Bimetallic Co–Ni/TiO2 catalysts for selective hydrogenation of cinnamaldehyde

Article abstract of DOI:10.1007/s11164-018-3517-7

Bimetallic Co–Ni catalysts in the composition range Co_(1?x)Ni_x with x = 0.0, 0.2, 0.3, 0.4, 0.5, 0.6, 0.8 and 1.0, with total metal loading of 15% w/w and supported on TiO₂-P25, have been prepared by chemical reduction of the metal acetates by glucose in aqueous alkaline medium and characterized by XRD, TEM, TPR, XPS and H₂-TPD techniques. Selective hydrogenation of cinnamaldhyde (CAL) to hydrocinnamaldehyde (HCAL), cinnamyl alcohol (COL) and hydrocinnamyl alcohol (HCOL) has been investigated at 20?bar pressure, in the temperature range 120–140?°C. Co/Ni crystallite sizes in the range 6.0 ± 1?nm are observed by TEM. TPR and XPS results indicate the formation of nanoscale Co–Ni alloys, which tend to weaken M–H bond strength, as revealed by H₂-TPD measurements. Ni/TiO₂ displays very high conversion of CAL (86.9%) with high selectivity (78.7%) towards HCAL formation at 140?°C. Co/TiO₂, on the other hand, exhibits relatively lower CAL conversion (55%) and higher selectivity (61.3%) for COL formation at the same temperature. However, bi-metallic Co–Ni catalysts in the composition range x = 0.3–0.6 display very high conversion (> 98%) due to alloy formation and weakening of M–H bonds. Bimetallic Co_0.7Ni_0.3 catalyst displays high conversion of CAL (98.1%) and high selectivity (82.9%) towards HCOL. Overall CAL hydrogenation activity at 140?°C, when expressed as TOF, displays a maximum value at the composition Co_0.5Ni_0.5. Activity and selectivity patterns have been rationalized based on the reaction pathways observed on the catalysts and the influence of Co–Ni alloy formation and M–H bond strength. Thus, a synergetic effect, originating from an appropriate composition of base metal catalysts and reaction conditions, could result in hydrogenation activity comparable with noble metal based catalysts.

Full text of DOI:10.1007/s11164-018-3517-7

Research on Chemical Intermediates
https://doi.org/10.1007/s11164-018-3517-7(0123456789(). -, vo Vl
)
(
0
1
2
3
4
5
6
7
8
9
(
)
.
-
,
v
o
l
V
)
Bimetallic Co–Ni/TiO catalysts for selective hydrogenation
2
of cinnamaldehyde
Saranya Ashokkumar, et al. [full author details at the end of the article]
Received: 7 March 2018 / Accepted: 18 June 2018
Ó Springer Nature B.V. 2018
Abstract
Bimetallic Co–Ni catalysts in the composition range Co(1-x)Ni with x = 0.0, 0.2,
x
0
TiO -P25, have been prepared by chemical reduction of the metal acetates by
.3, 0.4, 0.5, 0.6, 0.8 and 1.0, with total metal loading of 15% w/w and supported on
2
glucose in aqueous alkaline medium and characterized by XRD, TEM, TPR, XPS
and H -TPD techniques. Selective hydrogenation of cinnamaldhyde (CAL) to
2
hydrocinnamaldehyde (HCAL), cinnamyl alcohol (COL) and hydrocinnamyl alco-
hol (HCOL) has been investigated at 20 bar pressure, in the temperature range
120–140 °C. Co/Ni crystallite sizes in the range 6.0 ± 1 nm are observed by TEM.
TPR and XPS results indicate the formation of nanoscale Co–Ni alloys, which tend
to weaken M–H bond strength, as revealed by H -TPD measurements. Ni/TiO2
2
displays very high conversion of CAL (86.9%) with high selectivity (78.7%)
towards HCAL formation at 140 °C. Co/TiO , on the other hand, exhibits relatively
2
lower CAL conversion (55%) and higher selectivity (61.3%) for COL formation at
the same temperature. However, bi-metallic Co–Ni catalysts in the composition
range x = 0.3–0.6 display very high conversion ([ 98%) due to alloy formation and
weakening of M–H bonds. Bimetallic Co0.7Ni0.3 catalyst displays high conversion
of CAL (98.1%) and high selectivity (82.9%) towards HCOL. Overall CAL
hydrogenation activity at 140 °C, when expressed as TOF, displays a maximum
value at the composition Co Ni . Activity and selectivity patterns have been
0
.5 0.5
rationalized based on the reaction pathways observed on the catalysts and the
influence of Co–Ni alloy formation and M–H bond strength. Thus, a synergetic
effect, originating from an appropriate composition of base metal catalysts and
reaction conditions, could result in hydrogenation activity comparable with noble
metal based catalysts.
Keywords Co–Ni bimetallic catalysts Á Titania Á Co–Ni alloys Á Cinnamaldehyde
hydrogenation Á Metal-hydrogen bond strength
Electronic supplementary material The online version of this article (https://doi.org/10.1007/s11164-
018-3517-7) contains supplementary material, which is available to authorized users.
123

S. Ashokkumar et al.
Introduction
Selective hydrogenation of cinnamaldehyde (CAL) to cinnamyl alcohol (COL) and
hydro-cinnamaldehyde (HCAL) is an important reaction from industrial as well as
academic points of view [1,2]. Bimetallic catalysts, in general, display high activity
and selectivity for this reaction [3]. Noble metal (Pt and Ru) based bimetallic
catalysts with promoters like Co, Sn exhibit high activity and selectivity for COL
[
4–7]. Non-noble metal based bimetallic catalysts containing Co, Ni and Cu have
also been studied extensively [8–12]. Considering the higher cost of noble metal
based catalysts, attempts are being made to develop bimetallic catalysts based on Ni
and Co [13–15]. Earlier work in this direction by Reddy et al. [13] on 10% Co-10%
Ni supported on SiO for vapor phase hydrogenation of CAL resulted in 63%
2
conversion, with 59% selectivity to HCAL and 34% to HCOL and no COL
formation. Hui et al. [14] in their studies on unsupported tri-component alloys with
varying Co, Ni and B contents as catalysts, have reported maximum CAL
conversion of 64.6% for Ni38.1Co26.3B35.6 catalyst with nearly total selectivity to
HCAL and again no COL. Malobela et al. [15] have observed that bimetallic Co–Ni
catalysts of specific composition (5 w/w % each of Ni and Co) on MWCNT support
displays 63% CAL conversion with 62% selectivity towards COL.
While titania supported Ni catalyst [16] displays high activity for CAL
conversion with high selectivity for HCAL, the corresponding cobalt catalyst
displays lower conversion and good selectivity for COL [17]. Bimetallic catalysts
with optimum Co–Ni composition could lead to higher activity and selectivity, and
this aspect has not been explored so far. A systematic study involving the whole
range of Co–Ni composition is needed to arrive at the optimum one. Accordingly,
mono and bi-metallic Co–Ni catalysts in the composition range Co(1-x)Ni with
x
x = 0.0, 0.2, 0.3, 0.4, 0.5, 0.6, 0.8 and 1.0, supported on TiO -P25, have been
2
explored to arrive at the compositions that would give high CAL conversions,
comparable to that of noble metal catalysts and good selectivity for COL.
Experimental
Materials
TiO -P-25 (Evonik), Ni(CH COO) .4H O (CDH), Co(CH COO) .4 H O (CDH), D-
2
2
3
2
2
3
2
glucose (Merck), methanol, liquor ammonia (Qualigens) and Cinnmaldehyde
Aldrich) were used as such.
(
Preparation of mono metallic catalysts
The 0.743 g cobalt acetate or 0.746 g of nickel acetate and 40 ml of aqueous D-
glucose solution (0.15 M) were mixed and stirred for 30 min at room temperature.
Then 10 ml of liquor ammonia was added drop-wise to the mixture. On refluxing
2
?
the mixture for 5 h at 80 °C, its color turned to black, indicating reduction of Co
Ni ions to Co /Ni . As known, D-glucose acts as reducing as well as capping agent
/
2
?
0
0
1
23

Bimetallic Co–Ni/TiO
2
catalysts for selective…
2?
18]. Reduction of Co and Ni ions to their respective metallic state was
2
?
[
confirmed by UV–Vis spectroscopic study (Fig. S1), which showed the absence of
absorption maxima due to the respective metal ions [19, 20]. Ni/Co nanoparticles,
stabilized in alkaline medium, were then anchored on to the support by adding 1 g
of TiO (P25) and stirring continued for 2 h. The mixture was cooled to ambient
2
temperature, centrifuged, washed with anhydrous ethanol and dried at 60 °C for
2
prior to all characterization and hydrogenation experiments.
4 h. The catalysts were then pre-reduced in hydrogen gas flow at 300 °C for 3 h,
Preparation of Co–Ni bimetallic catalysts
Co1-xNi bimetallic catalysts with different atomic fractions with x = 0.0, 0.2, 0.3,
x
0
.4, 0.5, 0.6, 0.8 and 1.0 were prepared using the same green chemistry route
followed for monometallic catalysts. In all the bimetallic catalysts total metal
loading (Co ? Ni) was maintained at 15% w/w. Appropriate quantities of Co and
Ni acetates, as defined by the value of x, were mixed in aqueous glucose solution
and subjected to simultaneous chemical reduction, which was confirmed by UV–
visible spectroscopy (Fig. S1). After washing the final catalysts with ethanol and
drying at 60 °C, the catalysts were pre-reduced in hydrogen gas flow at 300 °C for
3
h. All characterization and hydrogenation experiments were carried out with pre-
reduced catalysts.
Characterization of catalysts
Co and Ni contents in the catalysts were estimated by the ICP-OES technique, using
a Perkin Elmer Model Optima 5300 DV unit, after extraction of the metals with
aqua regia.
Powder X-Ray diffraction patterns for the catalysts were recorded by using a
Rigaku Miniflex II X-ray diffractmeter with Cu-Ka (k = 0.15418 nm) radiation in
the 2h range of 10°–80° and at a scan rate of 3°/min.
Temperature programmed reduction (H -TPR) and temperature programmed
2
desorption (H -TPD) runs were carried out using a Micromeritics Autochem II 2920
2
chemisorption analyser. For TPR, the catalysts were pre-calcined in air at 300 °C
for 3 h. Then 50 mg of calcined catalyst was pre-treated at 300 °C in high purity Ar
gas (25 cc/min) for 1 h and then cooled to room temperature in an Ar flow. The gas
was changed to 10% H2 in Ar (25 cc/min) at room temperature. After the
stabilization of the baseline, TPR was started from RT to 700 °C with a heating rate
1
0 °C/min.
For H TPD measurements, 50 mg of catalyst was reduced in hydrogen flow
2
(25 cc/min) at 300 °C for 4 h and cooled to ambient temperature. Ar flow (30 cc/
min) was then introduced, and the catalyst was purged for 30 min. After the
stabilization of the baseline, TPD of H was recorded up to 700 °C at a temperature
2
ramp of 10 °C/min.
Transmission electron micrographs were recorded using a JEOL 3010 model
microscope. Few milligrams of the reduced samples (1–2 mg) were dispersed in a
few mL (1–2 mL) of ethanol by ultra-sonication for 15 min and a drop of the
123

S. Ashokkumar et al.
dispersion was placed on a carbon coated copper grid and allowed to dry in air at
room temperature.
Based on the mean crystallite size measured from TEM data, Co/Ni metal
dispersion were calculated using the formula [21]
6
00 Â M
Co=Ni
Dispersion ð%Þ ¼ q
;
ð1Þ
 dnm  a
 Na
Co=Ni
Co=Ni
where M is the molecular weight of the metal (Co/Ni), a is the atomic surface area
-
20
2
of Co/Ni (6.62/6.49 9 10
m /atom), q
is the density of the metals, N is
Co/Ni a
Avogadro’s number, and dnm is the average crystallite diameter (in nm) estimated
from TEM data.
XPS spectra of the catalysts were recorded using an Omicron ESCA Probe
spectrometer with Mg K X-rays (ht = 1253.6 eV). The samples were spotted as
a
drop cast films on a sample stub. The base pressure of the analysis chamber during
-
10
the scan was 2 9 10
were 20 and 100 eV, respectively. The spectra were recorded with a step width of
.05 eV. Data were processed with the Casa XPS program (Casa Software Ltd.UK),
millibar. The pass energies for individual and survey scans
0
and calibrated with reference to the adventitious carbon peak (284.9 eV) of the
sample.
Hydrogenation of cinnamaldehyde
Catalysts were evaluated for liquid phase hydrogenation of cinnamaldehyde (CAL)
in a 100 mL Parr reactor (Model-4848). Initially, the reactor was filled 15 g of iso-
propanol and 1.65 g of water as solvent. Use of this mixture as solvent [7] resulted
in very little formation of acetals during hydrogenation. Also, 1.2 g of CAL and
1
three times with H gas and then pressurized to 20 bar. Normally the reaction was
50 mg of catalyst were then added to the solvent mixture. The reactor was purged
2
carried out for 1 h in the temperature range 120–140 °C. A 1 h time is noted from
the moment the reactor reaches the specified temperature. After each reaction, the
reactor is allowed to cool naturally to room temperature, catalyst and reaction
products were separated by filtration, and the product stream was analyzed in a
Perkin Elmer Clarus-500 GC equipped with a ZB-1 capillary column and FID.
Calculation of conversion The reaction mixture containing (15 g iso-
propanol ? 1.65 g water and 1.2 g CAL) was used as standard, representing the
initial concentration of CAL. Analysis of this mixture was carried out before every
reaction run. Then 0.4 lL standard is injected into GC and its area is noted. 0.4 lL
of reaction products is then analyzed, and the conversion is calculated as follows:
AS À ARP
%
Conversion ¼
 100:
ð2Þ
AS
A —Area of CAL in standard mixture, A —Area of CAL in reaction product
s RP
1
23

Bimetallic Co–Ni/TiO catalysts for selective…
2
ASP
%
Selectivity ¼ P
 100
ð3Þ
Products
P
A —Area of product whose selectivity to be calculated, Products—sum of areas
SP
of all products
Rates of hydrogenation (r) and Co/Ni metal dispersion (calculated using formula-
) values were utilized to compute Turn over Frequency (TOF) using the formula-2
21]
1
[
r
TOF ¼
;
ð4Þ
ntot  Dispersionð% Þ
where r is the rate of hydrogenation (moles converted per second), ntot is the total
number of Co/Ni moles in the reactor.
Results and discussion
Chemical composition
Actual chemical composition of the catalysts, in terms Co and Ni contents, as
analyzed by ICP-AES technique, are given in Table S1. Actual values are close to
the expected values, thus confirming the chemical composition of the catalysts.
X-ray diffraction
XRD patterns for monometallic Co and Ni catalysts and six different bimetallic
catalyst formulations are given in Fig. 1. According to JCPDS data, metallic Ni
exhibits major d-lines, at 2h = 44.505°, 51.84° and 76.37° (04-0850) and metallic
Co at 2h = 44.227°, 51.53° and 75.865° (89-7093). X-ray diffractogram for Ni/TiO
2
displayed characteristic d-lines due to support TiO -P-25 and in addition, only one
d-line at 44.38° due to metallic Ni. The other two lines at 2h = 51.84° and 76.37°
2
are not observed due to the weak intensity. Likewise, in the case of Co/TiO ,
2
besides the d-lines due to TiO -P-25 phase, an extra d-line at 2h = 44.18° attributed
2
to metallic Co is noticed. Additional d-lines at 2h = 51.53° and 75.865° are not
observed. Both monometallic catalysts prepared in this work exhibit fcc crystal
structure.
All the six bimetallic Co–Ni catalysts exhibit d-lines in the range
2
h = 44.1–44.3°, which is within the 2h values of 44.18° and 44.38° observed in
this work for Co and Ni metals, respectively. Such shifts in the d-lines in the case of
titania supported Ni–Co bimetallic catalysts was observed by Takanabe et al. [22]
and attributed to alloy formation. The 2h values in the range 44.1–44.3° observed
for bimetallic catalysts are also close to the major d-line at 2h = 44.48°, reported for
Ni–Co alloys (071-074-5694). Since the crystallite sizes are small, the d-lines are
broader, but the formation of alloys is imminent. Malobela et al. [15] in their studies
on 5% Ni-5% Co catalysts on carbonaceous supports, (graphite, MWCNT and
activated carbon) have also observed the formation of Ni Co alloy.
5
0
50
123

S. Ashokkumar et al.
Fig. 1 X-ray diffractograms for a Co/TiO
2
b Co0.8 Ni 0.2/TiO
h Ni
2
c Co0.7 Ni 0.3/TiO
2 2
d Co0.6 Ni0.4/TiO e
Co0.5 Ni 0.5/TiO f Co0.4 Ni 0.6/TiO g Co0.2 Ni 0.8/TiO
2
2
2
1
/TiO
2
No d-lines due to Co with hcp structure are observed, though Co rich catalysts
may contain unalloyed Co phase [23]. According to Zhang et al. [24], hcp phase
may appear when Co crystallite size is larger than 17 nm. In the present work, since
crystallite size of Co is \ 17 nm, only fcc structure is favored. Crystallite size
values calculated using the Debye–Schererr equation, are in the range of 5–11 nm
(
Table 1). However, when prepared in unsupported form, Co nanoparticles with hcp
as well as fcc structure are observed, while in presence of titania support, only the
fcc phase is observed.
Transmission electron microscopy
Figure 2 presents transmission electron micrographs along with histograms for all
eight catalyst samples. Finely dispersed Co/Ni crystallites in the size range
6 ± 1.0 nm could be observed clearly on the titania support. While monometallic
Co or Ni catalyst displays crystallites of mean size 6.5 and 6.9 nm, respectively, in
1
23

Bimetallic Co–Ni/TiO catalysts for selective…
2
Table 1 Crystallite size and TOF values for Co–Ni catalysts
a
Cryst. size (nm)
b
Cryst. size (nm)
-1 -1 c
TOF 9 10 (h )
Catalysts
Co
1
8.0
5.6
6.5
5.5
5.1
7.0
6.4
5.0
6.1
6.9
1.08
1.44
1.54
1.58
1.80
1.44
1.51
1.69
Co0.8 Ni0.2
Co0.7 Ni0.3
Co0.6 Ni0.4
Co0.5 Ni0.5
Co0.4 Ni0.6
Co0.2 Ni0.8
5.5
8.1
11.0
7.4
6.2
Ni
1
10.0
a
b
c
XRD, TEM, at 140 °C
Fig. 2 TEM micrographs with histographs for a Co/TiO
2
b Co0.8 Ni 0.2/TiO
h Ni/TiO
2
c Co0.7 Ni0.3/TiO
i Co0.7 Ni0.3/TiO
2
d Co0.6
(Used)
Ni 0.4/TiO e Co0.5 Ni 0.5/TiO f Co0.4 Ni 0.6/TiO g Co0.2 Ni 0.8/TiO
2
2
2
2
2
2
123

S. Ashokkumar et al.
Co–Ni bi-metallic catalysts, a relatively smaller size range, 5.0 to 5.5 nm
crystallites are observed. Crystallite size values obtained by TEM and XRD are
comparable (Table 1).
Temperature programmed reduction
TPR profiles for mono- and bimetallic catalysts are presented in Fig. 3. For better
clarity, as recorded TPR profiles for individual catalysts are given in Fig. S2.
Monometallic Ni/TiO shows major reduction maximum at 387 °C, attributed to the
2
2
reduction of Ni ions that have interacted with the support (Profile-h). Reduction
?
2
?
of dispersed Ni ions is indicated at 230 and 300 °C, as shoulders to the major
2
?
peak. Continuation of reduction beyond 450 °C is due to the reduction of Ni ions
with strong interaction with the support. Reduction of dispersed Co ions (Profile-
2
?
a) is indicated in the form of shoulders at 250 °C and a peak at 333 °C, followed by
2
?
another peak at 380 °C due to Co ions that have interacted with the support.
2
?
Reduction maximum at 488 °C is due to Co with relatively stronger interaction
Fig. 3 H
Ni 0.5/TiO
2
-TPR Profiles for a Co/TiO
f Co0.4 Ni 0.6/TiO g Co0.2 Ni 0.8/TiO
2
b Co0.8 Ni 0.2/TiO
2
c Co0.7 Ni 0.3/TiO
2 2
d Co0.6 Ni 0.4/TiO e Co0.5
2
2
2
h Ni /TiO
1
2
1
23

Bimetallic Co–Ni/TiO catalysts for selective…
2
with the support. Addition of Ni to the extent of 0.2 atom fraction (Co0.8Ni0.2
profile-b), results in shifting of the reduction processes to lower temperatures.
,
2
?
2?
Dispersed Ni and Co ions get reduced at 230, 330 °C, and the reduction peaks
2
?
2?
at 387 and 380 °C, observed for mono metallic Ni and Co respectively, merge
at 341 °C, possibly due to simultaneous reduction and/or alloy formation. Besides,
2
?
the high temperature reduction peak 488 °C due to Co is shifted to 447 °C,
2
?
indicating weaker Co-support interaction. Since Ni
ions undergo reduction
= -0.277 V vs. SHE)
2
ENi(II)/Ni = -0.250 V vs. SHE) prior to Co ions (E
?
(
[
Co(II)/Co
2
25], Ni metal crystallites catalyze the reduction of Co with the nascent hydrogen
?
2
?
formed on the Ni surface, and; hence, the alloy formation and Co reduction at
lower temperature (488–447 °C) is facilitated. Further addition of Ni to Co0.7Ni0.3
(
profile-c) increases overall reducibility, with the Ni metal catalyzing the reduction
2?
of Co ions. Due to facile reduction, the peaks at 330, 378, 420 and 480 °C merge,
indicating simultaneous Ni and Co reduction and alloy formation. Bimetallic
2
?
2?
catalyst composition Co0.6Ni (profile-d) with higher Ni content presents a distinct
0.4
2
?
2?
peak at 311 °C due to Co and Ni reduction and alloy formation along with a
2
?
shoulder at 280 °C and another peak due to supported Co getting reduced at lower
temperature, i.e., 430 °C. With the composition Co0.5Ni0.5 (profile-e), besides the
peak at 390 °C due to Co–Ni alloy, and high temperature reduction peak at 505 °C,
is observed. With further increase in Ni content (Co Ni ) alloy content decreases
0
.4 0.6
2
?
and its reduction peak merges (profile-f) with that due to reduction of Ni at
10 °C, due to strong interaction with the support. TPR profile-g for bi-metallic
composition Co0.2Ni shows very little alloy formation and a major reduction peak
5
0.8
2
?
at 535 °C that is due to Ni strongly bound to the support.
To summarize, TPR profiles indicate maximum Co–Ni alloy formation and high
reducibility in the composition range Co0.7Ni0.3–Co0.5Ni0.5 and minimum/weaker
interaction with the support. Outside this composition range, interaction of Co/Ni
with the support is prominent. Co–Ni alloy formation and increase in reducibility
could influence the activation of the reactants and the nature of adsorbed hydrogen,
which, in turn affect the activity.
Alloy formation in Co–Ni bimetallic systems and its influence on the activity for
various reactions has been reported earlier for other reactions like, steam methane
reforming [26], dry reforming of methane with CO [22, 27], steam reforming of
2
alcohols [28] and acetic acid [29], hydrogenation of CO [30–32], methane partial
oxidation [33], hydrogenation of furfural [34] and hydrogenation of benzaldehyde
[
35]. In such reactions, Ni–Co alloys suppress coke formation and retard
deactivation, possibly by hydrogenation of coke precursors. Temperature pro-
grammed surface hydrogenation studies [36] support such a mechanism, which
essentially involves generation of active hydrogen for hydrogenation of coke
precursors.
XPS analysis
Figures 4 and 5 present XPS profiles for Co2p3/2 and Ni2p3/2 energy levels,
respectively, for selected Co–Ni bimetallic and monometallic Co and Ni catalysts.
Binding energy (BE) values observed for these catalysts are tabulated in Table 2.
123

S. Ashokkumar et al.
2 2 2 2
Fig. 4 XPS Profiles for Co2p level a Co/TiO b Co0.7 Ni0.3/TiO c Co0.6 Ni0.4/TiO d Co0.5 Ni0.5/TiO
2 2 2 2
Fig. 5 XPS Profiles for Ni 2p level a Ni/TiO b Co0.5 Ni0.5/TiO c Co0.6 Ni0.4/TiO d Co0.7 Ni0.3/TiO
Table 2 XPS binding energy
Catalysts
Co 2p3/2 (eV)
Ni 2p3/2 (eV)
data for Co–Ni catalysts
Co
778.4
778.8
778.9
778.7
–
–
Co0.7 Ni0.3
Co0.6 Ni0.4
Co0.5 Ni0.5
Ni
852.2
852.3
852.3
852.8
1
23

Bimetallic Co–Ni/TiO catalysts for selective…
2
BE value of 778.4 eV observed for Co2p3/2 energy level in monometallic Co/TiO2
Fig. 4a) indicates that Co is in metallic state and is in line with the values reported
(
in the literature [22, 37]. Small amounts of Co in oxidized state, observed at higher
BE, are due to surface oxidation by atmospheric oxygen during handling of the
catalysts. Introduction of Ni results in the shifting of Co2p3/2 BE to the higher side,
by 0.3-0.5 eV (Fig. 4b–d; Table 2). The shift in BE of metallic Co could be due to
charge transfer from Co to Ni in the formation of alloys. Similarly, the BE value of
8
52.8 eV observed for Ni2p3/2 energy level in monometallic Ni/TiO (Fig. 5a)
2
indicates that Ni is in a metallic state and is also in line with the values reported in
the literature [22, 38]. Again, small amounts of Ni in an oxidized state are observed,
due to surface oxidation during handling of the catalysts. Shifts in BE values for the
Ni2p3/2 level in bimetallic Co–Ni catalysts (Fig. 5b–d; Table 2) show a reverse
trend compared to that observed for Co 2p3/2 level, a decrease of 0.5–0.6 eV with
respect to that for monometallic Ni/TiO . Changes in BE values observed for both
2
Co 2p3/2 and Ni2p3/2 are indicative of nanoscale Co–Ni alloy formation. XPS
studies on bimetallic Co–Ni supported on anatase TiO by Takanabe et al. [22] did
2
not reveal changes in BE values, possibly due to the catalyst preparation method
2
used (incipient wetness). Simultaneous chemical reduction of Co
?
2?
and Ni ,
adopted in the present work, has resulted in more effective nanoscale interactions
between Co and Ni. No significant changes in the surface composition of Co and Ni
are observed. Corresponding to the bulk composition of Co0.5Ni0.5, the surface
atomic composition of Co and Ni, calculated from XPS data, turns out to be 44% Co
and 56% Ni. A slight surface enrichment of Ni could be due to the lower surface
energy of Ni vis-a- vis that for Co [38].
H2 temperature programmed desorption
H -TPD studies have been used to investigate the type and strength of active
2
catalytic centers for hydrogen chemisorption. Figure 6 displays H -TPD profiles
2
TiO supported Co and Ni monometallic and six bimetallic catalysts. Distinct
2
desorption peaks observed on all the catalysts indicate the presence of metal sites
that bind hydrogen with varying strength. Peaks in the temperature range
1
00–200 °C are attributed to desorption from dispersed Co/Ni metal sites and
those at higher temperatures, to desorption of spilled over hydrogen on support sites
39–41]. Monometallic Co displays desorption peak at 118 °C and Ni at 134 °C.
With the addition of a 0.2 atomic fraction of Ni to Co (Co Ni ), the low
[
0
.8 0.2
temperature H TPD peak shifts to 112 °C. With further addition of Ni (Co0.7Ni0.3
2
and Co Ni0.4) desorption maxima is shifted to lower temperatures, 110 and
0
.6
1
Co0.5 Ni and Co Ni desorption maxima are again at lower temperatures, 121
06 °C, respectively, indicating gradual weakening of M–H bond strength. With
0.5
0.4
0.6,
and 112 °C, respectively. TPD maximum for Co0.2Ni0.8 however is at 134 °C,
similar to that observed for monometallic Ni/TiO . Thus for specific Co–Ni
compositions, weakening of M–H bond strength, possibly due formation of Co–Ni
alloys, is observed. Weaker M–H bonds may favour facile hydrogenation.
2
123

S. Ashokkumar et al.
Fig. 6 H
Ni 0.5/TiO
2
-TPD Profiles for a Co/TiO
f Co0.4 Ni 0.6/TiO g Co0.2 Ni 0.8/TiO
2
b Co0.8 Ni 0.2/TiO
2
c Co0.7 Ni 0.3/TiO
2 2
d Co0.6 Ni 0.4/TiO e Co0.5
2
2
2
h Ni /TiO
1
2
Hydrogenation of cinnamaldehyde
Table S2 gives CAL conversion and selectivity for different products, HCAL, COL
and HCOL at different temperatures on mono- as well as bimetallic catalysts.
Formation of acetals as a side product has been minimized (2.8% max) by using iso-
propanol and water mixture as solvent. CAL conversion and selectivity for various
products at 140 °C on all catalysts are presented in Fig. 7.
Hydrogenation of cinnamaldehyde on monometallic Co and Ni catalysts
Activity and selectivity values for Co and Ni monometallic catalysts tabulated in
Table S2 are in line with the reported data [16, 17, 42, 43]. Ni/TiO displays very
2
high conversion of CAL (86.9%) with high selectivity (78.7%) towards HCAL
formation at 140 °C. Co/TiO , on the other hand, exhibits relatively lower CAL
conversion (55%) and higher selectivity (61.3%) for COL formation at the same
2
1
23

Bimetallic Co–Ni/TiO catalysts for selective…
2
Fig. 7 CAL conversion and product selectivity values for Co–Ni catalysts. (Reaction conditions: 15 g
isopropanol ? 1.65 g water and 1.2 g CAL, Temp. 140 °C, Pressure 20 bar)
temperature. Similar trends for both catalysts are observed at other temperatures as
well (Table S2).
Singh and Vannice [44] have observed volcano type correlation between %
d-character (d) of the metals and the initial TOF values for liquid phase
hydrogenation of citral, a typical unsaturated aldehyde like cinnamaldehyde, on a
series of silica supported Group VIII metal catalysts. The % d-character (d), as
described and calculated by Pauling [45], represents electronic structure of metals
and gives a measure of the contribution of d-electrons to the hybrid spd orbitals of
the metal. Correlations between d values for Group VII metals and hydrogenation
[
optimum % d character value displays maximum initial TOF, Co/SiO with lowest
46] and hydrogenolysis [47] activities have been reported. While Pd/SiO with
2
2
%
d-character value with respect to Co/SiO , displays higher initial TOF [44].
d-character value shows lowest activity. Ni/SiO , with relatively higher %
2
2
In the present work, the difference in the activity of Ni and Co catalysts with
-
1
TOF values of 1.69 and 1.08 h respectively, observed for hydrogenation on
cinnamaldehyde (Table 1) could be explained on the basis of their % d-character
values that influence the adsorption and activation of the unsaturated aldehydes.
According to Delbecq and Sautet [48], the selectivity for hydrogenation of C=O
versus C=C is related to the width of the d-band [48, 49]. Os, Ir, Ru and Pt with
higher d-band width exhibit higher selectivity towards C=O hydrogenation, while
Pd, Rh and Ni with lower width, preferentially activate C=C for hydrogenation.
Accordingly, Co with higher d-band width (4.0 eV) vis- a` -vis Ni (3.0 eV) exhibits
higher selectivity for C=O hydrogenation. However, it is pertinent to note that the
concept of d-band width could explain the selectivity for C=O versus C=C
hydrogenation in the case of 3d, 4d and 5d metals, but with some exceptions. Pd
(4.1 eV) and Rh (4.4 eV) with d-band width closer to that for Co (4.0), display
higher selectivity for C=C hydrogenation. Hence, d-band width alone could not be
123

S. Ashokkumar et al.
the factor responsible for higher selectivity for C=O hydrogenation observed with
Co.
Adsorption of CAL through C=O could be the primary route [42] for COL
formation on Co catalysts. Of the five possible modes of adsorption of a, b-
unsaturated aldehydes like CAL, proposed by Delbecq and Sautet [48], the three
possible modes for C=O adsorption are, namely, on-top g , di-r g and p
g
CO 2
leading to the formation of COL. In the absence of experimental data on the mode
1
2
of adsorption of CAL on the Co surface, on-top g could be the likely mode, based
1
on the in situ IR spectral data for adsorption of acetone on the Co surface [50]
though adsorption in pCOg2 mode could not be ruled out. Adsorption of CAL on the
Ni surface could be through pC=C– g2 or rC=C– g2, resulting in the formation of
HCLA. Since hydrogenation of C=C is kinetically favored vis- a` -vis that of C=O,
conversion of CAL on Co is lower than that on Ni. This aspect is also reflected in
experiments when neat HCAL or COL is used as feeds on Co (Table 3). While
3
3.8% of HCAL undergoes conversion on Co in 1 h, only 14.4% COL is converted
during that period. The corresponding behavior is just the reverse on Ni, with 57.9%
COL conversion and only 17.5% conversion of HCAL. It is clear that in the absence
of conjugated C=C and C=O bonds, hydrogenation of C=C is preferred on Co while
hydrogenation of C=O is facile on Ni. Similar trends are observed when mixtures of
HCAL and COL and CAL and COL are used as feed (Table 3). Thus, competitive
adsorption of products (HCAL and COL) on Co and Ni [42] also influences the
reaction pathways and hence the selectivity.
Hydrogenation of cinnamaldehyde on bimetallic Co and Ni catalysts
Activity and selectivity data for Co–Ni bimetallic catalysts at different reaction
temperatures are given in Table S2. Graphical representation of CAL conversion/
selectivity data for Co–Ni series of catalysts at 140 °C are presented in Fig. 7 and
discussed further. Under identical reaction conditions, substitution of a 0.2 atomic
fraction of Co with Ni (Co0.8Ni0.2) results in significant improvement in CAL
conversion (55–87.9%) vis- a` -vis Co/TiO , accompanied by sharp decrease in
2
selectivity for COL (61.3–2.9%) and increase in selectivity for HCAL
(
14.7–76.7%). On further substitution of Co by Ni to yield Co0.6Ni0.4 composition,
Table 3 Influence of feedstock composition on conversion (%)
Feed stock composition
Catalysts
100%CAL 100%HCAL 100%COL 50%HCAL ? 50%COL 50%CAL ? 50%COL
Co1.0
26.1
33.8
99.1
78.0
17.5
14.4
63.4
47.9
57.9
40.7 ? 25.4
97.6 ? 56.6
69.8 ? 31.8
21.3 ? 30.9
62.5 ? 5.8
98.0 ? 3.6
88.5 ? 90.6
55.8 ? 73.9
Co0.7 Ni0.3 84.0
Co0.6 Ni0.4 51.4
Ni1.0
62.1
Reaction conditions-(Reaction conditions-Reaction conditions: Temp-120 °C, 15 g isopropanol ? 1.65 g
water and 1.2 g CAL)
1
23

Bimetallic Co–Ni/TiO catalysts for selective…
2
nearly the same CAL conversion (88.4%) is maintained, but significant changes on
selectivity of products is observed. While the selectivity for HCAL decreases from
7
6.7% to 36.1%, selectivity for COL increases to 16.6% along with an increase in
selectivity towards HCOL (46.9%), indicating further hydrogenation of COL to
HCOL.
Bi-metallic catalyst with an intermediate composition, namely, Co0.7Ni0.3
,
exhibits interesting behavior, with high CAL conversion (98.1%), and selectivity to
COL and HCOL at 17 and 82.9% respectively, indicating very high overall
hydrogenation activity. TPR studies on this catalyst reveal increase in overall
reducibility (Figs. 3, S2).
The catalyst with composition with relatively higher Ni content, Co0.5Ni0.5
,
maintains nearly same CAL conversion (96%), but selectivity for HCAL is a
maximum at 80.4%, followed by 15.5% selectivity to HCOL and very low
selectivity to COL (2.2%). With further increase in Ni, the catalyst Co0.4Ni0.6, high
CAL conversion is maintained at 98.6% with HCAL (71.8%) and HCOL (24.6%) as
major products. The catalyst with composition Co0.2Ni0.8 displays CAL conversion
and a product selectivity pattern which closely resemble those for monometallic Ni.
To summarize, at a constant reaction temperature of 140 °C, an exceptional
increase in CAL conversion up to [ 98% is observed in composition range x = 0.3
to 0.6. Maximum selectivity to HCOL (82.9%) is realized with Co0.7Ni0.3, while
maximum selectivity to HCAL (80.4%) with Co0.5Ni0.5. While Co rich phases
facilitate further hydrogenation of COL to HCOL, Ni rich phases exhibit high
selectivity towards HCAL. Overall CAL hydrogenation activity at 140 °C, when
expressed as TOF (Table 1), displays a maximum value at the composition
Co0.5Ni0.5. Similar trends are observed at 120 and 130 °C as well.
In the present work, H -TPD studies (Fig. 6) have shown that, on Co(1-x)Nix
2
catalysts with x values in the range 0.3–0.6, the desorption of hydrogen from metal
sites occurs at lower temperatures compared to that on monometallic Co or Ni.
Thus, both Co–Ni alloy formation, as observed by TPR and XPS and weaker M–H
bonds improve the CAL conversion, vis- a` -vis monometallic Co and Ni. Weaker M–
H bonds, especially in Co0.7Ni0.3 and Co0.6Ni0.4, facilitate further hydrogenation of
COL to HCOL. On these catalysts, higher selectivity to COL (42.8 and 25.9%,
respectively) could be achieved at lower reaction temperature, 120 °C (Table S2)
Relevance of alloy formation in achieving higher CAL conversion is further
supported by carrying out CAL conversion on mechanical mixture of separately pre-
reduced Co/TiO and Ni/TiO . Mechanical mixture of pre-reduced component
2
2
catalysts, corresponding to the composition Co0.7Ni0.3 at 140 °C, showed CAL
conversion of 32.6% with selectivity to HCAL, HCOL, COL and ACTL being 68.6,
1
9
5.8, 10.5 and 5.0%, respectively, compared to the very high CAL conversion of
8.1% on Co0.7Ni0.3 prepared by simultaneous reduction of the metal acetates by
glucose, wherein alloy formation is facilitated.
As observed in the TPR and XPS measurements, Co–Ni alloy formation in the
composition range x = 0.3–0.6 enhances overall reducibility by controlling the
degree of interaction of the metals with the support. Alloying also shifts the d-band
centre, thus modifying the electronic character and influencing the nature of
adsorption of the reactants [51].
123

S. Ashokkumar et al.
Availability of active hydrogen with weaker M–H bond strength is known
26–36] to increase the activity for hydrogenation, facilitating the hydrogenation of
coke precursors in steam reforming of methane and dry reforming with CO . Such a
[
2
synergetic effect between Co-& Ni resulting in improved reducibility, dispersion
and alloy formation is responsible for the high activity of Co–Ni bimetallic catalysts
for hydro de-oxygenation (HDO) of bio-oils [52] and hydrogen production by
decomposition of cellulose and glycerol [53, 54]. Base metal catalysts with
appropriate composition of Co and Ni emerge as highly effective catalysts for
different hydrogenation and hydrogenolysis reactions and economically viable
alternatives to noble metal based catalysts.
It is pertinent to mention at this stage that Malobela et al. [15] could achieve COL
selectivity of 62% on Co–Ni (5% w/w each) catalyst at lower CAL conversion
(63%) on MWCNT support, which does not facilitate adsorption of CAL via the
C=C bond, due to the increase in electron density around Co/Ni.
In contrast to this work, Hui et al. [14] in their studies on ternary system Ni–Co–
B observed a moderate increase in activity (64.6% maximum) for CAL
hydrogenation over a broad composition of (Co/Co ? Ni) ranging from 0.2 to 0.5
and no change on selectivity to HCAL. H -TPD studies by Hui et al. [14] showed
2
that on Co–B desorption of hydrogen occurs at a lower temperature compared to
that on Ni–B indicating that the presence of Co with Ni could lead to weaker M–H
bonds, resulting in higher activity, which is in line with the observations in the
present work
Recycling and stability of the bimetallic catalyst
The used catalyst (Co0.3Ni0.7) after 1 h reaction time is separated and washed with
iso-propyl alcohol several times and hydrogenation of CAL is carried out again. The
catalyst displays nearly the same conversion and selectivity as that of the fresh
catalyst for four such cycles (Table S3). ICP-OES data (Table S1) shows that the
metal composition is intact with no leaching. No significant change in the crystallite
size of Co/Ni is observed on the TE micrograph (Fig. 2i) for the used catalyst,
indicating the stability of the catalyst under reaction conditions.
Conclusions
Selective hydrogenation of cinnamaldehyde has been investigated on a series of bi-
metallic Co–Ni catalysts in the composition range Co(1-x)Ni with x = 0.0, 0.2, 0.3,
x
0
TiO -P25. TPR and XPS results indicate the formation of nanoscale Co–Ni alloys
.4, 0.5, 0.6, 0.8 and 1.0, with total metal loading of 15% w/w and supported on
2
and increase in reducibility of Co and Ni in the composition range x = 0.3–0.6,
besides weakening of M–H bond strength, as revealed by H -TPD measurements.
2
Unlike the monometallic Co and Ni catalysts, bi-metallic Co–Ni catalysts in the
composition range x = 0.3–0.6 display very high conversion ([ 98%) of cin-
namaldehyde (CAL). Bimetallic Co0.7Ni0.3 catalyst displays high conversion of
CAL (98.1%) and high selectivity (82.9%) towards hydrocinnamyl alcohol
1
23

Bimetallic Co–Ni/TiO catalysts for selective…
2
(
displays a maximum value at the composition Co Ni . Co–Ni alloy formation,
HCOL). Overall CAL hydrogenation activity at 140 °C, when expressed as TOF,
0
.5 0.5
which modifies the electronic structure, d-band characteristics and M–H bond
strength, is responsible for the observed changes in activity and selectivity. The
catalysts display stable activity and selectivity for four cycles and no changes in the
metal composition or metal crystallite size are observed after use. A synergetic
effect, originating from an appropriate composition of Co and Ni and reaction
conditions, results in highly effective and economically viable base metal catalysts
for hydrogenation reactions, in comparison with noble metal based catalysts.
Acknowledgements The authors are thankful to the Department of Science and Technology, Govt. of
India, for establishing the research facilities at the National Centre for Catalysis Research for carrying out
this work and IIT Madras for providing infrastructure support.
Compliance with ethical standards
Conflict of interest The authors declared that they have no conflict of interest.
References
1. P. Gallezot, D. Richard, Catal. Rev. Sci. Eng. 40, 81 (1998)
2. A. Giroir-Fendler, D. Richard, P. Gallezot, in Studies in Surface Science and Catalysis, vol. 41, ed.
by J. Barrault, C. Bouchoule, D. Duprez, C. Montassier, M. Guisnet, G. P e´ rot (Elsevier, Amsterdam,
1988), pp. 171–178
3
4
5
6
7
8
. M.G. Prakash, R. Mahalakshmy, K.R. Krishnamurthy, B. Viswanathan, Catal. Today 263, 105 (2016)
. W. Yu, Y. Wang, H. Liu, W. Zheng, J. Mol. Catal. A 112, 105 (1996)
. B. Coq, P.S. Kumbhar, C. Moreau, P. Moreau, M.G. Warawdekar, J. Mol. Catal. 85, 215 (1993)
. B. Coq, P.S. Kumbhar, C. Moreau, P. Moreau, F. Figueras, J. Phys. Chem. 98, 10180 (1994)
. R. Basale, Ph.D. Thesis, (Tech. Univ. Eindhoven, Netherland, 2013)
. P. M a¨ ki-Arvela, L.P. Tiainen, M. Lindblad, K. Demirkan, N. Kumar, R. Sj o¨ holm, T. Ollonqvist, J.
V a¨ yrynenSalmi, DYu. Murzin, Appl. Catal. A 241, 271 (2003)
9. X. Yuan, J. Zheng, Q. Zhang, S. Li, Y. Yang, J. Gong, AIChE J. 60, 1036 (2012)
1
0. E. Asedegbega-Nieto, B. Bachiller-Baeza, A. Guerrero-Ru ´ı z, I. Rodriguez-Ramos, Appl. Catal. A
00, 120 (2006)
3
1. H. Noller, W.M. Lint, J. Catal. 85, 25 (1984)
2. S.A. Ananthan, V. Narayanan, Int. J. Chem. 3, 55 (2011)
1
1
1
1
1
1
1
1
1
3. B.M. Reddy, G.M. Kumar, I. Ganesh, A. Khan, J. Mol. Catal. A: Chem. 247, 80 (2006)
4. L.I. Hui, M.A. Chun-Jing, L.I. He-Xing, Acta Chim. Sinica 64, 1947 (2006)
5. L.J. Malobela, J. Heveling, W.G. Augustyn, L.M. Cele, Ind. Eng. Chem. Res. 53, 13910 (2014)
6. W. Lin, H. Chang, L. He, Y. Yu, F. Zhang, J. Catal. 303, 110 (2013)
7. K.J.A. Raj, M.G. Prakash, T. Elangovan, B. Viswanathan, Catal. Lett. 142, 87 (2011)
8. M. Vasseem, N. Tripathy, G. Khang, Y. Halm, RSC. Adv. 3, 9698 (2013)
9. H.B. Boubaker, M. Mhamdi, E. Marceau, S. Khaddar-Zine, A. Ghorbel, M. Che, Y.B. Taarit, F.
Villain, Microporous Mesoporous Mater. 93, 62 (2006)
2
0. R. Li, X. Ren, H. Ma, X. Feng, Z. Lin, X. Li, C. Hu, B. Wang, J. Mater. Chem. A 2, 5724 (2014)
1. R.J. Isaifan, H.A.E. Dole, E. Obeid, L. Lizarraga, P. Vernoux, E.A. Baranova, Electrochem. Solid-
State Lett. 15, E14 (2012)
2
2
2. K. Takanabe, K. Nagaoka, K. Nariai, K. Aika, J. Catal. 232, 268 (2005)
3. M.M. Van Schooneveld, C. Campos-Cuerva, J. Pet, J.D. Meeldijk, J. van Rijssel, A. Meijerink, B.H.
Erne, F.M.F. de Groot, J. Nanopart. Res. 14, 991 (2012)
2
24. X.B. Zhang, Y.J. Zhang, F. Chen, Y.Z. Xiang, B. Zhang, L.Y. Xu, T.R. Zhang, Reac. Kinet. Mech.
Cat. 115, 283 (2015)
123

S. Ashokkumar et al.
5. D. Golodnitsky, N.V. Gudin, G.A. Volyanuk, J. Electrochem. Soc. 147(11), 4156 (2000)
6. X. You, X. Wang, Y. Ma, J. Liu, W. Liu, X. Xu, H. Peng, C. Li, W. Zhou, P. Yuan, X. Chen, Chem.
Cat. Chem. 6, 3377 (2014)
2
2
2
7. M. Phongaksorn, S. Tungkamani, N. Dharmasaroja, T. Chamni, R. Chuvareec, N. Kanjanabat, N.
Siri-Nguanc, Chem. Eng. Trans. 43, 925 (2015)
8. X. Hu, G.X. Lu, J. Mol. Catal. A 261, 43 (2007)
2
29. L.W. Chen, C.K.S. Choong, Z.Y. Zhong, L. Huang, Z. Wang, J.Y. Lin, Int. J. Hydrog Energy 37,
16321 (2012)
3
3
3
3
3
0. H. Woo, R. Srinivasan, L. Rice, R. De Angelis, P. Reucroft, J. Mol. Catal. 59, 83 (1990)
1. I. Tatsumi, E. Koichi, A. Hiromichi, Appl. Catal. 30, 225 (1987)
2. P.C. Das, N.C. Pradhan, A.K. Dalai, N.N. Bakshi, Catal. Lett. 98, 153 (2004)
3. H. Luo, H. Li, H.L. Zhuang, Chem. Lett. 5, 404 (2001)
4. L. Wei, C. Xianian, L. Jin, Y. Ying, Z. Yongfu, H. Zhaolong, Y. Zhengzhou, Adv. Malts. Res. 430,
917 (2012)
3
3
3
3
3
4
4
5. W. Lin, H. Chang, L. He, Y. Yu, F. Zhang, J. Catal. 303, 110 (2013)
6. D. Xu, W. Li, H. Duan, Q. Ge, H. Xu, Catal. Lett. 102, 229 (2005)
7. N.S. McIntyre, M.G. Cook, Anal. Chem. 47, 2208 (1975)
8. H.L. Skriver, N.M. Rosengaard, Phys. Rev. B 46, 7157 (1992)
9. S. Velu, S. Gangwal, Solid State Ion. 177, 803 (2006)
0. A.G. Boudjahem, S. Monteverdi, M. Mercy, D. Ghanbaja, M.M. Bettahar, Catal. Lett. 84, 115 (2002)
1. J. Liu, C. Li, F. Wang, S. He, H. Chen, Y. Zhao, M. Wei, D.G. Evans, X. Duan, Catal. Sci. Technol.
3, 2627 (2013)
4
2. Y. Nitta, K. Ueno, T. Imanaka, Appl. Catal. 56, 9 (1989)
3. C. Rudof, B. Dragoi, A. Ungureanu, A. Chirieac, S. Royer, A. Nastroc, E. Dumitriu, Catal. Sci.
Technol. 4, 179 (2014)
4
4
4
4
4
4
4
4. U.K. Singh, M.A. Vannice, J. Catal. 199, 73 (2001)
5. L. Pauling, Proc. R. Soc. A 196, 343 (1949)
6. O. Beeck, Faraday. Soc. Discuss. 8, 118 (1950)
7. J.H. Sinfelt, Science 195, 641 (1977)
8. F. Delbecq, P. Sautet, J. Catal. 152, 217 (1995)
9. C.S. Fadley, D.A. Shirley, 3rd materials research symposium—electronic density of states. Nov. 3–6,
1969, Gaithersburg, MD
50. G. Blyholder, D. Shihabi, J. Catal. 46, 91 (1977)
51. T. Bligaard, J.K. Nørskov, in Chemical Bonding at Surfaces and Interfaces, 1st Edn., ed. by A.
Nilsson, L.G.M. Petterson, J.K. Nørskov (Elsevier, Amsterdam, 2008)
52. T.M. Huynh, U. Armbruster, L.H. Nguyen, D.A. Nguyen, A. Martin, J. Sustain. Bioenergy Syst. 5,
151 (2015)
53. M. Zhao, T.L. Church, A.T. Harris, Appl. Catal. B 101, 522 (2011)
54. Y. Wang, M. Chen, Z. Yang, T. Liang, S. Liu, Z. Zhou, X. Li, Appl. Catal A. Gen. 550, 214 (2018)
Affiliations
Saranya Ashokkumar ,2 • Vivekanandan Ganesan
1
2 •
1
1
Krishnamurthy K. Ramaswamy
•
Viswanathan Balasubramanian
&
1
Viswanathan Balasubramanian
bviswanathan@gmail.com
National Centre for Catalysis Research (NCCR), Indian Institute of Technology, Madras,
Chennai 600036, India
2
Department of Chemistry, A V C College, Mayiladuthurai 609305, India
1
23