi/HZSM-5 (T-3),
but was much higher than the reduction temperature of
RuO2 on T-1 (140 °C), indicating a much stronger interac-
tion between RuO2 and NiO NPs on T-2. The order for
the strength of interaction between RuO2 and NiO NPs
on T-1, T-2 and T-3 was: T-3 > T-2 >T-1. Therefore, it can
be concluded that the hydrogen transfer distance be-
tween Ru and Ni centers followed the trend: T-3 < T-2 <
T-1, indicating that the reduction temperature of NiO was
strongly influenced by the length of the hydrogen transfer
process.
5Ruꢀ1Ni
5Ruꢀ2Ni
5Ruꢀ3Ni
5Ruꢀ7.5Ni
0
1
2
3
4
5
6
7
8
5% Ruꢀx% Ni/ HZSMꢀ5 x%
Figure 6. (a) FT-IR spectra of adsorbed CO on 5%Ru-
x%Ni/HZSM-5 with varying Ni contents at ambient tem-
perature. (b) Normalized areas of CO adsorption for Ruδ+-
(CO) species in IR spectra on 5%Ru-x%Ni/HZSM-5 sam-
ples with varying Ni contents.
The x%Ru-5%Ni/HZSM-5 (x=0-7.5) samples were
tested for guaiacol hydrogenolysis at the identical condi-
tions (Fig. 4d). Within Ni contents of 0-3 wt%, the guaia-
col hydrogenolysis rates were proportional to the concen-
trations of the “RuO2-NiO” species, further suggesting
that the reduced interacted Ru and Ni NPs may be the
active sites. However, such hydrogenolysis rates were
lowered with 5%-7.5% Ni contents on RuNi/HZSM-5
Apart from the characterization of TPR-H2 profiles,
the Ru and Ni interaction was also detected by the EELS
lines of three RuNi/HZSM-5 samples of T-1, T-2, and T-3
6
ACS Paragon Plus Environment

Page 7 of 11
ACS Catalysis
Based on the above results, we formulated a reaction
(Fig. 7c). The average distances between Ru and Ni parti-
1
2
3
4
5
6
7
8
cles determined by EELS are defined to be the adjacent
distances between Ru-rich and Ni-rich centers, as indi-
cated from Fig. 7c. On T-1, the average distance of Ru and
Ni particles as detected by the SEM-EELS was around 150-
180 μm. On the T-2 and T-3 samples, the average distanc-
es of Ru and Ni atoms were ca. 50-100 nm and 3.3 nm,
respectively, as analyzed by the STEM-EELS measure-
ments. The trend in Ru- and Ni- particle distance among
T-1, T-2, and T-3 (T-3 < T-2 < T-1) shown from EELS re-
sults are highly consistent with that from TPR-H2 profiles
in Fig. 7b.
mechanism for the hydrogenolysis of C–O bonds over
RuNi centers in the aqueous phase at low hydrogen pres-
sures. The Ni NPs showed a very low activity (3.2
mmol·g−1·h−1) because of their relative poor hydrogen dis-
sociation ability and the low concentration of H2 in water.
However, the addition of Ru to Ni dramatically enhanced
the hydrogenolysis rates to 118 mmol·g−1·h−1. As shown in
Fig. 8, in the co-impregnated bimetallic RuNi/HZSM-5
catalyst, it is speculated that Ru can fast dissociate low-
concentrated hydrogen in water to form H· radicals that
are rapidly transferred to the adjacent Ni atoms, thereby
activating these inert Ni centers at low hydrogen pressure.
Thus, the hydrogenolysis capability was greatly influ-
enced by the distance between Ru and Ni (i.e., the length
of the hydrogen radical transfer process).
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 8. The supposed schematics of the C–O hydro-
genolysis mechanisms of guaiacol over Ni/HZSM-5 and
RuNi/HZSM-5.
Next, we try to explore how the HZSM-5 support acts
as a holder to store and transfer the H· radicals between
Ru and Ni NPs by the TPD-H2 coupled with MS meas-
urement. First we investigated the stability of HZSM-5 as
a function of temperature, we observed that at round
400 °C the Brönsted acid site (BAS) of HZSM-5 was dehy-
drated to generate water, as evidenced by the appearance
of H2O signal on the MS detector (Fig. S5). In addition, in
our previous work we also observed structural changes of
HZSM-5 during regeneration by air-calcination and hy-
drogen-reduction process. 22 The acid site concentrations
changed in the following trend: (i) the LAS increased and
the BAS decreased, (ii) the sum of BAS and LAS (ca. 90
μmol/g) was nearly unchanged. Therefore, there are
structural changes to HZSM-5 during calcination in flow
air, i.e., above 450 °C, the dehydration of BAS formed
Lewis acid sites. In fact, this gained former result highly
agrees with our conclusion in this work on the loss of BAS
of HZSM-5 at around 400 °C as shown from the TPD-MS
(Fig. S5). Meanwhile, the produced H2O would release H2
due to the recombination of H protons (derived from H2O)
during the mass-spectrometric detection, which would
seriously affect the detection of H2 signal, and thus, we
set the desorption temperature lower than 300 °C. Fig. 9a
displayed TPD/H2-MS profiles of HZSM-5, RuNi/HZSM-5,
and physically mixed RuNi/HZSM-5 and HZSM-5 samples.
It showed that no hydrogen was adsorbed on the HZSM-5.
But it was surprisingly noted that a significantly increased
concentration on desorbed H2 signal was observed on the
physically mixed sample compared to RuNi/HZSM-5. It is
suggested that the additional HZSM-5 may act as an H·
Figure 7. (a) Three approaches for constructing
RuNi/HZSM-5
RuNi/HZSM-5 (T-3) samples. (b) TPR profiles of
RuNi/HZSM-5 (T-1), RuNi/HZSM-5 (T-2), and
(T-1),
RuNi/HZSM-5
(T-2),
and
RuNi/HZSM-5 (T-3) samples. (c) The average distances
between Ni and Ru particles detected by the EELS meas-
urements on T-1, T-2 and T-3, (d) A schematic representa-
tion of the reduction of NiO and catalytic properties of
catalysts with different Ru and Ni distance.
The guaiacol hydrogenolysis activities were evaluated
on the T-1, T-2, and T-3 samples at optimal conditions
(240 °C and 2 bar H2) in aqueous phase. The hydrogenoly-
sis rates increased as follows: 46 mmol·g−1·h−1 (T-1), 60
mmol·g−1·h−1 (T-2), and 118 mmol·g−1·h−1 (T-3). The closer
distance between Ru and Ni is, the higher hydrogenolysis
activity attained, as shown in Fig. 7d. These results
strongly indicated that the hydrogenolysis capacity of
RuNi was also particularly determined by the length of
the hydrogen transfer process.
7
ACS Paragon Plus Environment

ACS Catalysis
Page 8 of 11
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 9. (a) The H2-TPD-MS profiles of HZSM-5, RuNi/HZSM-5, and physical mixture of RuNi/HZSM-5 and HZSM-5. (b)
The speculated mechanism of transfer and reservoir of H species on RuNi/HZSM-5.
radical acceptor and transferor, after H2 dissociation on
Ru and Ni NPs. Therefore, hydrogen spillover occurs on
the physical mixture of RuNi/HZSM-5 and HZSM-5. It
was also observed that the maximum H2 desorption tem-
perature on physically mixed RuNi/HZSM-5 and HZSM-5
(130 °C) was much lower than that on RuNi/HZSM-5
(150 °C). This demonstrates a much weaker interaction of
adsorbed H atoms on HZSM-5 compared to adsorbed H
atoms on interacted Ru and Ni NPs. With the increasing
amount of the HZSM-5 mixed into RuNi/HZSM-5 powder,
the desorption curve on RuNi/HZSM-5 and 4 HZSM-5 can
be clearly fitted into two peaks (Fig. S6), with one appear-
ing at 76 °C and another appearing at 118 °C. They were
assigned to the desorbed H atoms on HZSM-5 and de-
sorbed H atoms on Ru and Ni NPs, respectively. This re-
sult is fitted with the statement of Prins23 that the H-
spillover species are strongly bonded to a metal particle
and only moderately to a non-reducible support. Based
on this recognition, the desorption temperature for H· on
HZSM-5 would be much lower than that on interacted Ru
and Ni NPs, and this is proved by our experimental re-
sults (Fig. 9a).
It has been reported that defects on non-reducible
oxides by high temperature treatment may anchor H at-
oms.24 In line with it, Choi et al.24 reported that surface
hydroxyls, presumably Brönsted acids, are proved as a
main mediator for migration of H atoms on the surface of
HZSM-5. As shown in Fig. 9b, we depicted that the sur-
face defects of Si-OH groups and the intact Si-OH groups
including the Brönsted acids on HZSM-5 acted as an ex-
cellent holder to anchor and transfer the H· radicals. In
addition, it is reported that H· radical spillover on non-
reducible aluminum oxides is mediated by three coordi-
nated centers and restricted to relative short distances
from one metal to another metal.25 This also suggests that
the distance between the Ru and Ni centers may greatly
influence the transfer of H· radicals, thus affecting the
hydrogen spillover process. This is also evidenced by our
designed experiments, when exploring the relationship
between varying Ru and Ni distances and the correspond-
ing hydrogenolysis activities, as displayed at Fig. 7. There-
fore, the close interaction of Ru–Ni NPs can greatly in-
crease the reaction rates by fully activating the capacity of
metals via optimum H· transfer for the hydrogenolysis
reaction at low hydrogen concentrations.
Blocking the inactive isolated Ni/HZSM-5 species on
RuNi/HZSM-5 to form uniform interacted Ru and Ni
NPs with shorted H· transfer distance. Abundant iso-
lated Ni/HZSM-5 species were observed in the co-
impregnated RuNi/HZSM-5 sample as evidenced by the
reduction temperature peaks at 321 and 425 °C in the
TPR-H2 profile (Fig. 3a). The isolated Ni/HZSM-5 catalyst
showed poor activity towards the hydrogenolysis of guai-
acol (3.2 mmol·g−1·h−1) at 2 bar H2 in water. On one hand,
the presence of Ni islands hindered the formation of ac-
tive interacted Ru and Ni NPs, or partially covered active
interacted Ru and Ni NPs (Fig. 10a), which would lower
their hydrogenolysis activity as confirmed from the re-
sults in Fig. 4d. On the other hand, the presence of Ni
island would lengthen the H· transfer distance on Ru and
Ni NPs. In order to block the presence of these inactive,
isolated Ni species (RuNi/HZSM-5, Fig. 10a) while simul-
taneously maximizing the number of RuNi sites with
shorter H· transfer distance, we used an improved synthe-
sis method. In this method, the Ru precursor was firstly
introduced over the HZSM-5 support after which the Ni
precursor was incorporated. The catalyst was finally air-
calcined and hydrogen-reduced in order to generate the
active catalyst (Ru-Nimodified/HZSM-5, 5 wt% Ru-5 wt% Ni,
Fig. 10b). The Ru-Nimodified/HZSM-5 catalyst showed guai-
acol hydrogenolysis rates of 152 mmol·g−1·h−1 (versus 118
mmol·g−1·h−1 for RuNi/HZSM-5).
We speculated that the initial incorporation of the
Ru salt would favor ion-exchange between Ru ions and H+
on HZSM-5, and the subsequent introduction of Ni ions
would result in well-dispersed interacted Ru and Ni NPs
on Ru-Nimodified/HZSM-5. The absence of Ni islands and
the presence of abundant Ru–Ni active sites on the im-
proved sample with shorted H· transfer distance would be
responsible for the much higher Caromatic–O hydrogenoly-
sis rates.
8
ACS Paragon Plus Environment

Page 9 of 11
ACS Catalysis
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 10. Formation mechanism and accelerated rates of guaiacol hydrogenolysis at 240 °C and 2 bar H2 with (a)
RuNi/HZSM-5, (b) modified Ru-Ni/HZSM-5. Characterization of RuNi/HZSM-5 and modified Ru-Ni/HZSM-5 samples by
(c) IR of pyridine adsorption, (d) TPR-H2, and (e) IR of adsorbed CO at ambient temperature.
To confirm our hypothesis, we conducted a series of
characterizations on the catalyst samples. The sizes of
interacted Ru and Ni NPs on Ru-Nimodified/HZSM-5 (d =
4.5 ± 0.8 nm) were comparable to those on RuNi/HZSM-5
(d = 4.5 ± 1.2 nm), as shown on TEM images (Fig. S7),
eliminating their influence on the reaction. The acid sites
of the samples were studied by analyzing the IR spectra of
pyridine adsorption (Fig. 10c). The BAS on Ni/HZSM-5
(0.005 mmol·g−1) were almost completely occupied by Ni
ions as compared to HZSM-5 (BAS: 0.175 mmol·g−1), while
Ru/HZSM-5 showed a significantly higher concentration
of BAS (0.020 mmol·g−1). These results indicated that the
Ni ions have significantly stronger proton-exchange ca-
pacities compared to Ru ions. In line with this hypothesis,
we confirmed the presence of RuO2–NiO and isolated
NiO islands species on the co-impregnated RuNi/HZSM-5
catalyst by TPR-H2 profile (i.e., reduction peaks from
207 °C, 310 °C, and 409 °C, respectively). In contrast,
Ru@Ni/HZSM-5 did not exhibit the reduction peaks cor-
responding to isolated Ni/HZSM-5 (Fig. 10d). The pres-
ence of isolated Ni species was also demonstrated by the
IR spectra of CO adsorption on RuNi/HZSM-5 (i.e., dis-
tinguishing peak of Ni0 species at 2049 cm−1, Fig. 10e),
while this peak was basically absent for Ru-
Nimodified/HZSM-5. All these results suggest that more uni-
form RuNi NPs with shorted H· transfer distances are
formed by this improved way, and thus, achieve much
higher C-O bond hydrogenolysis rates.
CONCLUSIONS
By means of a new route involving H· radical transfer
mechanism in aqueous phase, we remarkably enhanced
the C–O cleavage rates on lignin-derived guaiacol using a
bimetallic RuNi catalyst. In-depth XRD, HR-TEM, STEM-
HAADF, STEM-EELS, XAFS, and DFT calculations charac-
terization confirmed that the co-impregnated individual
Ru and Ni NPs were highly dispersed and strongly inter-
acted with each other, although alloyed phases were not
formed. The closely contacted Ru and Ni NPs on HZSM-5
were subsequently proven to be the active sites. The dis-
sociated H· radicals on Ru atoms were readily transferred
to the adjacent inert Ni sites as a result of the interacting
electronic structures of Ru and Ni particles, as evidenced
by TPR-H2, IR spectra of CO adsorption, and XPS analyses.
We observed a relationship between the H· radical trans-
fer length on Ru and Ni NPs and the C–O bond cleavage
activity. That is the increased hydrogenolysis rates upon
shortening the H· transfer length. The structure design of
9
ACS Paragon Plus Environment

ACS Catalysis
the catalyst allowed it to be more sensitive to lower H2
Page 10 of 11
2. Luo, Z. C.; Zhao, C. Catal. Sci. Technol., 2016, 6, 3476–
3484.
1
2
3
4
5
6
7
8
concentrations in water. The H· radicals transferring and
anchoring was realized by the Si-OH groups and defects
of HZSM-5, as demonstrated by TPD/H2-MS experiments.
To further shorten the H· transfer length over uniformly
formed Ru and Ni nanoparticles, Ni islands were blocked
by initially occupying the BAS of HZSM-5 by ion-
exchanged Ru ions, thereby maximizing the guaiacol hy-
drogenolysis rates. The new catalytic system involving
H· radical transfer in aqueous phase reactions provides a
promising approach for selective hydrogenolysis of bio-
mass molecules over multi-functional catalysts.
3. Huang, Y.; Yan, L.; Chen, M.; Guo, Q. Fu, Y. Green
Chem., 2015, 17, 3010–3017.
4. Shao, Y.; Xia, Q. N.; Dong, L.; Liu, X. H.; Han, H.; Par-
ker, S. F.; Cheng, Y. Q.; Daemen, L. L.; Ramirez-Cuesta, A.
J.; Yang, S. H.; Wang, Y. Q. Nat. Commun. 2017, 8, 16104-
16112.
5. Akpa, B. S.; Agostino, C. D.; Gladden, L. F.; Hindle, K.;
Manyar, H.; McGregor, J.; Li, R.; Neurock, M.; Sinha, N.;
Stitt, E. H.; Weber, D.; Zeitler, J. A.; Rooney, D. W. J.
Catal., 2012, 289, 30-41.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
6. Michel, C. Gallezot, P. ACS Catal., 2015, 5, 4130-4132.
7. Luo, Z. C.; Wang, Y. M.; He, M. Y.; Zhao, C. Green
Chem., 2016, 18, 433-441.
8. Zhang, J. G.; Teo, J.; Chen, X.; Asakura, H.; Tanaka,T.;
Teramura, K.; Yan, N. ACS Catal., 2014, 4, 1574-1583.
9. Pichaikaran, S.; Arumugam, P. Green Chem., 2016, 18,
2888.
10. Sun, J. M.; Karim, A. M.; Zhang, H.; Kovarik, L.; Li, X.
H. S.; Hensley, A. J.; McEwen, J. S.; Wang, Y. J. Catal., 2013,
306, 47-57.
11. Luo, Z. C.; Zheng, Z. X.; Wang, Y. C.; Sun, G.; Jiang, H.;
Zhao, C. Green Chem., 2016, 18, 5845-5858.
12. Mori, K.; Miyawaki, K.; Yamashita, H. ACS Catal., 2016,
6, 3128−3135.
13. He, X.; Kong, L. T.; Li, J. H.; Li, X. Y.; Liu, B. X. Acta
Mater. 2006, 54, 3375-3381.
14. Li, D.; Ichikuni, N.; Shimazu, S.; Uematsu, T. Appl.
Catal., A, 1999, 180, 227–235.
15. Boonyasuwat, S.; Daniel, T. O.; Steven. E. R.; Crossley,
P.; Catal. Lett., 2013, 143, 783–791.
16. (a) Yang, Y.; Ma, J. X.; Wu, F. Int. J. Hydrogen Energy,
2006, 31, 877-882; (b) Mohan, V.; Venkateshwarlu, V.;
Pramod, C. V.; Raju, B. D.; Rao, K. S. R. Catal. Sci. Tech-
nol., 2014, 4, 1253–1259.
17. Garacia, P. B.; Sancho, C. G.; Molina, A. I.; Catellon, E.
R. A.; Lopez, J. Appl. Catal., A, 2010, 381, 132–144.
18. Shen, J. Y.; Sayari, A.; Kaliaguine, S. Appl. Spectrosc.
1992, 46, 1279-1287.
19. Crisafulli, C.; Scire, S.; Magglore, R.; Mnico, S.; Gal-
vagno, S. Catal. Lett., 1999, 59, 21–26.
20. Okal, J.; Zawadzki, M.; Tylus, W. Appl. Catal. B: Envi-
ron., 2011, 101, 548-559.
21. Salagre, C.; Fierro, J. L. G.; Medina, F.; Sueiras, J. E. J.
Mol. Catal. A: Chem., 1996, 106, 125-134.
22. (a) Zhao, C.; Kasakov, S.; He, J. Y.; Lercher, J. A. J.
Catal., 2012, 296, 12-23; (b) Nash, M. J.; Shough, A. m.;
Fickel, D. W.; Doren, D. J. Lobo, R. F. J. Am. Chem. Soc.,
2008, 130, 2460–2462.
23. Prins, R. Chem. Rev. 2012, 112, 2714−2738.
24. Lm, J.; Shin, H.; Jang, H.; Kim, H.; Chio, M. Nat.
Commun. 2014, 5, 3370-3378.
25. Karim, W.; Spreafico, C.; Kleibert, A.; Gobrecht, J.;
Vandevondele, J.; Ekinci, Y.; Bokhoven, J. A. van. Nature,
2017, 541, 68–71.
ASSOCIATED CONTENT
Supporting Information
Catalyst recycling data, TPD-MS profile of HZSM-5, EXAFS
raw data, the deconvolution of FT-IR spectra of adsorbed CO,
TPD-MS profile of HZSM-5, TEM images of RuNi/HZSM-5
samples. This material is available free of charge via the In-
ternet at http://pubs.acs.org.
AUTHOR INFORMATION
Corresponding Author
*Email: czhao@chem.ecnu.edu.cn
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENT
This research was supported by the National Key Research
and Development Program of China (Grant No.
2016YFB0701100), the Recruitment Program of Global Young
Experts in China, Natural Science Foundation of China
(Grant No. 21573075), the Shanghai Pujiang Program (Grant
No. PJ1403500), the Foundation of Key Laboratory of Low-
Carbon Conversion Science & Engineering, Shanghai Ad-
vanced Research Institute, Chinese Academy of Sciences. The
synchrotron radiation based XAFS experiment was carried
out at the BL08B2 of SPring-8 with the approval of Japan
Synchrotron Radiation Research Institute (Proposal
No.2016B3302).
REFERENCES:
1. (a) Zakzeski, J.; Bruijnincx, P. C. A.; Jongerius, A. L.;
Weckhuysen, B. M. Chem. Rev., 2010, 110, 3552-3599; (b)
Song, Q.; Wang, F.; Cai, J.; Wang, Y.; Zhang, J.; Yu, W.; Xu,
J. Energy Environ. Sci., 2013, 6, 904-1007; (c) Patwardhan,
P. R.; Brown, R. C.; Shanks, B. H. ChemSusChem, 2011, 4,
1629-1636; (d) Ma, R.; Hao,W.; Ma, X.; Tian, Y.; Li, Y. An-
gew. Chem. Int. Ed., 2014, 53, 1-7; (e) Bosch, S. V. D.;
Schutyser, W.; Vanholme, R.; Driessen, T.; Koelewijn, S. F.;
Renders, T.; Meeter, B. D.; Huijgen, W. J. J.; Dehaen, W.;
Courtin, C. M.; Lagrain, B.; Boerjan, W.; Sels, B. F. Energy
Environ. Sci., 2015, 8, 1748-1763; (f) Besson, M.; Gallezot,
P.; C. Pinel, Chem. Rev., 2014, 114, 1827-1870; (h) Ruppert,
A. M.; Weinberg, K.; Palkovits, R. Angew. Chem. Int. Ed.,
2012, 51, 2564-2601; (i) Wang, A. Q.; Zhang, T. Acc. Chem.
Res., 2013, 46, 1377–1386.
10
ACS Paragon Plus Environment

Page 11 of 11
ACS Catalysis
Table of Content:
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
11
ACS Paragon Plus Environment