Bimetallic Effects of Silver-Modified Nickel Catalysts and their Synergy in Glycerol Hydrogenolysis

Article abstract of DOI:10.1002/cctc.201600239

A series of Ag-modified Ni catalysts was prepared, and the role of bimetallic effects on the selective cleavage of C?O and C?C bonds was investigated in glycerol hydrogenolysis, a model reaction for the utilization of biomass-derived polyols. Compared to Raney Ni, the optimal Raney Ni₆Ag catalyst was more active in glycerol conversion and efficient for C?O bond cleavage to afford a higher selectivity toward C₃ products (1,2-propanediol and lactic acid; 88 %) at the same ～70 % conversion of glycerol, especially 1,2-propanediol (73 %). On Ni catalysts, the hydrogenation of adsorbed species was the most likely rate-limiting step in glycerol hydrogenolysis under alkaline conditions. The Ag additive enhanced the hydrogenation ability of Ni catalysts, which should be related to the higher strength of the adsorption of hydrogen and the lower strength of the adsorption of substrates on Ni sites because of the formation of Ni-Ag alloys. A mechanistic interpretation was presented for the excellent catalytic behavior.

Full text of DOI:10.1002/cctc.201600239

DOI: 10.1002/cctc.201600239
Full Papers
Bimetallic Effects of Silver-Modified Nickel Catalysts and
their Synergy in Glycerol Hydrogenolysis
Bin Chen,[a, b, c] Bin Zhang,[b, c] Yibo Zhang,*[a, b] and Xiangguang Yang*[a, b]
A series of Ag-modified Ni catalysts was prepared, and the role
of bimetallic effects on the selective cleavage of CÀO and CÀC
bonds was investigated in glycerol hydrogenolysis, a model re-
action for the utilization of biomass-derived polyols. Compared
to Raney Ni, the optimal Raney Ni6Ag catalyst was more active
in glycerol conversion and efficient for CÀO bond cleavage to
afford a higher selectivity toward C3 products (1,2-propanediol
and lactic acid; 88%) at the same ~70% conversion of glycerol,
especially 1,2-propanediol (73%). On Ni catalysts, the hydroge-
nation of adsorbed species was the most likely rate-limiting
step in glycerol hydrogenolysis under alkaline conditions. The
Ag additive enhanced the hydrogenation ability of Ni catalysts,
which should be related to the higher strength of the adsorp-
tion of hydrogen and the lower strength of the adsorption of
substrates on Ni sites because of the formation of Ni-Ag alloys.
A mechanistic interpretation was presented for the excellent
catalytic behavior.
Introduction
The selective hydrogenolysis of glycerol, the main byproduct
of biodiesel production by transesterification,[1] to useful chem-
icals (i.e., propanediols (1,2- and 1,3-PD)) has been significant
industrially.[2,3] This reaction is also very important as an appro-
priate model reaction to study the selective cleavage of CÀO
and CÀC bonds in the catalytic dehydroxylation of biomass-de-
rived polyols,[4–6] which is regarded as an important process for
future biorefineries.[7,8] To better understand the reaction
mechanism of hydrogenolysis as a guideline to select suitable
metal catalysts and cocatalysts to improve the product selec-
tivity, intensive studies on various metal catalysts, which
mainly include Ru-,[9–14] Cu-,[12,15–19] Pt-,[9,12,20,21] and Pd-based
catalysts,[22,23] for glycerol hydrogenolysis have been conduct-
ed. Ni is a promising catalyst for hydrogenolysis because of its
low cost, high activity, and resistance to poisoning[24,25] but it
shows poor hydrothermal stability compared to noble-metal
catalysts[26] and low hydrogenolytic activity toward CÀO bond
compared to Cu catalysts.[12]
Numerous strategies for the stabilization of Ni particles in
hydrothermal processes, such as aqueous-phase reforming
(APR), have been investigated. Raney Ni (R-Ni) is a prominent
hydrogenation catalyst in industrial application, not only for its
activity[27,28] but also for its dramatic stability and insolubility in
aqueous-phase reactions as it loses only ꢀ50% of its initial ac-
tivity for APR over 48 h at 2258C for which the catalysts com-
posed of Ni supported on Al2O3, SiO2, or ZrO2 lost ꢀ90% of
their initial activity (deactivation because of Ni sintering and
leaching).[29] The adjustment of the pH of the solutions with
base (i.e., KOH) is an elegant strategy to stabilize the Ni nano-
particles under aqueous solution conditions, and at pH>8,
achieved by the addition of at least 0.5m base, the Ni sintering
(particle growth) through Ostwald ripening as a result of the
acidic APR conditions was almost prevented, which corre-
sponds to the suppression of Ni leaching.[30]
As CÀO bond breaking is avoided by using Ni- and Ru-based
catalysts, which exhibit high hydrogenolysis activity for the CÀ
C bond in comparison with Cu-based catalysts and are ideal
for H2 production from biomass-derived feedstocks through
the APR process, the main difficulty in using Ni catalysts for
the production of desired C3 chemicals (i.e., 1,2-PD) through
CÀO bond rupture from glycerol hydrogenolysis is the selec-
tive cleavage of CÀO and CÀC bonds over Ni sites.[12,29–31] One
alternative to enhance the efficiency of Ni-based catalysts for
the formation of C3 products is to engineer the metal function
in the catalysts. This objective can be realized by the alloying
of Ni with other transition metals. Previous studies demonstrat-
ed that bimetallic catalysts exhibited a decreased tendency for
CÀC bond hydrogenolysis than the respective monometallic
ones.[15,32] Thus, the investigation of second metal components
that promote CÀO bond cleavage or inhibit CÀC bond cleav-
age is regarded as the key to accomplish a complete cleavage
[a] B. Chen, Dr. Y. Zhang, Prof. X. Yang
State Key Laboratory of Rare Earth Resources Utilization
Changchun Institute of Applied Chemistry
Chinese Academy of Sciences
Changchun 130022 (P.R. China)
E-mail: yibozhang@ciac.ac.cn
xgyang@ciac.ac.cn
[b] B. Chen, Dr. B. Zhang, Dr. Y. Zhang, Prof. X. Yang
Laboratory of Green Chemistry and Process
Changchun Institute of Applied Chemistry
Chinese Academy of Sciences
Changchun 130022 (P.R. China)
[c] B. Chen, Dr. B. Zhang
University of Chinese Academy of Sciences
Beijing 100049 (P.R. China)
Supporting information for this article can be found under:
http://dx.doi.org/10.1002/cctc.201600239.
ChemCatChem 2016, 8, 1929 – 1936
1929
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
ratio of CÀO/CÀC bonds in glycerol hydrogenolysis over Ni-
based catalysts.
In this work, a commercial R-Ni catalyst was modified with
Ag to develop a water-tolerant catalyst, which had better se-
lectivity toward C3 products (1,2-PD) and lactic acid (LA)
through CÀO bond cleavage. A certain amount of NaOH
(0.5m) was added to obtain stable Ni catalysts. The results of
catalyst characterization studies and catalytic activity measure-
ments allow us to elucidate the reaction mechanism in glycerol
hydrogenolysis and document the influence of Ag to control
the cleavage of CÀO and CÀC bonds over the Ni-based cata-
lysts.
Figure 1. H2-TPR profiles of a) Ag/Al2O3 calcined at 4008C, b) R-NiAg reduced
at 7008C, c) R-Ni6Ag reduced at 7008C, d) R-Ni34Ag reduced at 7008C, and
e) R-Ni reduced at 2608C.
Results and Discussion
Catalyst characterization
The characteristics of the catalysts used in this study are sum-
marized in Table 1. H2 temperature-programmed reduction
(TPR) profiles were measured to investigate the redox proper-
ties and phase composition of the catalysts. Four distinct
peaks appeared in the TPR profile of Ag/Al2O3 (Figure 1). The
reduction peaks in the temperature range of 100–4008C were
assigned to large Ag oxide particles located outside the Al2O3
matrix, and the other peaks at above 6208C were possibly
caused by the interaction of Ag with the Al2O3 support.[33]
Compared to Ag/Al2O3, R-Ni was reduced at ~2358C, which
shifted toward a higher temperature at around 2548C with the
addition of Ag. This indicated that a close interaction between
Ni and Ag could lead to the decrease of the reducibility of Ni.
The peak at above 3338C was assigned to the reduction of Ag
oxide, which was ascribed to Ag2O by Jeong and Kang.[34] If
the Ni/Ag ratio decreased from 34:1 to 6:1, the reduction peak
for Ag2O was shifted toward a higher temperature;[35] however,
the reduction peak for Ag2O shifted toward a lower tempera-
ture as more Ag was added (Ni/Ag=1:1).[36] This could be at-
tributed to the size shift of the Ag2O particles.[33]
XRD measurements were conducted to verify the metal crys-
tallites of the catalysts. The unmodified R-Ni that contained Sn
and Al showed typical diffraction peaks near 2q=44, 45, and
518, which were characteristic of Ni-Sn alloys,[29] Ni-Al alloy,[37,38]
and metallic Ni (200),[38] respectively (Figure 2A). The addition
of Ag to R-Ni had significant effects on the XRD pattern of the
catalysts. All the XRD patterns of R-NixAg catalysts had four
main characteristic peaks of face-centered cubic (fcc) crystalline
Figure 2. XRD patterns of catalysts; A) wide-scan XRD pattern and B) fine
scan of the Ag(111) peak: a) Ag/Al2O3 calcined at 4008C, b) R-NiAg reduced
at 7008C, c) R-Ni6Ag reduced at 7008C, d) R-Ni34Ag reduced at 7008C, and
e) R-Ni reduced at 2608C.
Table 1. Characterization of catalysts.
Catalyst
Ni[a] [wt%] Al[a] [wt%] Sn[a] [wt%] Ag[a] [wt%] Ag/Ni bulk atomic ratio[b] Ag/Ni surface atomic ratio[c] BET surface area [m2 gÀ1
]
R-Ni[d]
R-Ni34Ag
R-Ni6Ag
R-NiAg
17.5% Ag/Al2O3
64.1
50.2
45.6
30.8
–
12.7
16.6
14.3
13.2
–
20.4
30.5
27.6
21.5
–
0
0
0
21.1
42.4
40.2
32.6
103.7
2.7
12.6
34.5
–
0.03
0.15
0.61
–
2.30
2.71
4.87
–
[e]
[a] Determined by using ICP-OES. Others: Na, Sm, Mo (in trace amounts), which were not considered in Raney Ni-based catalysts. [b] Determined by using
ICP-OES. [c] Determined by using XPS. [d] Raney Ni prereduced at 2608C. [e] Catalyst from Ref. [58].
ChemCatChem 2016, 8, 1929 – 1936
www.chemcatchem.org
1930
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
Ag at around 2q=38.2, 44.4, 64.6, and 77.68, which can be in-
dexed to the (111), (200), (220), and (311) planes, respective-
ly.[39] Moreover, the peak of the Ag(111) plane shifted to
a higher angle, which indicated the formation of Ni-Ag alloys
after reduction (Figure 2B). This result was consistent with the
H2-TPR study that confirms the close interaction of Ni-Ag
(Figure 1).
X-ray photoelectron spectroscopy (XPS) was performed to
clarify the chemical state of the Ni and Ag species on the sur-
face of Ag-modified R-Ni, catalysts and the corresponding
Ni2p, Al2p, and Ag3d spectra of the R-Ni-based catalysts are
given in Figure 3. The Ni2p3/2 peak could be mainly deconvo-
luted into two peaks at binding energies (BEs) of 852.5 and
855.7 eV, which correspond well to Ni0 and Ni2+ species, re-
spectively.[37,40,41] However, the spectra presented in Figure 3A
suggested that a small amount of Ni species was concentrated
on the surface of the R-NixAg samples as there was no noticea-
ble peak in the XPS spectra. This can be attributed to three sit-
uations: (a) the pronounced Al2O3 enrichment (which existed
as hydrated alumina before calcination)[42] on the surface of
catalysts, as supported by the spectra shown in Figure 3B,
which was removed under reaction conditions (Figure S1), the
Al2p core-level BEs measured at around 74.4 and 72.6 were
typically ascribed to Al2O3 (Al3+) and metallic Al (Al0) species,
respectively;[37,41] (b) the metallic Al (Al0) enrichment on the
surface of Ni-Al alloys (because of the lower surface energy of
Al with respect to Ni)[37] (Figure S1); (c) the Ag species enrich-
ment on the surface of Ni-Ag alloys, which was because of the
lower surface energy of Ag with respect to Ni and reported to
be thermodynamically favorable by DFT studies and Monte
Carlo simulations,[43–46] whereas for other metal dopants such
as Pt there was Ni enrichment on the surface of the Ni-based
alloys (see Figure S2). The surface Ag/Ni atomic ratio of R-
Ni6Ag catalyst was 2.71, which was dramatically higher than
the bulk Ag/Ni atomic ratio of 0.15 and slightly increased as
more Ag was added (Table 1). Herein, all the cases above must
coexist. As the BE of bulk metallic Ag (Ag0) was 368.1 eV and
that of Ag2O (Ag+) was 367.7 eV,[47] the Ag present in the high-
valence state (Figure 3C) should be attributed to the passiva-
tion treatment, and an interaction between Ni and Ag could
not be ruled out (as supported by H2-TPR and XRD results).
H2 and NH3 temperature-programmed desorption (TPD)
were performed, and the results were compared for the R-Ni
and R-Ni6Ag catalysts to further analyze the changes of the hy-
drogenation activity and the acidity of the Ag-modified Ni cat-
alysts, respectively. The H2-TPD profiles showed that the addi-
tion of Ag increased the desorption temperature, which indi-
cates a stronger metalÀH bond on the surface of the R-Ni6Ag
catalysts than that on R-Ni as the temperature of desorption is
related to the metalÀH bond strength (Figure 4).[48] The shift of
electronic effects that results from the formation of Ni-Ag
alloys may lead to this decreasing desorption behavior. R-Ni is
known as a Ni-based alloy with few acidic functional groups.
The NH3-TPD profiles exhibited that the desorption peak shift-
ed to lower temperature, which indicated a weaker acid
strength on the surface of R-Ni6Ag catalysts than R-Ni, which
was attributed to a new acid site occurrence, likely Ag2O (Ag+),
Figure 3. XPS spectra of A) Ni2p, B) Al2p, and C) Ag3d: a) R-NiAg reduced at
7008C, b) R-Ni6Ag reduced at 7008C, c) R-Ni34Ag reduced at 7008C, and d) R-
Ni reduced at 2608C.
as shown by XPS (Figure 3C), on the surface of Ag-modified R-
Ni catalysts.
Low- and high-magnification SEM images of the R-Ni-based
catalysts demonstrated
a
predominant porous region
(Figure 5) in agreement with previous studies that showed
that fresh R-Ni is a high-surface-area catalyst that consists of
reduced Ni particles.[29] After drying and passivation, R-Ni had
À1
a
BET surface area of 21.1 m2 gcat
,
which increased to
42.4 m2 gcat as a small amount of Ag was added to R-Ni and
decreased slightly as more Ag was added (Table 1). In addition,
the BET results of R-NiAg suggested that the loading of large
amounts of Ag to R-Ni does not decrease the surface area of
the catalysts by pore blockage dramatically.
À1
ChemCatChem 2016, 8, 1929 – 1936
www.chemcatchem.org
1931
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
Notably, the Ag-modified R-Ni catalysts exhibited a higher
selectivity for the production of total C3 products, especially
1,2-PD, than R-Ni for glycerol hydrogenolysis after calcination
at 7008C (Table S1), which improved from 70.9 and 48.4% to
87.6 and 65.9% at a Ni/Ag ratio of 6:1 (Table 2, entries 1 and
3), respectively, and a LA selectivity of ~22% was maintained.
Actually, 1,2-PD can be generated from the hydrogenation of
2-hydroxyacrolein over metal sites, whereas LA must derive
from the cooperative production of metal and hydroxyl ions
through the Cannizzaro reaction from pyruvaldehyde, the
keto-form of 2-hydroxyacrolein (Table S2).[9,13,49,50]
As the product distribution is sensitive to the level of con-
version, the influence of the reaction time on glycerol hydro-
genolysis over R-Ni6Ag was investigated (Figure 6). At around
70% conversion of glycerol in the presence of base, R-Ni6Ag
promoted approximately 88 and 73% selectivity to total C3
products and 1,2-PD after 1 h, respectively. As the conversion
of glycerol neared completion, ethanol was observed as the
product of the further degradation of diols, which was consis-
tent with the findings of Montassier et al.[9,28,51] Methanation
was the main side reaction of glycerol hydrogenolysis, and
trace amounts of 1-propanol and methanol were observed in
the liquid phase.
Figure 4. H2- and NH3-TPD profiles of a) R-Ni6Ag reduced at 7008C and b) R-
Ni reduced at 2608C.
Figure 5. SEM images of catalysts shown at low and high magnification
(insets): A) R-Ni reduced at 2608C and B) R-Ni6Ag reduced at 7008C.
Figure 6. Glycerol conversion and product selectivity over R-Ni6Ag as a func-
tion of reaction time. Reaction conditions: 20 wt% glycerol, 0.5m NaOH,
substrate/catalyst=8 (weight ratio), pH =4 MPa (RT), T=2108C.
2
Catalytic performance
Table 2. Effect of Ag addition to Raney Ni catalysts.[a]
The effect of the addition of Ag to R-Ni catalyst on
the catalytic activity and C3 products selectivity is
shown in Table 2. In comparison with that over R-Ni,
the conversion of glycerol was promoted by the ad-
dition of small amount of Ag at a Ni/Ag ratio of 34:1,
which indicates that the addition of Ag improves the
overall rate of glycerol consumption (Table 2,
Entry Catalyst Conversion[b] Carbon balance[c]
Carbon selectivity
[%]
1,2-PD[b] LA[d] EG[b] Ethanol[b]
[%]
[%]
1
2
3
4
5
R-Ni[e]
R-Ni34Ag 78.8
R-Ni6Ag 78.0
R-NiAg
Ag/Al2O3 18.3
66.1
85.7
87.2
96.1
84.4
99.4
48.4
54.2
65.9
47.0
53.1
22.5 4.9
22.0 5.9
21.7 5.9
23.6 5.5
33.7 7.3
2.5
1.6
1.5
1.3
2.7
69.2
entry 2). Importantly, R-Ni6Ag showed
a higher
carbon balance than R-Ni in the hydrogenolysis of
glycerol, which increased from 85.7 to 96.1%, and
maintained a higher conversion at ꢀ78.0%. An in-
crease of the Ag content, however, had a negative
effect on the conversion, which decreased slightly
from 78.8 to 69.2% (Table 2, entries 2–4).
[a] Reaction conditions: 20 wt% glycerol, 0.5m NaOH, substrate/catalyst (weight
ratio)=8, pH =4 MPa (RT), T=2108C, t=6 h; 1,2-PD: 1,2-propanediol; EG: ethylene
2
glycol; LA: lactic acid (in the form of lactate). [b] Quantified by using GC–FID. Error
represents 95% confidence limits. [c] Others: methane, methanol, 1-propanol.
[d] Quantified by using HPLC. Error represents 95% confidence limits. [e] Raney Ni pre-
reduced under 2608C.
ChemCatChem 2016, 8, 1929 – 1936
www.chemcatchem.org
1932
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
Effects of Ag dopant: Understanding the network of glycer-
ol hydrogenolysis over Ni catalysts
The production of C3 products through CÀO bond rupture
was caused by the dehydration of glyceraldehyde before hy-
drogenation (for 1,2-PD production) or the Cannizzaro reaction
(for LA production; Scheme 1). Although undesirable CÀC
bond rupture occurred through Ni-catalyzed retro-aldol con-
densation that acted as the crucial barrier to obtain a high C3
product selectivity. Strategies to enhance the efficiency of Ni
catalysts in this study for catalytic dehydroxylation reaction
can be either metal function modification for the inhibition of
retro-aldol condensation or acid/base sites adjustment for the
enhancement of dehydration. As revealed by the H2-TPR, XRD,
and XPS results, a portion of the Ag species interacted with
the Ni moieties to form Ni-Ag alloys after reduction, whereas
other Ag species, likely Ag2O (Ag+), were present on the sur-
face of R-NixAg catalysts. Consequently, contiguous Ni ensem-
bles on the catalyst surface were diluted with Ag species,
which led to less contiguous Ni ensembles and more Ni mono-
mer sites on the surface. Contiguous Ni ensembles were re-
ported to be more active in CÀC bond cleavage than Ni mono-
mer sites, whereas Ni monomer sites were more suitable for
hydrogenation reactions.[55,56] Additionally, as suggested by the
weakened dehydrogenation properties of Ni catalysts with the
addition of Ag, the formation of Ni-Ag alloys may also inhibit
undesired CÀC bond cleavage by effectively decreasing the
strength of adsorption of the intermediates on Ni sites as the
bond energy of the COÀH bond (EOÀH ꢀ104 kcalmolÀ1) is gen-
erally higher than that of the CÀC bond (ECÀC ꢀ83 kcal
molÀ1).[57] However, the Ag catalyst was effective for the cleav-
age of the CÀO bond,[58] and the surface Ag species likely act
as new acid sites to promote the dehydration activity as R-Ni is
known as a Ni-based catalysts with few acidic functional
groups. Thus, it was significant that both the Ni metal function
and catalyst acidity were altered by the formation of Ni-Ag
alloys, and the synergistic effect of the Ni metal function and
catalyst acidity was proposed to be responsible for the im-
proved C3 product selectivity. However, Montassier et al. con-
cluded that CÀO bond cleavage proceeded through a base-
catalyzed instead of acid-catalyzed dehydration.[9,51] If this as-
sumption is reasonable, the dehydration process must be total-
ly directed by the added base. To confirm the role of catalyst
acidity on catalytic properties in the presence of base, in this
work, ZnO-modified R-Ni catalysts were prepared by an incipi-
ent impregnation method as the addition of ZnO to NiMo cat-
alysts could improve the catalytic activity and selectivity be-
cause of the Lewis acidity of ZnO.[59] The acidity of the catalysts
played a critical role to promote dehydration products but had
little effect on the rate of glycerol conversion in the presence
of base (Table S3).
On Ni-based catalysts, hydrogenolysis activity is highly exhibit-
ed in the CÀC bond and CÀO bond breaking is avoided.[12,31]
One way to design a better Ni catalyst for the catalytic dehy-
droxylation of biomass-derived polyols is to realize a higher
cleavage ratio of CÀO/CÀC bonds. The investigation of second
metal components for Ni-based catalysts is a promising option
to influence the tendency for CÀO and CÀC bond break-
ing.[15,32] The addition of Ag to R-Ni caused an increase in activ-
ity and superior selectivity toward C3 products through CÀO
bond rupture, especially 1,2-PD, for glycerol hydrogenolysis at
a Ni/Ag ratio of 6:1.
Proposed reaction mechanisms and selectivity challenges for
the dehydration products in glycerol hydrogenolysis at high
pH are outlined in Scheme 1. In the presence of base, the de-
Scheme 1. Proposed mechanism for the hydrogenolysis of glycerol over Ni-
based catalysts under alkaline conditions.[7]
hydrogenation of glycerol to glyceraldehyde intermediate on
the metal catalyst was suggested as the initial step and was
enhanced as more base was added.[9] Previous studies revealed
that Ag addition inhibited the dehydrogenation ability of Ni
catalysts because of the higher activation energy of Ag for al-
cohol dehydrogenation in comparison with Ni.[52,53] This was
consistent with the improved carbon balance and higher gly-
cols yield (1,2-PD and ethylene glycol; Table 2), which was at-
tributed to the suppression of further hydrogenolysis of prod-
ucts by inhibiting the dehydrogenation step.[54] However, from
the catalytic results, Ag-modified R-Ni catalysts were more
active in glycerol conversion, which suggests that the dehydro-
genation activity of R-Ni-based catalysts was not the key factor
to control the rate of glycerol conversion in the presence of
base.
Ag-modified R-Ni catalysts were efficient for 1,2-PD produc-
tion and maintained a LA selectivity of ~22% compared to R-
Ni (Table 2). As the formation of 1,2-PD proceeds through two
consecutive steps: dehydration of the glyceraldehyde on the
acid sites followed by hydrogenation of the produced 2-hy-
droxyacrolein/pyruvaldehyde intermediates on the Ni metal (or
bimetallic Ni-Ag), whereas LA is produced by the Cannizzaro
reaction from pyruvaldehyde and well controlled by base
(Table S2),[9,13,49,50] as depicted in Scheme 1, the activity of hy-
ChemCatChem 2016, 8, 1929 – 1936
www.chemcatchem.org
1933
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
drogenolysis and the selectivities to 1,2-PD and LA in C3 prod-
ucts were proposed to be sensitive to the hydrogenation activ-
ity of the Ni-based catalysts with controlled base addition. To
verify this point of view, herein, Pt-modified R-Ni catalysts were
prepared similarly according to the method for Raney Ni-Ag
catalysts as supported bimetallic Ni-Pt catalysts exhibited
a lower dehydrogenation activity,[60] but higher hydrogenation
activity than the monometallic Ni catalysts.[61,62] Compared to
R-Ni, at a higher conversion of glycerol around 90%, Pt-modi-
fied R-Ni catalysts promoted 62.5% selectivity to 1,2-PD and in-
hibited the selectivity to LA, which decreased from 22.5 to
14.3% at a Ni/Pt ratio of 17:1 (Table S4, entry 5). This indicated
that, instead of dehydrogenation or dehydration reactions, hy-
drogenation reactions over the Ni-based catalysts were the
most likely rate-limiting step in glycerol hydrogenolysis under
alkaline conditions. Moreover, the hydrogenation activity of
the Ni-based catalysts was crucial to control the selectivities to
1,2-PD and LA in C3 products. As reported previously, glycerol
hydrogenolysis performed over a Cu/SiO2 catalyst under basic
conditions (1m NaOH) yielded LA as the main product
(85%).[51] A Ru catalyst modified with sulfur yielded 75% of
1,2-PD and 13% of LA in the presence of NaOH after 2 h at
2408C.[63] A 64% selectivity to LA was obtained at a low con-
version over a Pt/C catalyst at 1808C and 4 MPa in the pres-
ence of CaO or NaOH.[9,51] Notably, the Ag additive can en-
hance the hydrogenation ability of Ni catalysts because of the
alloying of Ni with Ag. As suggested by surface science studies
and DFT calculations, the binding energies of unsaturated hy-
drocarbons on metals correlate with hydrogenation activity.
The formation of Ni-Ag alloys may not only decrease the
strength of the adsorption of unsaturated intermediates effec-
tively (indicated by the inferior dehydrogenation activity of Ag-
modified Ni catalysts) but it also enhances the strength of the
adsorption of hydrogen on Ni active sites (supported by H2-
TPD; Figure 4), which leads to an increased amount of ad-
sorbed hydrogen on the Ni metal for hydrogenation reactions.
Therefore, the addition of Ag to Ni catalysts introduced Ag
species on the surface of R-Ni-based catalysts and formed Ni-
Ag alloys after reduction, which led to a superior catalytic per-
formance (Scheme 2). However, the atomic ratio of Ni/Ag in R-
NixAg should be properly balanced in terms of catalytic activity
and C3 product selectivity. If the Ag-modified Ni catalysts were
prepared with the appropriate Ni/Ag atomic ratio of 6:1,
a higher selectivity to C3 products, especially 1,2-PD, was ach-
ieved, which increased from 70 and 48% to 88 and 73%, re-
spectively. Notably, for the production of C3 products through
CÀO bond breaking, especially 1,2-PD, a desirable Ni catalyst
should not only promote the hydrogenation activity of unsatu-
rated bonds (C=O and C=C) but also inhibit the dehydrogena-
tion activity to retard further hydrogenolysis and/or parallel
side reactions such as APR. The higher hydrogenation and infe-
rior dehydrogenation ability of Ni-based catalysts was crucial
to obtain a superior catalytic activity and yield of C3 products,
especially 1,2-PD, in glycerol hydrogenolysis, consistent with
a previous study on cellulose.[54] The effects of the doping of
Ag to Ni catalysts on the improvements of activity and selectiv-
ity in glycerol hydrogenolysis can be concluded as follows:
(1) ensemble effect, Ni atoms were substituted by Ag atoms
during preparation to form Ag species enrichment on the sur-
face of Ni-Ag alloys after calcination, which led to the dilution
of contiguous Ni sites that were more favorable for the cleav-
age of CÀC bonds than Ni monomer sites; (2) electronic effect,
the formation of Ni-Ag alloys increased the amounts of ad-
sorbed hydrogen through the decreased strength of adsorp-
tion of the substrates and/or the increased strength of adsorp-
tion of hydrogen on Ni sites, which resulted in the improved
hydrogenation ability but inhibited dehydrogenation ability of
Ni catalysts.
Conclusions
Raney Ni-Ag catalysts can be selected to achieve a good activi-
ty and selectivity for the production of C3 products, especially
1,2-propanediol, by the aqueous-phase hydrogenolysis of bio-
mass-derived glycerol. Ag-modified Raney Ni catalysts exhibit-
ed an increased activity for glycerol hydrogenolysis and en-
hanced selectivity to C3 products compared to the unmodified
Raney Ni catalyst. The most suitable Ni-based catalyst for CÀO
bond cleavage in glycerol hydrogenolysis was obtained at
a Ni/Ag atomic ratio of 6:1. Characterization of the catalysts by
using H2 temperature-programmed reduction, XRD, and X-ray
photoelectron spectroscopy indicated that the dilution of Ni
active sites and the existence of Ni-Ag alloys on the surface of
Ag-modified Ni catalysts were crucial for the excellent catalytic
performance. It is likely that further advances in the in situ
characterization of catalysts will lead to a deeper understand-
ing of reaction mechanisms and the generation of new cata-
lysts for the selective hydrogenolysis of CÀO and CÀC bonds in
the catalytic dehydroxylation of biomass-derived polyols.
Experimental Section
Catalyst preparation
The R-NixAg catalysts were prepared similarly according to
a method described elsewhere.[64,65] Briefly, commercial R-Ni (Da
Cheng Co., Ltd, China) was pretreated at 2608C for 2 h in a flowing
5% H2/Ar mixture (>99.99%, Changchun Juyang Gas Co., Ltd,
China), before the addition of appropriate amounts of AgNO3 (A.R.,
Sinopharm Chemical Reagent Co., Ltd, China) in deionized water
to reduced R-Ni under a N2 atmosphere (>99.999%, Changchun
Juyang Gas Co., Ltd, China), and subsequent heating in a 100 mL
stainless-steel autoclave with an inner Teflon coating to 1508C for
2 h. After reaction, the solvent was removed by centrifugation. The
catalyst was then washed and stored under deionized water.
Scheme 2. A schematic for preparation of R-Ni-Ag catalysts and their catalyt-
ic behavior in glycerol hydrogenolysis.
ChemCatChem 2016, 8, 1929 – 1936
www.chemcatchem.org
1934
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
Before we collected catalytic reaction data or catalyst characteriza-
tion, all catalysts were reduced in a flowing 5% H2/Ar mixture for
2 h at 7008C (heating at 58CminÀ1). Before contact with air, the
catalysts were passivated by flushing with 0.4% O2/Ar mixtures for
2 h at RT.
China). The HPLC system was equipped with a refractive index de-
tector (RID; RID-10A, Shimadzu) and an Aminex HPX-87H column
(300”7.8 mm, Bio-Rad) and 5 mm H2SO4 was used as the mobile
phase with a flow rate of 0.7 mLminÀ1 at 608C. The major liquid-
phase products detected by both GC–FID and HPLC were 1,2-PD,
ethylene glycol, and ethanol; other byproducts, which included
methanol and 1-propanol, were present in trace amounts. LA was
observed by HPLC. Methane was the main gas observed by GC–
TCD. The quantification of the liquid products was conducted by
using both GC–FID and HPLC based on calibration curves of the
standard compounds. The carbon balance was defined as the per-
centage of carbon accounted for in the liquid phase after 6 h reac-
tion. The conversion of glycerol and selectivity of each product
were determined as follows:
Catalyst characterization
XRD studies were conducted by using a Bruker D8 ADVANCE dif-
fractometer with a CuKa source (40 kV and 10 mA) in the 2q range
of 20–808 with a scan speed of 18minÀ1. The metal loading on the
catalysts was investigated by using inductively coupled plasma op-
tical emission spectroscopy (ICP-OES) by using a Thermo iCAP
6300 spectrometer. The BET surface areas of the catalysts were de-
termined by using N2 adsorption by using a Micromeritics ASAP-
2010 apparatus. SEM images were collected by using a Philips XL-
30 field-emission scanning electron microscope. XPS data were ob-
tained by using a Thermo ESCALAB 250 spectrometer equipped
with an Al source that operates at 15 kV and 150 W. The surface
concentrations were analyzed by using the Ni2p3/2 and Ag3d5/2
peak areas and sensitivities.[64,66]
moles of consumed glycerol
moles of initial glycerol
conversion ð%Þ ¼
selectivity ð%Þ ¼
 100
moles of carbon in specific product
moles of carbon in consumed glycerol
 100
ð1Þ
TPR was detected by using a TCD-GC (GC-8A, Shimadzu). TPD was
measured by using a mass spectrometer (QIC-20, Hiden). Before
the H2-TPR run, 50 mg of the sample was pretreated with air (>
99.99%, Changchun Juyang Gas Co., Ltd, China) at 1508C for 1 h.
After cooling to RT, the sample was flushed for 1 h to reach
a stable background and then heated to 8508C (heating at
108CminÀ1) in flowing 5% H2/Ar mixtures (50 cm3 (STP)minÀ1). For
the H2-TPD analysis, 50 mg of the sample was prereduced for 1 h
at the appropriate temperature in flowing 5% H2/Ar mixtures.
After cooling to RT, the sample was flushed for 1 h to reach
a stable background and then heated to 8508C (heating at
108CminÀ1) in flowing Ar (>99.999%, Changchun Juyang Gas Co.,
Ltd, China; 50 cm3 (STP)minÀ1). For NH3-TPD, 50 mg of the sample
was pretreated with air at 1508C for 1 h. After cooling to RT, the
sample was flushed with 0.2% NH3/Ar mixtures for 30 min and
then Ar for 1 h to reach a stable background. Desorption was then
Acknowledgements
The authors gratefully acknowledge the financial support of the
National Natural Science Foundation of China (21177120).
Keywords: biomass
· heterogeneous catalysis · nickel ·
reaction mechanisms · silver
[1] C. W. Chiu, M. A. Dasari, W. R. Sutterlin, G. J. Suppes, Ind. Eng. Chem. Res.
2006, 45, 791.
[2] T. Werpy, G. Petersen, Top value added chemicals from biomass, Vol. 1,
Results of screening for potential candidates from sugars and synthesis
gas, US DOE Report, 2004.
[3] Y. Nakagawa, K. Tomishige, Catal. Sci. Technol. 2011, 1, 179.
[4] J. ten Dam, U. Hanefeld, ChemSusChem 2011, 4, 1017.
[5] A. M. Ruppert, K. Weinberg, R. Palkovits, Angew. Chem. Int. Ed. 2012, 51,
2564; Angew. Chem. 2012, 124, 2614.
[6] Y. H. Shen, S. Wang, C. Luo, H. Liu, Prog. Chem. 2007, 19, 431.
[7] C. Luo, S. Wang, H. Liu, Angew. Chem. Int. Ed. 2007, 46, 7636; Angew.
Chem. 2007, 119, 7780.
performed by heating the sample to 8508C (heating at 108CminÀ1
)
in flowing Ar (50 cm3 (STP)minÀ1). The TPD curves were related to
the monitored ion intensities of m/z=2 (H2+) and 17 (NH3+).
Catalytic reaction
[8] J. Sun, H. Liu, Green Chem. 2011, 13, 135.
[9] E. P. Maris, R. J. Davis, J. Catal. 2007, 249, 328.
[10] D. G. Lahr, B. H. Shanks, J. Catal. 2005, 232, 386.
[11] T. Miyazawa, Y. Kusunoki, K. Kunimori, K. Tomishige, J. Catal. 2006, 240,
213.
[12] M. A. Dasari, P. P. Kiatsimkul, W. R. Sutterlin, G. J. Suppes, Appl. Catal. A
2005, 281, 225.
[13] J. Feng, W. Xiong, B. Xu, W. D. Jiang, J. B. Wang, H. Chen, Catal.
Commun. 2014, 46, 98.
[14] Y. M. Li, L. Ma, H. M. Liu, D. H. He, Appl. Catal. A 2014, 469, 45.
[15] J. Chaminand, L. Djakovitch, P. Gallezot, P. Marison, C. Pinel, C. Rosier,
Green Chem. 2004, 6, 359.
The hydrogenolysis experiments were performed in a stainless-
steel autoclave with a capacity of 50 mL. The reactant solutions
(20 wt% glycerol (99%, Aladdin), solvent: H2O, with base additives)
and catalysts were placed in the reactor to give a total volume of
ꢀ20 mL. Before it was put into a silicone oil bath, the reactor was
flushed with H2 (>99.999%, Changchun Juyang Gas Co., Ltd,
China) and pressurized to 4.0 MPa (RT). After the reactor was
heated to 2108C in the oil bath, stirring with a magnetic stirrer at
1200 rpm was started, which was defined as the beginning of the
reaction. The reaction was maintained under these conditions for
6 h to ensure that a steady state was achieved.
[16] L. Y. Guo, J. X. Zhou, J. B. Mao, X. W. Guo, S. G. Zhang, Appl. Catal. A
2009, 367, 93.
[17] E. S. Vasiliadou, T. M. Eggenhuisem, P. Munnik, P. E. de Jongh, K. P.
de Jong, A. A. Lemonidou, Appl. Catal. B 2014, 145, 108.
[18] S. H. Zhu, X. Q. Gao, Y. L. Zhu, Y. W. Li, Green Chem. 2016, 18, 782.
[19] M. L. Dieuzeide, M. Jobbagy, N. Amadeo, Ind. Eng. Chem. Res. 2016, 55,
2527.
[20] J. ten Dam, K. Djanashvili, F. Kapteijn, U. Hanefeld, ChemCatChem 2013,
5, 497.
[21] W. Oberhauser, C. Evangelisti, R. P. Jumde, R. Psaro, F. Vizza, M. Bevilac-
qua, J. Filippi, B. F. Machado, P. Serp, J. Catal. 2015, 325, 111.
After the reaction, the reactor was cooled to RT. Gas samples were
collected with a gas-bag and analyzed by using GC with a thermal
conductivity detector (TCD; GC-8A, Shimadzu) equipped with an
active carbon column. Liquid samples from the reaction mixture
after centrifugation were analyzed by using GC with flame ioniza-
tion detection (FID; GC-8A, Shimadzu) and HPLC (LC-20AB, Shimad-
zu). The GC–FID was equipped with a polyethylene glycol (PEG)
column (20 m”0.25 mm”0.25 mm; Dalian Zhonghuida Co., Ltd,
ChemCatChem 2016, 8, 1929 – 1936
www.chemcatchem.org
1935
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
[22] T. Jiang, Q. Huai, W. Y. Ying, T. C. Xiao, F. H. Cao, Biomass Bioenergy 2015,
78, 71.
[23] X. R. Li, C. Zhang, H. Y. Cheng, W. W. Lin, P. J. Chang, B. Zhang, Q. F. Wu,
Y. C. Yu, F. Y. Zhao, ChemCatChem 2015, 7, 1322.
[24] B. C. Miranda, R. J. Chiment¼o, J. B. O. Santos, F. Gispert-Guirado, J.
Llorca, F. Medina, F. L. Bonillo, J. E. Sueiras, Appl. Catal. B 2014, 147, 464.
[25] W. Yu, J. Zhao, H. Ma, H. Miao, Q. Song, J. Xu, Appl. Catal. A 2010, 383,
73.
[46] F. Calvo, E. Cottancin, M. Broyer, Phys. Rev. B 2008, 77, 121406.
[47] L. H. Tjeng, M. B. J. Meinders, J. Van Elp, J. Ghijsen, G. Sawatzsky, R. L.
Johnson, Phys. Rev. B 1990, 41, 3190.
[48] J. Greeley, M. Mavrikakis, J. Phys. Chem. B 2005, 109, 3460.
[49] J. Ten Dam, F. Kapteijn, K. Djanashvili, U. Hanefeld, Catal. Commun.
2011, 13, 1.
[50] F. Auneau, C. Michel, F. Delbecq, C. Pinel, P. Sautet, Chem. Eur. J. 2011,
17, 14288.
[26] T. van Haasterecht, C. C. I. Ludding, K. P. de Jong, J. H. Bitter, J. Energy
Chem. 2013, 22, 257.
[51] C. Montassier, D. Giraud, J. Barbier, Polyol conversion by liquid phase het-
erogeneous catalysis over metals, Elsevier, Amsterdam, 1988, p. 165.
[52] I. Brihta, P. Luetic, Croat. Chem. Acta 1957, 29, 419.
[53] P. Fuderer-Leutic, I. Brihta, Croat. Chim. Acta. 1959, 31, 75.
[54] G. F. Liang, L. M. He, H. Y. Cheng, W. Li, X. R. Li, C. Zhang, Y. C. Yu, F. Y.
Zhao, J. Catal. 2014, 309, 468.
[55] A. Roberti, V. Ponec, W. M. H. Sachtler, J. Catal. 1973, 28, 381.
[56] J. Sinfelt, J. Catal. 1973, 29, 308.
[57] Y. R. Luo, Comprehensive Handbook of Chemical Bond Energies, CRC,
Boca Raton, 2007.
[27] A. Perosa, P. Tundo, Ind. Eng. Chem. Res. 2005, 44, 8535.
[28] A. Y. Yin, X. Y. Guo, W. L. Dai, K. N. Fan, Green Chem. 2009, 11, 1514.
[29] J. W. Shabaker, G. W. Huber, J. A. Dumesic, J. Catal. 2004, 222, 180.
[30] T. van Haasterecht, C. C. I. Ludding, K. P. de Jong, J. H. Bitter, J. Catal.
2014, 319, 27.
[31] M. Besson, P. Gallezot, C. Pinel, Chem. Rev. 2014, 114, 1827.
[32] W. Q. Yu, J. Xu, H. Ma, C. Chen, J. Zhao, H. Miao, Q. Song, Catal.
Commun. 2010, 11, 493.
[33] Z. P. Qu, X. D. Zhang, F. L. Yu, X. C. Liu, Q. Fu, J. Catal. 2015, 321, 113.
[34] H. Jeong, M. Kang, Appl. Catal. B 2010, 95, 446.
[58] J. X. Zhou, J. Zhang, X. W. Guo, J. B. Mao, S. G. Zhang, Green Chem.
2012, 14, 156.
[35] N. V. Parizotto, K. O. Rocha, S. Damyanova, F. B. Passos, D. Zanchet,
C. M. P. Marques, J. M. C. Bueno, Appl. Catal. A 2007, 330, 12.
[36] D. L. Sun, Y. Yamada, S. Sato, Appl. Catal. A 2014, 475, 63.
[37] H. Lei, Z. Song, X. H. Bao, X. H. Mu, B. N. Zong, E. Z. Min, Surf. Interface
Anal. 2001, 32, 210.
[38] H. Hu, M. H. Qiao, S. Wang, K. N. Fan, H. X. Li, B. N. Zong, X. X. Zhang, J.
Catal. 2004, 221, 612.
[39] X. J. Song, D. M. Zhang, Energy 2014, 70, 223.
[40] K. W. Park, J. H. Choi, B. K. Kwon, S. A. Lee, Y. E. Sung, H. Y. Ha, S. A.
Hong, H. Kim, A. Wieckowski, J. Phys. Chem. B 2002, 106, 1869.
[41] A. Iriondo, J. F. Cambra, V. L. Barrio, M. B. Guemez, P. L. Arias, M. C. San-
chez-Sanchez, R. M. Navarro, J. L. G. Fierro, Appl. Catal. B 2011, 106, 83.
[42] S. D. Robertson, J. Freel, R. B. Anderson, J. Catal. 1972, 24, 130.
[43] A. Rapallo, G. Rossi, R. Ferrando, A. Fortunelli, B. C. Curley, L. D. Lloyd,
G. M. Tarbuck, R. L. Johnston, J. Chem. Phys. 2005, 122, 194308.
[44] M. Harb, F. Rabilloud, D. Simon, J. Chem. Phys. 2009, 131, 174302.
[45] M. Gaudry, E. Cottancin, M. Pellarin, J. LermØ, L. Arnaud, J. R. Huntzinger,
J. L. Vialle, M. Broyer, J. L. Rousset, M. Treilleux, P. MØlinon, Phys. Rev. B
2003, 67, 155409.
[59] J. Y. Hu, Y. Q. Fan, Y. Pei, M. H. Qiao, K. N. Fan, X. X. Zhang, B. N. Zong,
ACS Catal. 2013, 3, 2280.
[60] M. Salciccioli, W. Yu, M. A. Barteau, J. G. Chen, D. G. Vlachos, J. Am.
Chem. Soc. 2011, 133, 7996.
[61] L. He, Y. Q. Huang, A. Q. Wang, Y. Liu, X. Y. Liu, X. W. Chen, J. J. Delgado,
X. D. Wang, T. Zhang, J. Catal. 2013, 298, 1.
[62] W. W. Lonergan, D. G. Vlachos, J. G. Chen, J. Catal. 2010, 271, 239.
[63] B. Casale, A. M. Gomez, Patent EP 92830357.7.
[64] J. W. Shabaker, D. A. Simonetti, R. D. Cortright, J. A. Dumesic, J. Catal.
2005, 231, 67.
[65] X. Q. Pan, Y. B. Zhang, Z. Z. Miao, X. G. Yang, J. Energy Chem. 2013, 22,
610.
[66] A. Y. Yin, C. Wen, W. L. Dai, K. N. Fan, Appl. Catal. B 2011, 108, 90.
Received: February 29, 2016
ChemCatChem 2016, 8, 1929 – 1936
www.chemcatchem.org
1936
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim