Selective Conversion of Cellulose to Hydroxyacetone and 1-Hydroxy-2-Butanone with Sn–Ni Bimetallic Catalysts

Article abstract of DOI:10.1002/cssc.201900172

The high-value-added chemicals hydroxyacetone (HA) and 1-hydroxy-2-butanone (HB) were produced from agricultural waste over a Ni₃Sn₄-SnO_x catalyst. The Sn–Ni intermetallic compound and SnO_x acted as the active sites for HA and HB production by selectively cleaving the target C?C and C?O bonds. Approximately 70 % of the total HA and HB yield was obtained by selective hydrogenolysis of cellulose. This strategy expands the application of cellulose towards renewable production of high-value C₃ and C₄ keto-alcohols from cellulosic biomass.

Full text of DOI:10.1002/cssc.201900172

Accepted Article
Title: Selective Conversion of Cellulose to Hydroxyacetone and 1-
Hydroxy-2-Butanone Using Sn-Ni Bimetal Catalysts
Authors: Longlong Ma, Haiyong Wang, Changhui Zhu, Qiying Liu, Jin
Tan, Chenguang Wang, and Zheng Liang
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: ChemSusChem 10.1002/cssc.201900172
Link to VoR: http://dx.doi.org/10.1002/cssc.201900172
A Journal of
www.chemsuschem.org

10.1002/cssc.201900172
ChemSusChem
Selective Conversion of Cellulose to Hydroxyacetone and 1-
Hydroxy-2-Butanone Using Sn-Ni Bimetal Catalysts
Haiyong Wang1, 2, 3, 4, Changhui Zhu1, 2, 3, 4, Qiying Liu1, 2, 3,5*, Jin Tan1, 2, 3, Chenguang Wang1, 2, 3
Zheng Liang1,2,3, Longlong Ma1, 2, 3
,
*
1 Guangzhou Institute of Energy Conversion, Chinese Academy of Sciences, Guangzhou 510640,
P. R. China
2 CAS Key Laboratory of Renewable Energy, Guangzhou 510640, P. R. China
3 Guangdong Key Laboratory of New and Renewable Energy Research and Development,
Guangzhou 510640, P. R. China
4 University of Chinese Academy of Sciences, Beijing 100049, P. R. China
5 Dalian National Laboratory for Clean Energy, Dalian 116023, P. R. China
*Author for correspondence: Qiying Liu (liuqy@ms.giec.ac.cn), Longlong Ma
(mall@ms.giec.ac.cn)
Hydroxyacetone (HA) and 1-hydroxy-2-butanone (HB), both of which contain
carbonyl and hydroxyl groups, are becoming increasingly important as intermediates,
due to their diversified applications in fine chemicals production such as medicine and
food additives. We developed a strategy to obtain these two products from direct
cellulose conversion over Sn-Ni intermetallic compound (IMC) catalysts. It was
confirmed that the Sn-Ni IMC and SnOx acted as the active sites for HA and HB
production via selectively cleaving the target C-C and C-O bonds. About 70 % of total
HA (33.2%) and HB (35.6%) yield could be obtained by selective hydrogenolysis of
cellulose. This strategy expands the application of cellulose for affording a new route
to renewably produce the high valued C3-C4 keto-alcohols from cellulosic biomass.
With the growing shortage of fossil resources and the increase in greenhouse gas
emissions, developing green, rich, renewable carbon sources to replace fossil carbon is
the tendency in both the academic and industrial communities.[1,2] Biomass resources
are the abundant and sustainable carbon source in nature for the production of high
valued platform chemicals.[3,4] Cellulose is the most abundant component in
1
This article is protected by copyright. All rights reserved.

10.1002/cssc.201900172
ChemSusChem
lignocellulosic biomass. Its linear polysaccharide polymer with numerous hydroxyl
groups makes it act as a favorable feedstock for the production of high value-added fine
chemicals.[5-14] Fine chemicals, such as pesticides, coatings, drug intermediates, and
food additives, used in daily life, are almost petrochemical products.[15] Currently,
developing new catalytic strategy to match with the conversion of biomass are
increasingly attracted.[16,17]
Most researchers focused on studies for enhancing the activities of hydrogenation
and selectively cleaving C-C bond of the catalysts to catalyze the hydrolysis of cellulose.
Traditional precious-metal catalysts, such as Ru, Pt, Pd are usually used, and they
exhibited relatively high activity. Due to the high activity of the catalysts, the
production from cellulose are polyols and C2-C6 alcohols.[18-21] For some compounds,
the unsaturated C=O bond makes them exhibit special nature and application.
Accordingly, a bifunctional catalyst with selectively breaking C-C bond and partial
hydrogenation would be a good choice for biomass catalytic conversion. Keto-alcohols,
which possess hydroxyl and activity unsaturated carbonyl groups, are among the high
value-added organic compounds and could potentially be used for many organics
synthesis. In particularly, hydroxyacetone (HA) is an important intermediate used to
produce polyols, acrolein, methylglyoxal, acetone and lactic acid or directly used as
food addictive, reducing agent and so on.[22-26] The HA produced mainly by oxidation
of 1,2-propanediol which is produced mainly from petro-propylene via epoxy
compounds.[25,27,28] It also can be obtained by direct conversion cellulose over Ni based
catalysts such as Ni-SnOx/Al2O3 or signal-atom Ni catalysts.[27,29,30] In Liu’s work, they
2
This article is protected by copyright. All rights reserved.

10.1002/cssc.201900172
ChemSusChem
investigated the catalytic performances of Ni-SnOx of with Sn/Ni ratios in a range of 0-
2.0, and they proposed that the SnOx species were responsible for the selective cleavage
of the C-C bonds in glucose and fructose.[23] 1-hydroxy-2-butanone (HB) is mainly used
to synthesis ethambutol, a most common anti-tuberculosis medicine.[30] It is produced
mainly by oxidation of 1,2-butanediol over CuO or enzyme catalyst.[31] The HA and
HB are also found in the preparation of polyols by non-selective pyrolysis of biomass,
while the both amounts are small with the yields below 10%.[32] For producing HA and
HB from direct cellulose hydrogenolysis, developing new catalyst is critical for
selectively cleaving the target C-C bond while protecting the target C=O and C-OH
bonds survived from partial hydrogenation.
In this work, we for the first time report the synchronized production of HA and
HB with a total yield of 70 % through selective cleavage of the C-C bonds by retro-
aldol condensation over a Ni3Sn4-SnOx catalyst in water. The bimetallic Sn-Ni IMC
was prepared by controlling the Sn/Ni ratio and annealing temperatures. The formation
of Ni3Sn4 and their synergistic effect with Sn oxide species endow the catalysts
unexpected performance in hydrogenolysis of cellulose.
The investigations were started with hydrogenolysis of cellulose on single metal
and bimetal Sn-Ni IMC catalysts. The influence of reaction conditions such as reaction
time, temperature, and initial H2 pressure was also investigated (Figure 1, Figure S1
and Table S3). The maximum yield of keto-alcohol was obtained at 240 oC, 4 MPa H2,
and with a reaction time of 1 h. Thus, all the catalysts were evaluated under these
optimized conditions. Table 1 showed the results of cellulose conversion and the
3
This article is protected by copyright. All rights reserved.

10.1002/cssc.201900172
ChemSusChem
products distribution over various Sn-Ni catalysts. Catalysis by Ni@C demonstrated a
low conversion of cellulose (52.3 %), and the main products were ethylene glycol (EG,
15.3 %) and hexitols (18.3 %). Other products including glucose, sorbitol, 1,2-
propanediol (1,2-PDO), glycerol (Gly), lactic acid (La), HA, HB, methylglyoxal (MG)
and acetoin (Ace), and the total yield of these products was very low (< 5 %). Only
small amount products were detected after the cellulose conversion with Sn@C as the
catalyst. The conversion of cellulose was 30 % while the total yield of keto-alcohols
was only 5 % (Entry 3, Table 1). Compared to the experiment in which the Ni@C was
solely used (Entry 1, Table 1), the products distribution and yields remained almost
unchanged over the Ni@C+Sn@C, and EG and hexitol were the main products (Entry
4, Table 1). It is interesting for the 3Sn-Ni@CT catalysts. When the metal Sn and Ni
were introduced into the catalyst simultaneously, the distribution of products
significantly changed (Entry 4-9, Table 1). The cellulose conversion has increased to
more than 90 %. The yields of HA and HB were obviously increased over the 3Sn-
Ni@CT catalysts. At low annealing temperature, the total yield of HA and HB was 36.2 %
(Entry 5, Table 1). As the annealing temperature of catalyst increased, the yield of HA
and HB increased. The highest total yield of keto-alcohols was up to 72 % over the 3Sn-
Ni@C600 catalyst (Entry 8, Table 1). However, with further increasing the annealing
temperature, the yields were decreased to 39.3 %. The carbon balance based on total
organic carbon (TOC) analysis of the liquid phase showed that there were some
amounts of carbon missing. The missing carbon was due to the formation of gas
products and the possibly unidentified products (e.g. humins) derived from the
4
This article is protected by copyright. All rights reserved.

10.1002/cssc.201900172
ChemSusChem
condensation or degradation of glucose and other sugar intermediates. [33]
Table 1. Conversion of cellulose and the products distribution over different NiSn@C catalysts
Conv.
(%)
Yield (C-mol%)
Carbon balance
(C-mol%)
Catalyst
EG
1,2-PDO
Gly
La
HA
MG
Ace
HB
Tk-a
C6
Ni@C
52.3
67.2
30.1
62.9
70.1
82.1
90.2
96.2
95.4
94.8
15.3
9.1
0
3.1
11.6
0
0.2
0
0
12.6
0
0.3
0
0.4
0
0.1
0
0.4
0
1.2
0
18.3
10.1
0.8
75.1
79.8
25.9
62.8
63.4
52.7
70.7
77.4
49.3
43.0
Ni-NiO@Ca
Sn@C
0
3.8
0
0
1.2
0.2
1.1
5.0
0.2
19.1
Sn@C+Ni@C
3Sn-Ni@C300
3Sn-Ni@C400
3Sn-Ni@C500
3Sn-Ni@C600
3Sn-Ni@C700
3Sn-Ni@C800
15.6
4.3
0.1
0.2
0.1
2.5
3.6
3.4
10.4
0.1
0.4
0.6
0.6
0.7
0
0
0
0
0
20.1
7.1
0.4
0.6
0.8
1.0
0.7
0.5
3.3
0.9
1.1
1.3
0.6
0.4
15.3
16.1
30.4
33.2
21.8
18.9
0.1
0.2
0.3
0.4
0.3
0.2
2.6
1.1
2.2
2.1
1.9
1.2
18.6 36.2
23.7 57.2
35.6 72.0
18.9 46.0
14.7 39.3
5.6
4.7
0.2
0.3
0.6
Reaction conditions: cellulose 0.2 g; catalyst 0.08 g; deionized water 20 mL; reaction temperature
o
240 C; reaction time 1 h and initial H2 pressure 4.0 MPa. Tk-a total yield of keto-alcohols; EG,
ethylene glycol; 1,2-PDO, 1,2-propanediol; Gly, glycerol; La, lactic acid; HA, hydroxyacetone; MG,
methylglyoxal; Ace, acetoin; HB, 1-hydroxy-2-butanone; C6, sorbitol.
a: main other products, glyceric acid 10.2 %
Screening on xSn-Ni@C600 catalysts, the catalytic performances of Sn/Ni catalyst
with different atomic ratio are investigated (Figure 1 A). Upon addition of a small
amount of Sn to Ni@C at the Sn/Ni ratio of 1.0, the yield of EG and sorbitol decreased
rapidly from 33.6 % to 8.8 %, meanwhile, the yields of C3-C4 keto-alcohols
significantly increased to 41.1 %. When Sn/Ni ratio was increased to 3.0, the sorbitol
yield sharply decreased to 0.5 % with the concurrent increase of the C3-C4 keto-alcohols
to 72 % for obtaining the main respective products of HA (33.2 %) and HB (35.6 %).
However, further increasing the Sn/Ni ratios to 4.0 and 5.0 led to an obvious decrease
in the C3-C4 keto-alcohols yields to 46.3% and 16.1 %, respectively. The catalytic
5
This article is protected by copyright. All rights reserved.

10.1002/cssc.201900172
ChemSusChem
behavior of the 3Sn-Ni@C600 catalyst is shown in Figure 1B, HA and HB were formed
at yields of 19.7% and 21.1% in 30 min, respectively. As the precursor of HA,
glyceraldehyde was detected during the reaction with the yield of 5.1% and then
disappeared at 1 h. The HA and HB were produced at yields of 33.2% and 35.6% in 1
1, respectively. Further prolonging the reaction time, the yield of keto-alcohols
decreased slightly.
80
60
40
20
0
100
80
60
40
20
0
MG
HA
La
Gly
1,2-PDO
EG
Ace
HB
100
80
60
40
20
0
100
80
60
40
20
0
A
B
C6
HB
Carbon balance
Ace
MG
HA
La
Gly
Carbon balance
C6
1,2-PDO
EG
@C
0.0
0.5
1.0
1.5
2.0
2.5
3.0
Ni@C
Sn
n-Ni@C n-Ni@C n-Ni@C n-Ni@C n-Ni@C
Reaction time (h)
1S
2S
3S
4S
5S
Figure 1. Effect of Sn/Ni atomic ratios (A) and reaction time over 3Sn-Ni@C600 (B) on cellulose
conversion and products yield.
Reaction conditions: cellulose 0.2 g; catalyst 0.08 g; deionized water 20 mL; reaction temperature
240 oC; reaction time (A) 1 h, (B) 0.5-2.5 h, and initial H2 pressure 4.0 MPa. EG, ethylene glycol;
1,2-PDO, 1,2-propanediol; Gly, glycerol; La, lactic acid; HA, hydroxyacetone; MG, methylglyoxal;
Ace, acetoin; HB, 1-hydroxy-2-butanone; C6, sorbitol, glucose.
Figure 2A showed the XRD patterns of monometallic and bimetallic Sn-Ni
catalysts. Clearly, the monometallic Ni@C sample exhibited the characteristic (111)
and (200) diffractions of crystalline metallic nickel (JCPDS 04-0850). With Sn
introduced (Sn/Ni ratio of 1.0 and 2.0), some new peaks which attributed to Sn-Ni IMC
(Ni3Sn2 and Ni3Sn4) and SnOx could be observed. The dominant Sn-Ni phase in these
6
This article is protected by copyright. All rights reserved.

10.1002/cssc.201900172
ChemSusChem
two catalysts was Ni3Sn2. In addition, the absence of metal Ni also reflected the
formation of bimetallic Sn-Ni IMC phases. At the Sn/Ni atomic ratio of 3.0 and 4.0, the
Sn-Ni IMC existed primarily in the form of Ni3Sn4. Further increasing the Sn/Ni ratio
to 5.0, the diffractions of Sn-Ni IMC disappeared, and diffractions of Sn oxide species
were observed. While for the Sn@C catalyst, the Sn species existed primarily in the
form of SnOx and metallic Sn. The annealing temperature also affected the component
of the Sn-Ni catalysts. Sn-Ni IMC catalysts were prepared via reduction-carbonation of
the precursor at different annealing temperatures in a nitrogen atmosphere. During the
carbonation process, CH4, CO and C from the precursor reduced the Sn2+ and Ni2+ to
metal Sn and Ni (Figure S2). Subsequently, the Sn-Ni IMC formed with the temperature
increased. As shown in Figure S3, the main peaks of 3Sn-Ni@C300 were SnOx and
o
Ni3Sn2. With the annealing temperature increased from 300 to 600 C, the peaks of
SnOx and Ni3Sn2 decreased and the intensity of the characteristic diffraction peaks
corresponding to Ni3Sn4 increased. The BET surface of the catalysts was also increased
o
from 2.1 to 403.5 m2/g with the annealing temperature increased from 300 C to 600
oC. Correspondingly, the keto-alcohol yields increased from 19.1 % to 72.0 %. These
results (Entry 5, Table 1) showed that the Ni3Sn4, rather than the Ni3Sn2, is the active
phase for the formation of HA and HB. However, further increasing the annealing
temperature led to a significant loss in keto-alcohols yield. The decrease of Ni3Sn4
content (Figure S3 and Table S1) may be responsible for the reduction of the keto-
alcohols.
7
This article is protected by copyright. All rights reserved.

10.1002/cssc.201900172
ChemSusChem
SnOx
A
B

Sn/C
Sn


Sn2+/4+
487.4
Sn3d5/2
Sn0
485.3
5Sn-Ni/C
5Sn-Ni/C
4Sn-Ni/C
4Sn-Ni/C
3Sn-Ni/C
2Sn-Ni/C
3Sn-Ni/C
Ni3Sn4
2Sn-Ni/C


1Sn-Ni/C
Ni3Sn2
1Sn-Ni/C
Ni/C
20
30
40
50
2 (o)
60
70
80 500
495
490
485
480
Binding Energy (eV)
Figure 2. XRD patterns (A) and Sn 3d (B) of xSn-Ni@C catalysts annealed at 600 oC.
The surface chemical states of the catalysts were studied by XPS analysis. As
shown in Figure 2B, two groups of peaks were detected on the surface of xSn-Ni@C600
catalysts. The higher binding energy at 487.4 eV (Sn 3d5/2) and 495.9 eV (Sn 3d3/2)
were assigned to the oxidized Sn species (Sn2+ and Sn4+), but the distinction between
Sn2+ and Sn4+ was difficult owing to the negligible difference in the binding energies.[34]
The other group of peaks with binding energy at 485.3 eV (Sn 3d5/2) and 493.8 eV (Sn
3d3/2) were attributed to the metallic Sn species (Sn0). The XPS results indicated that,
the Sn species on the Sn-Ni catalysts surface mainly existed in the form of SnOx, and a
not negligible part (2-20 %) Sn existed in the form of Sn-Ni IMC (Table S1). The Ni
2p3/2 region comprises the main peak, the shoulder, and its satellites at 853.0, 856.2 and
862.0 eV, respectively (Figure S4). The peak around 856 eV is attributed to Niσ+, which
could act as Lewis acid site in the reaction (Figure S5 and Table S4).[35,36]
For xSn-Ni@C600 catalysts (2.0 < x < 5.0), the Sn/Ni mole ratio was higher than
the nominal ones at 4:3 or 2:3 of Sn-Ni IMCs. Therefore, it could be speculated that not
all Sn formed the Sn-Ni IMC with Ni. In other words, some Sn-Ni intermetallic
particles might be formed non-uniformly or some SnOx might exist on the 3Sn-
8
This article is protected by copyright. All rights reserved.

10.1002/cssc.201900172
ChemSusChem
Ni@C600.[37] This result was consistent with the TEM images (Figure 3 and Figure S7).
The Ni3Sn4 IMC and SnOx particles with the average size 25.4 nm were clearly found
to be well dispersed in the carbon matrix. The lattice distances of 0.293 nm could be
attributed to the (111) plane of Ni3Sn4, which was in good agreement with the XRD
result.[27,38] SnOx particles were also observed, and the lattice distance of 0.334 nm was
attributed to the (110) plane of SnO2. The image of Figure 3B further shows that
crystalline Sn-Ni IMC was directly attached on SnO2 particles for forming the distinct
boundaries.
Figure 3. HRTEM images of 3Sn-Ni@C600 catalyst.
Table 2．Cellulose conversion and product distribution using different substrates over xSn-Ni@C600
Conv.
(%)
Yield (C-mol%)
Carbon balance
(C-mol%)
Catalyst
Substrate
EG
Gly
La
1,2-PDO
HA
MG
Ace
HB
Sor
3Sn-Ni@C600
3Sn-Ni@C600
sorbitol
glucose
0
0
0.1
0.1
0.2
0.1
0
0
3.3
0.9
0.2
1.4
0
0
2.6
0.6
0.1
0.9
0
0
0.1
0.2
0.1
0.1
0
0
35.2
28.1
2.6
46.6
0
0
0.1
0
0
0.1
0
0
36.1
30.2
1.5
2.4
0
100
7.6
9.7
2.7
5.3
-
-
96.7
86.4
65.4
97.9
0
88.1
80.1
30.1
58.2
-
a
3Sn-Ni@C600
Sn@Ca
glucose
glucose
0.1
0.1
0
0.4
0.1
0
3Sn-Ni@C600
3Sn-Ni@C600
3Sn-Ni@C600
3Sn-Ni@C600
Ni@C
fructose
erythritol
glyceraldehyde
glycerol
95.4
0
0
10.6
100
0
3.4
0
1.2
0
74.4
0
3.4
0
0
0
-
97.5
-
0
0
0
-
sorbitol
0
0
0
0
0
0
0
0
100
18.6
26.4
0
-
Ni@C
glucose
100
100
0
1.2
0.1
0
8.4
20.6
0
18.9
0
10.3
8.7
0
0
0
0
0
70.9
55.8
-
Ni@C
fructose
0
0
0
0
Ni@C
erythritol
0
0
0
0
0
9
This article is protected by copyright. All rights reserved.

10.1002/cssc.201900172
ChemSusChem
Ni@C
Ni@C
glyceraldehyde
glycerol
100
0
0
0
12.9
100
8.1
0
76.6
0
0
0
0
0
0
0
0
0
0
0
97.6
-
Reaction conditions: substrate 0.2 g, catalyst 0.08 g, deionized water 20 mL, reaction temperature
240 oC, reaction time 1 h and initial H2 pressure 4.0 MPa. EG, ethylene glycol; Gly, glycerol; La,
lactic acid; 1,2-PDO, 1,2-propanediol; HA, hydroxyacetone; MG, methylglyoxal; Ace, acetoin; HB,
1-hydroxy-2-butanone; Sor, sorbitol.
a: reaction time 10 min; including other products: fructose (8.9%), ethanol (1.3%), 1,2-butanediol
(0.7%), 2,3- butanediol (0.9%), and several unknown products.
In order to understand the reaction pathway for the conversion of cellulose to keto-
alcohols, reactions of the possible intermediates, such as sorbitol, glucose, fructose, and
other possible intermediates were performed using different catalysts under the
comparable reaction conditions to cellulose. Sorbitol was found to be inactive with a
negligible conversion after 1 h (Entry 1, Table 2) over the 3Sn-Ni@C600 catalyst. In the
presence of Sn-Ni intermetallic phase, glucose was converted to the keto-alcohols for
obtaining the total yield of 71.5 % (at 96.7 % conversion, Entry 2, Table 2), which
showed the similar product distribution to those obtained in cellulose reaction (Entry 8,
Table 1). The main products are HA (35.2 %), and HB (36.1 %). While using fructose
as a substrate, HA was formed dominantly in the yield of 46.6 % while trace amount of
HB (2.4 %) was obtained. These results indicated that HA was produced mainly by
direct retro-aldol reaction of fructose over these Sn-Ni catalysts. The reaction results of
erythritol and glycerol, the possible intermediates, suggested that they were inactive
(Entry 6 and 8, Table 2) for HA and HB production. For glyceraldehyde, the main
product was HA (74.4 %, Entry 7, Table 2). However, when using Sn@C physically
10
This article is protected by copyright. All rights reserved.

10.1002/cssc.201900172
ChemSusChem
mixed with Ni@C as the catalyst, the keto-alcohols were negligible (Table 1). The
results showed that Sn-Ni IMC catalyzed the hydrodeoxygenation of C-O in
intermediate and also has weak C-C bond cleavage capability for the formation of the
keto-alcohols. Therefore, glucose and fructose produced from glucose isomerization
are the key two intermediates for highly selective production of HA and HB,
respectively, through retro-aldol condensation followed by selective cleavage the C-O
bonds over the catalysts (Figure S11).
To identify the functions of different active sites such as SnOx, Ni metal, NiO and
Sn-Ni IMC on the catalysts. A series reactions using cellulose, sorbitol, glucose,
fructose, and other possible intermediates as the feedstock were performed over the
catalysts including Ni@C, Ni-NiO@C, NiO, Sn@C. The results are listed in Table 1,
Table 2 and Table S2. In the presence of Ni@C, despite glucose was completely
converted at 1 h, giving La and sorbitol as main products together with the minor C2-
C3 diols (Entry 10, Table 2) [39]. The C3-C4 keto-alcohols was negligible in this case.
For the Sn@C catalyst (Entry 4, Table 2), about 65 % of glucose was converted after
10 min of reaction. HA, HB and fructose were observed with yields of 2.6 %. 1.5% and
8.9 %, respectively. Correlating to the characterization results of Sn@C catalyst
discussed above (XRD, Figure 2A), the Sn metal and SnOx on the catalyst provide the
major active sites for the glucose isomerization and the retro-aldol condensation of
glucose in view of the C3-C4 keto-alcohols formation [27,40,41]. For Ni-NiO@C and
Sn@C catalysts, the unreduced Ni species and SnOx act as acid sites and catalyze the
degradation of cellulose (Entry 1, Table S2 and Entry 4, Table 2)[42]. Furthermore, NiO
11
This article is protected by copyright. All rights reserved.

10.1002/cssc.201900172
ChemSusChem
could also inhibit the C-O bond cleavage over Ni particle (Entry 10,11 and 13, Table 2,
and Table S2). Based on the results of reaction and characterizations on Sn-Ni catalysts,
it is clear that there is a synergetic effect between Sn-Ni IMC and Sn/Ni oxides in the
reactions involving cellulose degradation. SnOx and unreduced Ni catalyzed the
glucose isomerization to fructose. The unreacted glucose and produced fructose were
further degraded via the retro-aldol condensation on the sites of Sn-Ni IMC and metal
oxide on the catalysts. Finally, the C3 and C4 intermediates of glyceraldehyde, 1,3-
dihydroxyacetone and erythrose were hydrodeoxygenated over Sn-Ni IMC sites to
produce C3-C4 keto-alcohol.
Considered the results discussed above, it suggested that the formation of the C3
(HA) and C4 (HB) products in the cellulose reaction involved the isomerization of
glucose to fructose prior to its retro-aldol condensation on the SnOx species, which was
consistent with the observed conversion of glucose to fructose, as shown in Table 2, on
the Sn@C and 3Sn-Ni@C600 at 240 oC with 10 min reaction. Davis et al.[41] found that
Sn incorporated in the framework of zeolite Beta, performed the isomerization reaction
following an intramolecular hydride shift mechanism between the carbonyl containing
C-1 and hydroxyl bearing C-2 of glucose by a way of a five member complex. Therefore,
a similar mechanism of SnOx is proposed in Scheme 1 and Scheme 2, considering the
isomerization on Lewis acid sites (Figure S5 and Table S3) provided by SnOx. As it is
widely known that more vacant orbitals in d band would lead to a stronger ability of
adsorbing and bonding. With d band holes, it can form adsorption bonds with reaction
molecules to produce a reactive intermediate[43]. Therefore, the less vacant d bands, the
12
This article is protected by copyright. All rights reserved.

10.1002/cssc.201900172
ChemSusChem
weaker the hydrogenation capacity of the metal. Ni[3d9.44s0.6] has 0.6 holes, while d
band in Sn[4d105s25p2] is full[43,44]. For Sn-Ni IMC, electrons in d band of Sn can be
occupied the holes of d band in Ni metal, and then weaken the hydrogenation capacity
of Ni metal. Therefore, the C=O group can be maintained over Sn-Ni IMC with inferior
hydrogenation capacity.
Scheme 1. Isomerization mechanism of glucose to fructose on 3Sn-Ni@C600
.
13
This article is protected by copyright. All rights reserved.

10.1002/cssc.201900172
ChemSusChem
Scheme 2. Reaction scheme for conversion sugars to C3-C4 keto-alcohols over 3Sn-Ni@C600
.
The recyclability of these Ni promoted catalysts was also examined. As shown in
Figure S8. The results showed that the 3Sn-Ni@C600 catalyst exhibited good reusability,
as indicated by only slight losses in cellulose conversion and keto-alcohols yield. ICP
analysis of the reaction solution suggested that leaching of either Ni or Sn is negligible
(Entry 7, Table S1). In addition, the good stability of the Sn-Ni IMC could be attributed
to the strong interactions between the Ni and Sn species. A slight losses in cellulose
conversion and keto-alcohols yield were observed after six cycles either on 3Sn-
Ni@C600 (representative results in Figure S8). The characterizations of pristine and
spent catalysts were also studied (Figure S9). After 6 runs, the particle size of the Sn-
Ni intermetallic increased. Considering the residue of cellulose (Figure S9C), the TG-
DTA suggested that no obvious carbon deposits on the catalyst (Figure S9D and E).
14
This article is protected by copyright. All rights reserved.

10.1002/cssc.201900172
ChemSusChem
These results demonstrate that the Sn-Ni catalysts were stable and reusable under the
reaction conditions in this work.
In conclusion, we have found that Sn-Ni IMC-SnOx catalysts (xSn-Ni@CT) are
efficient in the acceleration of the hydrolysis of cellulose to keto-alcohols especially
HA (33.2 %) and HB (35.6 %) in water. The Sn-Ni IMC and SnOx synergistic catalyze
cellulose into HA and HB in an aqueous phase. Both the SnOx and Sn-Ni IMC sites are
active sites for cellulose conversion to keto-alcohols. The formation of keto-alcohols
are strongly depend on the synergy of Sn-Ni IMC and SnOx active sites on the surface
of the catalysts. Sn-Ni IMC catalyzes the hydrodeoxygenation of C-O in intermediates
and has weak C-C bond cleavage capability. The SnOx can promote the reactions
involving cellulose degradation, intermediates isomerization. Glucose and fructose are
the key intermediate for highly production of HB and HA, respectively. The xSn-Ni@C
catalysts reveal to be an efficient, reusable, and cheap catalyst make of abundant
biomass for the fine chemicals in water, and the process environmentally benign and
sustainable.
Acknowledgement
This work is financially supported by the National Natural Science Foundation of China
(51576199, 51536009), the DNL Cooperation Fund, CAS (DNL180302), the Natural
Science Foundation of Guangdong Province (2017A030308010) and the Local
Innovative and Research Teams Project of Guangdong Pearl River Talents Program
(2017BT01N092).
Reference
15
This article is protected by copyright. All rights reserved.

10.1002/cssc.201900172
ChemSusChem
[1] D. M. Alonso, J. Q. Bond, J. A. Dumesic, Green Chem 2010, 12, 1493-1513.
[2] M. J. Climent, A. Corma, S. Iborra, Cheminform 2014, 16, 516-547.
[3] J. Carlos Serrano-Ruiz, R. Luque, A. Sepulveda-Escribano, Chem. Soc. Rev. 2011, 40, 5266-
5281.
[4] C. Li, M. Zheng, A. Wang, T. Zhang, Energ. Environ. Sci. 2012, 5, 6383-6390.
[5] J. Pang, M. Zheng, R. Sun, L. Song, A. Wang, X. Wang, T. Zhang, Bioresour Technol 2015,
175, 424-429.
[6] H. Tang, N. Li, F. Chen, G. Li, A. Wang, Y. Cong, X. Wang, T. Zhang, Green Chem 2017, 19,
176-182.
[7] G. W. Huber, S. Iborra, A. Corma, Chem. Rev. 2006, 106, 4044-4098.
[8] P. A. Lazaridis, S. A.Karakoulia, C. Teodorescu, N. Apostol, D. Macovei, A. Panteli, A.
Delimitis, S. M. Coman, V. I. Parvulescu, K. S. Triantafyllidis, Appl Catal B-Environ 2017, 214, 1-
14.
[9] J. Xi, Q. Xia, Y. Shao, D. Ding, P. Yang, X. Liu, G. Lu, Y. Wang, Appl Catal B-Environ 2016,
181, 699-706.
[10] J. Keskivali, S. Rautiainen, M. Heikkila, T. T. T. Myllymaki, J.-P. Karjalainen, K. Lagerblom,
M. Kemell, M. Vehkamaki, K. Meinander, T. Repo, Green Chem 2017, 19, 4563-4570.
[11] B. Zhang, X. Li, Q. Wu, C. Zhang, Y. Yu, M. Lan, X. Wei, Z. Ying, T. Liu, G. Liang, F. Zhao,
Green Chem 2016, 18, 3315-3323.
[12] N. Ji, T. Zhang, M. Y. Zheng, A. Q. Wang, H. Wang, X. D. Wang, J. G. G. Chen, Angew Chem
Int Edit 2008, 47, 8510-8513.
[13] L. Gao, Y. Bao, S. Gan, Z. Sun, Z. Song, D. Han, F. Li, L. Niu, ChemSusChem 2018,
10.1002/cssc.201800695.
[14] M. Ventura, D. Williamson, F. Lobefaro, M. D. Jones, D. Mattia, F. Nocito, M. Aresta, A.
Dibenedetto, ChemSusChem 2018, 11, 1073-1081.
[15] M. Ventura, F. Nocito, E. D. Giglio,S. Cometa, A. Altomare, A. Dibenedetto, Green Chem
2018, 20, 3921-3926.
[16] L. I. I. Granone, F. Sieland, N. Zheng, R. Dillert, D. W. Bahnemann, Green Chem 2018, 20,
1169-1192.
[17] H. Kobayashi, T. Komanoya, S. K. Guha, K. Hara, A. Fukuoka, Appl Catal A Gen 2011, 409,
13-20.
[18] J. Yi, S. Liu, M. M. Abu-Omar, ChemSusChem 2012, 5, 1401-1404.
[19] C. Luo, S. Wang, H. Liu, Angew Chem Int Edit 2010, 119, 7780-7783.
[20] J. Sun, H. Liu, Chinese J Catal 2011, 13, 135-142.
[21] B. Gumina, F. Mauriello, R. Pietropaolo, S. Galvagno, C. Espro, Mol Catal 2018, 446, 152-
160.
[22] S. Sato, D. Sakai, F. Sato, Y. Yamada, Chem Let 2012, 41, 965-966.
[23] T. Deng, H. Liu, J Mol Catal A-Chem 2014, 388, 66-73.
[24] Y. Wang, M. Zhuang, S. Chen, W. Hu, J. Sun, J. Lin, S. Wan, Y. Wang, ACS Catal 2017, 7,
6038-6047.
[25] A. K. Kinage, P. P. Upare, P. Kasinathan, Y. K. Hwang, J.S.Chang, CatalCommun 2010, 11,
620-623.
[26] V. Zacharopoulou, E. S. Vasiliadou, A. A. Lemonidou, ChemSusChem 2018, 11, 264-275.
16
This article is protected by copyright. All rights reserved.

10.1002/cssc.201900172
ChemSusChem
[27] R. Sun, M. Zheng, J. Pang, X. Liu, J. Wang, X. Pan, A. Wang, X. Wang, T. Zhang, ACS Catal
2016, 6, 191-201.
[28] M. Zheng, J. Pang, R. Sun, A. Wang, T. Zhang, ACS Catal 2017, 7, 1939-1954.
[29] W. Liu, Y. Chen, H. Qi, L. Zhang, W. Yan, X. Liu, X. Yang, S. Miao, W. Wang, C. Liu, A.
Wang, J. Li, T. Zhang, Angew Chem Int Edit 2018, 57, 7071-7075.
[30] C. W. Chin, M. A. Dasari, G. J. Suppes, W. R. Sutterlin, AIChE J 2006, 52, 3543-3548.
[31] L. G. Lee, G. M. Whitesides, J Org Chem1986, 51, 25-36.
[32] A. Demirbas, Fuel Process Technol 2007, 88, 591-597.
[33] H. Wang, C. Zhang, Q. Liu, C. Zhu, L. Chen, C. Wang, L. Ma, ACS Energ Fuel 2018, 32,
11529-11537.
[34] J. M. Jacob, P. G. Corradini, E. Antolini, N. A. Santos, J. Perez, Appl Catal B-Environ 2015,
165, 176-184.
[35] X. Yan, J. Zheng, L. L. Zheng, G. Lin, H. Lin, G. Chen, B. Du, F. Zhang, Mater Res Bull 2018,
103, 150-157.
[36] J. W. Shabaker, D. A. Simonetti, R. D. Cortright, J. A. Dumesic, J Catal 2005, 231, 67-76.
[37] J. Liu, Y. Wen, P. A. van Aken, J. Maier, Y. Yu, NANO Let 2014, 14, 6387-6392.
[38] J. W. Jang, P. G. Kim, K. N. Tu, D. R. Frear, P. Thompson, J Appl Phys 1999, 85, 8456-8463.
[39] M. F. N. D'Angelo, V. Ordomsky, J. van der Schaaf, J. C. Schouten, T. A. Nijhuis,
ChemSusChem 2014, 7, 627-630.
[40] J. Pang, M. Zheng, X. Li, Y. Jiang, Y. Zhao, A. Wang, J. Wang, X. Wang, T. Zhang, Appl Catal
B-Environ 2018, 239, 300-308.
[41] R Y. Roman-Leshkov, M. Moliner, J. A. Labinger, M. E. Davis, Angew Chem Int Edit 2010,
49, 8954-8957.
[42] S.Van de Vyver, J. Geboers, W. Schutyser, M. Dusselier, P. Eloy, E. Dornez, J. W. Seo, C. M.
Courtin, E. M. Gaigneaux, P. A. Jacobs, B. F. Sels, ChemSusChem 2012, 5, 1549-1558.
[43] X. Shi, H. Huang, B. Jacobson, Y.-C. Chang, Q.-Z. Yin, C. Y. Ng, Astrophys J 2012, 747, 1-8.
[44] A. Onda, T. Komatsu, T. Yashima, Phys Chem Cheml Phys 2000, 2, 2999-3005.
17
This article is protected by copyright. All rights reserved.