Nickel boosts ring-opening activity of iridium

Article abstract of DOI:10.1002/cctc.201300844

A variety of bimetallic Ni-Ir catalysts were synthesised by preforming nanoparticles in the presence of polyvinylpyrrolidone, followed by deposition on γ-alumina and high-temperature polymer removal. The Ni-Ir (1:1 molar ratio) nanoparticles prepared by the hydrogen-sacrificial technique (Ir reduction on the preformed Ni nanoparticles with surface Ni hydride) allowed increasing indane ring opening activity per total amount of Ir as compared to monometallic Ir. The simultaneous reduction of Ni and Ir precursors was not as efficient. The catalysts were characterised with UV/Vis spectroscopy, TEM, temperature-programmed reduction, CO₂ temperature-programmed desorption, CO diffuse reflectance Fourier transform spectroscopy, X-ray photoelectron spectroscopy and CHN analysis. The study only explored the catalyst's metal function and allows saving rare and expensive iridium without loss of its outstanding performance as a ring-opening catalyst. Save the rare: To avoid inefficient use of rare and expensive catalytic metals, iridium atoms are placed only in the outermost layer of the nanoparticles, with inexpensive metal (nickel) inside, which boosts the catalytic performance.

Full text of DOI:10.1002/cctc.201300844

CHEMCATCHEM
FULL PAPERS
DOI: 10.1002/cctc.201300844
Nickel Boosts Ring-Opening Activity of Iridium
Hessam Ziaei-Azad and Natalia Semagina*[a]
A variety of bimetallic Ni–Ir catalysts were synthesised by pre-
forming nanoparticles in the presence of polyvinylpyrrolidone,
followed by deposition on g-alumina and high-temperature
polymer removal. The Ni–Ir (1:1 molar ratio) nanoparticles pre-
pared by the hydrogen-sacrificial technique (Ir reduction on
the preformed Ni nanoparticles with surface Ni hydride) al-
lowed increasing indane ring opening activity per total
amount of Ir as compared to monometallic Ir. The simultane-
ous reduction of Ni and Ir precursors was not as efficient. The
catalysts were characterised with UV/Vis spectroscopy, TEM,
temperature-programmed reduction, CO2 temperature-pro-
grammed desorption, CO diffuse reflectance Fourier transform
spectroscopy, X-ray photoelectron spectroscopy and CHN anal-
ysis. The study only explored the catalyst’s metal function and
allows saving rare and expensive iridium without loss of its
outstanding performance as a ring-opening catalyst.
Introduction
Catalytic ring opening (RO) is one of the key processes of in-
creasing the quality of heavy polyaromatic and naphthenic
compounds. During the process, a CÀC bond on the ring is
cleaved only once, leaving the number of carbon atoms in the
product unchanged.[1] RO increases the volume of the prod-
ucts, and if substituted CÀC bonds are the ones cleaved, it can
also improve the cetane number. These are two important in-
centives for the oil market.[2]
sitions; this unusual behaviour was attributed to be the effect
of support rather than the particle size.[13]
However, Ir is one of the rarest elements on earth, a concern
for the sustainable development. Is there a way to spare the
metal without losing its outstanding catalytic performance?
Could we place a much more inexpensive metal in the metal
nanoparticle core, surrounded by only one layer of the expen-
sive active noble metal, which would make all its atoms avail-
able for catalysis? Not with traditional preparation techniques
for heterogeneous catalysts. Such techniques, such as impreg-
nation, exert limited control over particle size, particle structure
and vicinity between the two metals.[19]
Since the 1970s, RO of cyclic compounds has been studied
extensively using the Pt-group supported catalysts.[1,3–7] The
RO of naphthenic rings relies on a metal-driven chemistry and
among the different metals studied, Ir is the most selective
and active ring-opening catalyst.[1,8,9] The RO of several cyclic
molecules such as methylcyclopentane, ethylcyclopentane,
methylcyclohexane, dimethylcyclopentane, dimethylcyclohex-
ane, indane, decalin, perhydroindane and trimethylcyclohexane
has been studied using Ir-supported catalysts.[1,2,5–7,10–14] Mech-
anism of the metal-catalysed RO of naphthenic rings depends
on the nature of the metal, its particle size, and support ef-
fects.[8,15] Iridium is reported to operate mostly through the di-
carbene mechanism, implying perpendicular adsorption of two
carbon atoms of a naphthenic ring, which results in the cleav-
age of unsubstituted secondary–secondary carbon atom
bonds.[15] The particle size effect in changing the mechanism of
RO is reportedly less pronounced for Ir,[16,17] as contrary to the
platinum.[17,18] A study of the support effects in the RO of 1,3-
dimethylcyclohexane revealed that Ir/SiO2 is selective towards
expected unsubstituted CÀC bond cleavage, whereas Ir/Al2O3
presented also a significant selectivity for RO at substituted po-
The controlled synthesis of bimetallic core–shell nanoparti-
cles, developed about a decade ago, presents an attractive so-
lution.[20] Colloid chemistry methods have synthesised bimetal-
lic nanoparticles in a successful, controlled way in recent years,
into various structures such as “random alloy” and “core–
shell”.[21] A significant number of core–shell nanoparticles for
catalytic applications have remained in the noble-metal
region: Au–Pd, Pd–Pt, Ru–Pt, Rh–Pt, Ir–Pt, Au–Pt, Au–Rh and
Pd–Ir.[22–30] Recent metal cores have comprised cheaper ele-
ments such as Ni, Co and Cu to prevent the waste of precious
metals at bulk positions. However, to the best of our knowl-
edge, synthesised nanoparticles have only been applied in low
temperatures (rarely exceeding 1208C), or synthesised by
using relatively complicated methods. Some examples include
a series of Co–M (M=Pd, Pt and Au) nanoparticles prepared
by colloid chemistry (redox–transmetallation),[31] Ni–Pt and Ni–
Pd nanoparticles synthesised by wet chemistry (polyol reduc-
tion) as the catalysts of proton-exchange membrane fuel cells
(operating typically at 50–1008C) or direct methanol fuel cells
(90–1208C),[32,33] polyol-assisted synthesis of Ni/Pd for hydroge-
nation of nitrobenzene to aniline at 278C,[34] Co/Pt nanoparti-
cles prepared by wet chemistry (redox–transmetallation) for
hydrogenation of several compounds at room temperature,[35]
Ni/Pd nanoparticles synthesised by combined electrochemistry
and wet chemistry methods, tested in a cross-coupling reac-
[a] H. Ziaei-Azad, Prof. N. Semagina
Department of Chemical and Materials Engineering
University of Alberta
9107-116 St., Edmonton, AB, T6G 2V4 (Canada)
Fax: (+1)780-492-2881
E-mail: semagina@ualberta.ca
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/cctc.201300844.
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2014, 6, 885 – 894 885

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
tion at 658C,[36] and Cu/Pt nanoparticles synthesised by electro-
deposition of Cu on carbon electrode followed by Pt shell for-
mation, used in a study of proton-exchange membrane fuel
cell catalysts.[37]
and expensive Ir, increase its activity per metal loading in the
reactor, and bring no loss to the RO product yield. We only tar-
geted the metal function of the material. We found that the
addition of Ni in the core position almost doubled RO activity
with the same amount of active Ir in the reactor.
The goal of the work was to replace a portion of rare and
expensive Ir with the more abundant and less expensive Ni,
without incurring any loss in catalytic performance per total Ir
loading in a reactor. Indane ring opening was used as a model
reaction, representing one of the routes in large-scale industri-
al fuel upgrading processes. Herein we report the synthesis of
bimetallic nanoparticles of Ni and Ir either through simultane-
ous reduction of metal precursors in the presence of the stabil-
ising polymer polyvinylpyrrolidone (PVP), yielding supposedly
alloy nanoparticles with random atom distribution, or through
so-called hydrogen-sacrificial technique, in which nickel hy-
dride, formed by hydrogenation of preformed nickel nanoparti-
cles, reduces and positions Ir atoms in the nanoparticle shell,
creating supposedly “Ni core–Ir shell” bimetallic particles
(Ni(c)Ir(s), Scheme 1). As demonstrated herein by a variety of
Results and Discussion
Characterisation of as-synthesised nanoparticles
The synthesised Ni, Ir and Ni–Ir nanoparticles were analysed by
UV/Vis spectroscopy and TEM. In Figure 1, the UV/Vis spectra
Figure 1. UV/Vis spectra taken during the preparation of Ni1Ir1 colloidal dis-
persion.
taken during the synthesis of Ni1Ir1 colloidal dispersion are
shown as an example. No peak was discernible in the UV/Vis
range for PVP or ethylene glycol at room temperature. If metal-
lic precursors were added to the solution, peaks appeared at
approximately 443 and 496 nm, which can be ascribed to the
ligand-to-metal charge-transfer absorption of metallic ions.[38]
As soon as the NaBH4 was added, the mixture turned dark
brown, a sign of nanoparticle formation. The UV/Vis spectrum
collected after heating to reflux confirmed that precursor
peaks disappear and absorption increases in the visible region.
Metallic nanoparticles absorb photons in UV/Vis region as
a result of the coherent oscillation of their conduction band
electrons.[39]
Scheme 1. Formation of Ni(c)Ir(s) nanoparticles by hydrogen-sacrificial
method.
characterisation techniques, after the high-temperature poly-
mer removal, the particles either retain or change their initial
size depending on the Ni/Ir ratio, and the supposedly core–
shell sample exhibits the expected nanoparticle shell enrich-
ment with Ir atoms. Indane RO was studied at atmospheric
pressure and 3508C (Scheme 2). The RO products, if the ring
was cleaved only once, were 2-ethyltoluene and n-propylben-
zene.
In Figure 2, size distribution histograms and an exemplary
TEM image of the supposedly Ni(c)Ir(s) nanoparticles are pre-
sented (other TEM pictures are shown in Figure S1 in the Sup-
porting Information). Particles are nearly spherical with
a narrow size distribution. The particles of the Ni(c)Ir(s) sample
have the largest size amongst all the synthesised nanoparticles
(3.2Æ0.6 nm), which implies the hydrogen-sacrificial technique
successfully deposited Ir atoms on preformed core Ni particles
(2.7Æ0.7 nm). The technique, as shown in Scheme 1, provides
the colloidal Ni core particles with hydrogen, which forms
NiÀH.[40] The hydride species then catalyse the reduction of iri-
dium ions on the surface of Ni core particles, forming core–
shell particles. We have recently used this technique to synthe-
sise Pd(c)Ir(s) particles.[30]
Herein, we did not focus on selectivity improvement or
cetane number increase, but on a material that will spare rare
Scheme 2. Reaction scheme for ring opening of indane at atmospheric pres-
sure and 3508C.
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2014, 6, 885 – 894 886

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
Figure 3. H2–TPR profiles of the selected g-Al2O3 supported catalysts. A posi-
tive inverted thermal conductivity detector (TCD) signal indicates hydrogen
consumption in the H2/Ar mixture.
calcined at 4208C, attributed to the reduction of O-rich com-
plexes.[41] The TPR spectrum of Ni/g-Al2O3 reveals a major peak
at 2198C (reduction of NiO phase),[43] followed by two smaller
peaks at 258 and 3168C. The latter peaks may show Ni species
interacting with the support or the reduction of bulk Ni, be-
cause Ni sinters upon calcination as shown below, or both.[13]
The deconvolution of bimetallic catalysts’ TPR profiles was
initiated by assuming the same reduction temperatures as in
monometallic samples; a fit could indicate the absence of bi-
metallic interactions. The TPR profile of Ni1Ir1 reveals three
peaks at 150, 210 and 2758C. The first two peaks appeared at
higher temperatures than those of mono-Ir, and the second
peak at 2108C can be also ascribed to the contribution from
the shifted 2198C Ni peak. This result agrees with the TPR
studies of Ir–Ni/g-Al2O3 catalysts, in which intimate contact be-
tween the oxide species of Ni and Ir delayed reduction of Ir
and facilitated that of Ni.[13] Such behaviour is a typical indica-
tion of alloy formation, which confirms our hypothesis of in-
trinsic bimetallic nanoparticle formation instead of the physical
mixture of mono-Ir and mono-Ni nanoparticles.[44] However,
the high-temperature reduction peak centred at 2758C re-
mained intact, which implies the presence of some segregated
mono-Ir oxygen-rich complexes as well.
Figure 2. Size distribution histograms and an exemplary TEM image:
a) Ni(c)Ir(s), b) Ni, c) Ni2Ir1, d) Ni1Ir1 and e) Ir. s=Standard deviation
Bimetallic interactions in the supported catalysts
The synthetic procedures for bimetallic nanoparticles may
form monometallic Ir and Ni nanoparticles, if the final bimetal-
lic system mixes Ni and Ir catalysts, as well as intrinsic bimetal-
lic Ni–Ir nanoparticles. The bimetallic interactions were probed
with H2 temperature-programmed reduction (H2–TPR) and CO2
temperature-programmed desorption (CO2–TPD) techniques,
X-ray photoelectron spectroscopy (XPS) and angle-resolved
XPS, diffuse reflectance infrared Fourier transform spectroscopy
(DRIFTS) of adsorbed CO and sintering studies. As shown
below, the catalytic performance indicates the formation of in-
trinsic bimetallic nanoparticles as well. For all the analyses,
samples were used as in the catalytic experiments, that is, cal-
cined and reduced at 4008C (apart from TPR).
In Figure 3, H2–TPR profiles of selected samples are shown
along with deconvoluted peaks for each graph. There are
three peaks in the TPR profile of monometallic Ir at 120, 200
and 2728C, respectively, after deconvolution. The peak at
1208C may represent the reduction of highly dispersed surface
Ir species, which agrees with other results.[6] The major peak at
2008C indicates the reduction of IrO2, which is formed during
the calcination period before the TPR,[41,42] whereas the peak
centred at 2728C resembles documented results for Ir/g-Al2O3
The deconvoluted TPR spectrum of Ni(c)Ir(s) reveals peaks at
165, 234, 275 and 3058C. The first two are at higher tempera-
tures than those of the monoforms and indicate bimetallic for-
mation. The peak at 2758C may indicate either some monome-
tallic Ir particles or the close contact between the Ir atoms in
the nanoparticle shell, resembling a monometallic Ir nanoparti-
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2014, 6, 885 – 894 887

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
cle’s surface. In addition to the formation of bimetallic species
discussed analogously for Ni1Ir1, the peak at 3058C may indi-
cate reduction of core Ni atoms in little or no contact with Ir
atoms. Thus, TPR profiles reveal bimetallic interactions in the
Ni–Ir samples, mostly indicating their intrinsic bimetallic
nature, with the possible coexistence of some monometallic Ir
structures. Simultaneous reduction for the alloy structure and
hydrogen-sacrificial technique for the core–shell does result in
different structures, despite the same molar ratio of Ni and Ir.
As Ni and Ir have different chemisorption behaviours toward
CO2,[45] carbon dioxide was used as a probe molecule on the
catalyst surface. In Figure 4 the CO2-TPD profiles of selected
Figure 5. DRIFT spectra of the adsorbed CO.
species, in which one Ir atom can accommodate two CO mole-
cules.[50] Monometallic Ni catalyst exhibits five peaks from 1824
to 2108 cmÀ1. The 1824 cmÀ1 and 1872 cmÀ1 bands may reflect
the bridged type of adsorbed CO;[49] the peaks at 1994 cmÀ1
and 2030 cmÀ1 the CO molecules adsorbed linearly on Ni parti-
cles.[52] The two peaks for each adsorption mode may be ex-
plained by CO adsorption on metal atoms with different coor-
dination numbers on a nanoparticle surface (terrace, edge and
corner atoms).[49] The weak band at 2108 cmÀ1 may reflect CO
adsorption on some Nid+ species resistant to reduction.[53]
In the DRIFT spectra of the two bimetallic samples, the band
at 1869 cmÀ1 in both Ni1Ir1 and Ni(c)Ir(s) is positioned be-
tween those of mono-Ir and mono-Ni (1865 and 1870 cmÀ1, re-
spectively), which can be related to the bridged adsorbed CO
molecules. However, neither sample gives any sign of the
1824 cmÀ1 peak of mono-Ni, but both present peaks at posi-
tions very close to that of the mono-Ir, which indicates Ir on
the samples’ surfaces. The bands of 2005 cmÀ1 in Ni1Ir1 and
2001 cmÀ1 in Ni(c)Ir(s) are close to the corresponding peak of
the mono-Ir (2008 cmÀ1). The small differences observed can
be related to the bimetallic interactions. All three, that is, Ir,
Ni1Ir1 and Ni(c)Ir(s), exhibit a peak at 2069 cmÀ1. Both bimetal-
lic samples also exhibit a peak at 2040 cmÀ1, with higher inten-
sity in Ni1Ir1 than in Ni(c)Ir(s). The 10 cmÀ1 difference with
monometallic Ni indicates changes in electronic properties of
Ni atoms in the bimetallic samples,[48,54,55] or new sites created
upon alloying.[54] Thus, like the TPR results, the DRIFT spectra
confirm bimetallic interactions in Ni–Ir samples, with possible
coexistence of monometallic Ir clusters.
Figure 4. CO2-TPD profiles of selected g-Al2O3 supported catalysts, after sub-
traction of the TPD data for the support.
supported catalysts are shown after subtracting the reference
TPD data for g-Al2O3 to eliminate the support effect. A relative-
ly sharp peak is observed for monometallic Ni catalyst at 808C,
which agrees with the literature, relating to the weakly ad-
sorbed CO2.[46,47] The Ir catalyst also exhibits a peak at 908C
but with a more pronounced tail than the Ni catalyst, which
continues up to approximately 3008C. Such tails indicate con-
tinuous desorption or dissociation of CO2 during TPD analy-
sis.[46] The core–shell structure’s desorption profile can neither
prove nor reject bimetallic interactions in the Ni(c)Ir(s) sample,
because it may be governed by tightly packed Ir atoms in the
nanoparticle shell. However, the Ni1Ir1 sample’s peak signifi-
cantly differs from that of the monoforms, which does indicate
bimetallic interactions.
In Figure 5, DRIFT spectra of the adsorbed CO for the same
catalysts at frequencies from 1800 to 2150 cmÀ1 along with de-
convoluted peaks are shown. Deconvolution for bimetallic
samples was performed in the same way as for TPR. For mono-
metallic Ir, peaks at 2069, 2008 cmÀ1 and 1865 cmÀ1 were de-
tected, which agrees with results reported for Ir catalysts.[48]
The band at 1865 cmÀ1 relates to the bridged type of adsorbed
CO,[49] and the one at 2008 cmÀ1 relates to linearly bonded
CO.[50,51] The band at 2069 cmÀ1 may reflect CO molecules ad-
sorbed linearly on large iridium crystallites,[51] or Ir–dicarbonyl
In Table 1, the XPS binding energy (BE) values of the Ni2p3/2
and Ir4f7/2 electrons are presented after deconvolution into
metallic and oxide components. The BE values of 852.4 and
855.2 eV in Ni/g-Al2O3 and 60.7 and 62.0 eV in Ir/g-Al2O3 match
reported numbers for metallic and oxide Ni and Ir, respective-
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2014, 6, 885 – 894 888

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
Table 1. XPS binding energy values for Ni and Ir (after peak deconvolu-
tion into metallic and oxide components).
Catalyst[a]
Binding energy [eV]
Ir4f7/2
Ni2p3/2
Ni
852.4
852.6
852.7
852.6
855.2
855.6
855.4
855.4
Ni2Ir1
Ni1Ir1
Ni(c)Ir(s)
Ir
60.4
60.7
60.6
60.7
61.5
61.8
61.9
62.0
[a] Support: g-Al2O3.
ly.[56,57,58] Bimetallic samples reveal minor but consistent shifts
of BE values from those of their monometallic counterparts,
approximately 0.2–0.3 eV higher than the BE values of Ni and
0.3 eV lower than from the BE values of Ir implying charge
transfer from Ni to Ir upon alloying, which can strengthen the
growing evidence of bimetallicity and new clusters. A similar
transfer of charge from Ni to Pt upon alloying has been report-
ed in the literature.[59]
Angle-resolved XPS (normal, 458 and 608 tilt) was also per-
formed on Ni(c)Ir(s). The normal beam provides the informa-
tion on the top 8–10 nm of the sample, and the tilted ones
reduce the depth by cos458 or cos608 factor (5.6–4 nm). As
some Ni(c)Ir(s) particles are close to the 4 nm range (3.2Æ
0.6 nm), we only expected some degree of Ir/Ni intensity ratio
increase upon tilting. Indeed, the ratio increased from 0.9 to
1.4, indicating Ir excess on the catalyst surface.
Alloying two metals may increase their resistance to sinter-
ing[41] as we evaluated as further evidence of bimetallicity. All
catalysts were calcined at 4008C, after depositing them on
g-Al2O3, to remove PVP. Monometallic Ni and Ir exhibited differ-
ent resistance. In Figure 6 and Figure S2, TEM images of fresh
and calcined catalysts are shown, and in Table 2 size changes
are listed. Monometallic Ni sintered severely after calcination
(into agglomerates of approximately 50 nm), whereas mono-Ir
retained its size. Oxidative agglomeration of Ir is known to
occur at calcination temperatures above 4008C.[41,60] The sam-
ples with the Ir-to-Ni molar ratio of equal or larger than 1, in-
Figure 6. TEM images of the a,c,e) fresh and b,d,f) calcined (2 h, 4008C) sup-
ported catalysts: a, b) Ni; c, d) Ni(c)Ir(s); e,f) Ir.
cluding the core–shell structure, did not sinter. This implies
that no monometallic Ni is present in the samples, and sup-
ports bimetallicity.
There is no evidence that the synthesised nanoparticles pre-
served their structure after the high-temperature oxidation and
reduction. On the contrary, particles with low Ir con-
tent sintered on calcination, and calcination probably
enriched the surface of all Ni-containing nanoparti-
Table 2. Size change and the stabiliser removal from the catalysts after calcination.[a]
cles with Ni. As the heat of sublimation of Ni
(101 kcalmolÀ1 [61]
is lower than that of Ir (160 kcal
)
Catalyst[b]
Nanoparticle size [nm]
Before[c] After calcination[d] [wt%]
Carbon content[e] Carbon rejection after
molÀ1),[62] the migration of Ni to the surface would
decrease the nanoparticles’ surface energy. Both Ni
and Ir have the face-centered cubic (fcc) crystal struc-
ture and are completely miscible in each other below
14558C.[63] Within the calcination–reduction tempera-
ture range used, we did not expect the bimetallic
nanoparticles to transform to a single thermodynami-
cally stable structure, which had previously been
shown to occur at approximately 700–800 K for Au–
Pt nanoparticles.[64] A study of the Pd–Pt nanoparti-
cles’ thermodynamic stability revealed that the trans-
formation process becomes significant at approxi-
calcination[f] [%]
Ni
2.7Æ0.7 sintered
(up to 50 nm)
2.5Æ0.5 agglomerates
(up to 10 nm)
8.40
–
97.4
Ni2Ir1
–
Ni1Ir1
Ni1Ir3
Ni(c)Ir(s)
Ir
2.3Æ0.5 no change
2.2Æ0.4 no change
3.2Æ0.6 no change
2.0Æ0.2 no change
7.01
–
3.99
6.90
99.6
–
99.7
99.7
[a] Calcination conditions: T=4008C, t=2 h, in air; [b] Support: g-Al2O3; [c] Unsup-
ported nanoparticles, based on TEM analysis; [d] Based on TEM analysis, see also
Figure 6; [e] Fresh catalyst; [f] Carbon content of pure calcined Al2O3 (0.08 wt%) was
subtracted from the numbers reported.
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2014, 6, 885 – 894 889

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
mately 1600 K.[65] The melting point of NiO is even higher than
that of Ni (19608C vs. 14558C, respectively); thus, the O2 at-
mosphere is not expected to promote sintering of the nano-
particles.
The addition of Ni to Ir with a molar ratio of equal or less
than 1 increased the catalytic activity. Above the specified
ratio, the activity dropped significantly as shown for Ni2Ir1.
However, the highest Ir-to-Ni ratio of 3 and the absence of Ni
did not lead to the highest activity. That was observed for the
1/1 Ni(c)Ir(s) sample, which had almost twice the activity of
monometallic Ir, and 50% more than the Ni1Ir1 sample with
the same metal ratio but synthesised by simultaneous metal
reduction. Thus, the hydrogen-sacrificial technique is para-
mount for achieving the highest activity. Since the amount of
Ir used was the same, and monometallic Ir does not sinter, the
observed activities further confirm bimetallicity, although the
presence of some monometallic Ir nanoparticles cannot be ex-
cluded.
The following catalytic reactions were performed at 3508C
after 4008C calcination and 3598C reduction. The characterisa-
tion analyses described above were conducted after calcina-
tion and reduction at 4008C (higher than the reaction temper-
ature); thus, the results cover the worst-case scenario of bimet-
allic particle rearrangement owing to hot-spot formation
during the reaction. The surface characteristics described can
be assumed to be close to the in situ surface arrangement of
Ni and Ir.
An interesting coincidence was found if calculating the turn-
over frequency (TOF) for the monometallic Ir catalyst. Based on
the diameter of an Ir nanoparticle (2.0 nm) and assuming an
ideal fcc cuboctahedron crystal with a unit-cell parameter of
0.384 nm, we found Ir dispersion based on van Hardeveld and
Hartog’s surface statics as 61%,[68] yielding TOFs of 0.7 mo-
Stabiliser removal before catalysis
The supported catalysts were calcined at 4008C in air for 2 h
before the catalytic reaction to remove PVP. PVP can be ad-
sorbed on the metal surface as a result of hydrophobic interac-
tions,[66] thus blocking part of the active surface sites. We did
not aim to find the best calcination conditions, but chose
a temperature higher than the free PVP decomposition tem-
perature, that is, approximately 3508C,[67] to ensure PVP’s re-
moval from particles capable of carbide formation. In Table 2,
the carbon content of the selected fresh and calcined catalysts
are presented analysed by CHN analysis. The calcination treat-
ment led to more than 99.6% removal of carbon in Ir-contain-
ing catalysts (97.4% for Ni), and XPS results revealed no detect-
able Cl or N traces in the samples. This confirms that calcina-
tion–reduction successfully rejected PVP and chlorine species.
l
indane molsurface IrÀ1> minÀ1. This value is almost the same as the
activity value of 0.8 molindane molIrÀ1 minÀ1 for the Ni(c)Ir(s) cata-
lyst. The similarity indicates that in the latter catalyst all Ir
atoms are accessible to the reactant, meaning they are in the
nanoparticle shell, with Ni supporting them. However, syner-
gism cannot be excluded, because some monometallic Ir nano-
particles with lower dispersion may be also present.
The core–shell sample also allowed the highest production
rate of single-cleavage products (i.e., 2-ethyltoluene and n-pro-
pylbenzene). The lower selectivities of this catalyst for 2-ethyl-
toluene and n-propylbenzene than those of the other catalysts
(see Table 4) should not be used for comparison, because they
were observed at much higher conversions. Ni addition did
not improve the ratio of n-propylbenzene to 2-ethyltoluene.
This is in-line with the dicarbene RO mechanism, which results
in the cleavage of unsubstituted secondary–secondary carbon
atom bonds, i.e., yielding 2-ethyltoluene for this specific reac-
tion, and agrees with the literature, which suggests a dicarbene
mechanism domination for both Ni and Ir in RO reactions.[15]
The observed low light hydrocarbon (lights) formation agrees
with the literature data on Ir-catalysed RO reactions.[1,7] As
a result of this work, we can replace the rare and expensive Ir
with Ni and increase the RO activity per total Ir loading
in the reactor.
Catalytic activity in indane RO
In Table 3 the catalytic activity results of the synthesised cata-
lysts are shown. As monometallic Ni has no activity for the RO,
the results are reported per mole of Ir, and the amount of Ir
was standardised (0.8 mg). The negligible Ni RO activity ob-
served is in-line with data presented in the literature: in an ex-
tensive study dealing with the RO of several five and six-mem-
bered-ring naphthenes,[1] the activity of Ir/Al2O3 for the RO of
methylcyclopentane at 2758C was found to be 50 times higher
than that of the Ni/Al2O3.
The results obtained can be compared with reported
data on the atmospheric RO of indane at 3258C using
a series of in-house Pt–Ir catalysts.[7,44,48] For the most
active bimetallic formulation of Pt5Ir95 (based on wt%)/
Table 3. Catalytic activity of the synthesised catalysts in atmospheric RO of inda-
ne.[a]
Catalyst[b] Ni/Ir loa-
ding [wt%]
Activity
Productivity
[10À1 molindane molIrÀ1 minÀ1] [10À1 mol(ET+PB)molIrÀ1 minÀ1[c]
g-Al2O3, the product selectivities at 55% indane conver-
sion at 275 min on stream were reported as 67.1, 8.8,
12.9, 3.1, 4.1, 0.4 and 3.6 (wt%) for 2-ethyltoluene,
n-propylbenzene, o-xylene, ethylbenzene, toluene, ben-
zene and lights, respectively.[7] These selectivities match
the selectivities of our Ni1Ir3 catalyst (or Ni9Ir91 in
wt%), as demonstrated in Table 4. This finding indicates
that Pt can be successfully replaced by Ni without loss
in selectivity. In terms of activity, using the reported
Ni
0.165/0
<0.1
1.3
5.6
5.1
8.3
<0.1
1.2
4.5
3.9
6.4
Ni2Ir1
Ni1Ir1
Ni1Ir3
Ni(c)Ir(s) 0.052/0.165
Ir 0/0.220
0.067/0.118
0.044/0.163
0.022/0.218
4.3
3.4
[a] Reaction conditions: T=3368C, 125 min on stream; samples with 0.8 mg of Ir
were used for each activity measurement. [b] Support: g-Al2O3. [c] ET=2-ethylto-
luene, PB=n-propylbenzene.
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2014, 6, 885 – 894 890

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
troscopy, CO diffuse reflectance
Table 4. Selectivities of the synthesised catalysts in atmospheric ring opening of indane.[a]
infrared
Fourier
transform
Catalyst[b]
Selectivity [wt%]
spectroscopy, temperature-pro-
grammed reduction and CO2
temperature programmed de-
sorption and CHN analysis were
performed on a variety of Ni–Ir
nanoparticles synthesised in the
presence of polyvinylpyrrolidone
and deposited on alumina
followed by high-temperature
polymer removal. The particles
were mostly bimetallic, and the
presence of Ir enhanced their
2-Ethyltoluene n-Propylbenzene o-Xylene Ethylbenzene Toluene Benzene Light hydrocarbons
Ni
<1% conversion
Ni2Ir1
Ni1Ir1
Ni1Ir3
Ni(c)Ir(s) 51
Ir 64
69
64
66
20
17
11
9
3
10
13
20
10
0
1
2
3
1
4
4
5
9
4
2
1
0
1
1
2
3
3
7
4
16
[a] Reaction conditions: T=3368C, 125 min on stream; samples with 0.8 mg of Ir were used for each activity
measurement; note that the selectivities are at different conversions (activities, Table 3). [b] Support: g-Al2O3.
values for Pt5Ir95 catalyst (H2 flow 4.8 dm3(n)hÀ1, H2/indane
stability towards sintering. The Ni(core)–Ir(shell) catalyst exhib-
ited twice higher activity in indane ring opening at 3508C than
monometallic Ir, and allowed higher yield of single cleavage
products. Ni presence did not change the dicarbene path for
ring-opening characteristic of Ir. The study has developed a ma-
terial that will spare rare and expensive iridium and enhance
its performance. Supposedly, the introduction of acidic func-
tion into the support may further improve ring opening yield
by the 6-member ring isomerisation to a 5-member ring.
molar ratio of 63, catalyst amount 0.3 g, sum of metal loadings
[44]
2 wt%
and 83% dispersion[48]) the TOF can be calculated as
1.2 molindane molsurface atomsÀ1 minÀ1
.
The TOF value for our
Ni(c)Ir(s) catalyst at the same 3258C can be estimated from
0.8 molindane molsurface atomsÀ1 minÀ1 at 3368C and the reaction ac-
tivation energy of 78 kJmolÀ1 (found previously for the mono-
metallic Ir catalyst[69]) as 0.6 molindane molsurface atomsÀ1 minÀ1. Con-
sidering the fact that the data stem from two different groups,
the value may be considered comparable to the value for the
Pt5Ir95 catalyst; however, the catalyst reported herein does
not contain Pt and has a lower amount of Ir.
Experimental Section
The TEM picture of exemplary Ni1Ir1 catalyst before and
after the catalytic reaction (200 min) is presented in Figure 7
and reveals no observable particle sintering after the reaction.
Materials
Hydrogen hexachloroiridate(IV) hydrate (H2Cl6Ir·xH2O, Sigma–Al-
drich), nickel chloride hexahydrate (NiCl2·6H2O, Sigma–Aldrich), pol-
yvinylpyrrolidone (PVP, average molecular weight 40000, Sigma–Al-
drich), sodium borohydride (NaBH4, Sigma–Aldrich), gamma alumi-
na (g-Al2O3, average pore size 58 ꢁ, specific surface area
155 m2 gÀ1, Sigma–Aldrich), ethylene glycol (Fischer Scientific) and
acetone (>99.7%, Fischer Scientific) were used as received. Nitro-
gen, argon and hydrogen of ultrahigh purity (99.999%) were pur-
chased from Praxair. Benzocyclopentane (indane, 95vol%, Sigma–
Aldrich) was distilled once.
Catalyst preparation
Ni and Ir mono- and bimetallic nanoparticles were preformed in
a colloidal dispersion before deposition on g-Al2O3. A modified
polyol process, which applies a strong reducing agent, NaBH4, was
adapted for the synthesis of nanoparticles in this study. The
method has precedent for Ni.[70] Quantities of 0.1 mmol of either Ni
or Ir metal precursor were dissolved in 25 mL of ethylene glycol
and added to a 500 mL three–neck flask. PVP was then added with
the PVP/metal ratio of 40:1 and 20:1 (molar basis) to the Ni and Ir
containing flasks, respectively. The final volume of the solution was
increased to 175 mL by adding extra ethylene glycol. The resulting
mixture was stirred vigorously under a constant flow of nitrogen
and heated until the temperature reached 130–1408C. NaBH4
(ꢀ0.38 g) was added to the flask at this temperature very quickly
with adequate care. The reduction reaction occurred immediately,
the solution turned into dark brown with vapour release. Heating
was continued until the mixture came to boil at approximately
1978C and then heated at reflux in air for 1 h to complete the syn-
thesis of PVP-protected nanoparticles. No precipitate was observed
at the end of synthesis.
Figure 7. TEM images of Ni1Ir1 catalyst: a) before the catalytic reaction (cal-
cined and reduced in situ) and b) after the 200 min on stream.
Conclusion
The correct method to synthesise bimetallic nanoparticles is
essential in developing a material with enhanced catalytic per-
formance. A 1:1 molar ratio of Ni/Ir led to doubled catalytic ac-
tivity only if Ir was placed preferably in the nanoparticle shell.
This was possible owing to applying a hydrogen-sacrificial
technique applied to preformed Ni core nanoparticles, as op-
posed to the simultaneous reduction of Ni and Ir precursors.
Transmission electron microscopy, X-ray photoelectron spec-
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2014, 6, 885 – 894 891

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
Bimetallic nanoparticles were synthesised with either simultaneous
reduction of metal precursors (co-reduction) or the successive re-
duction assisted by hydrogen (hydrogen-sacrificial reduction
method) to produce random alloy and core–shell particles, respec-
tively. The PVP/(Ir+Ni) ratio of 20:1 (molar basis) was used for all
the bimetallic nanoparticles. Three different Ni–Ir alloys with the
molar ratios of Ni/Ir=3:1, Ni/Ir=1:1 and Ni/Ir=1:3 were synthes-
ised by the co-reduction of the metallic precursors. In all cases,
quantities of 0.1 mmol of (Ni+Ir) were used and the same amount
of NaBH4 (0.38 g) was added at 130–1408C. The resulting colloidal
solution obtained was always macroscopically homogeneous and
transparent without any precipitate. As the final Ni/Ir molar ratios
in the g-Al2O3-supported catalysts were found as 1.9:1, 0.9:1 and
1:3, respectively, they are referred as Ni2Ir1, Ni1Ir1 and Ni1Ir3. In
addition to the alloy particles mentioned above, another bimetallic
nanoparticle composition, Ni(core)–Ir(shell) was synthesised (with
the molar ratio of 1:1) using the hydrogen-sacrificial reduction
method, and referred as Ni(c)Ir(s). For the latter combination, Ni
core nanoparticles were synthesised as detailed below. After the
heating time at reflux, colloidal solution was allowed to cool to RT
followed by purging with hydrogen for 1.5 h. This procedure cre-
ates Ni hydride,[40] which further served to reduce Ir and forms an
Ir shell around the preformed Ni nanoparticles. To build an Ir shell,
H2Cl6Ir·xH2O (0.1 mmol) was dissolved in ethylene glycol (ꢀ50 mL)
and added to the mixture of Ni nanoparticles drop by drop. A sy-
ringe pump (Fisher Scientific 01001) was used to transfer Ir solu-
tion to the synthesis flask with a flow rate of approximately
20 mLhÀ1. The synthesis flask was constantly purged with H2
during the addition of Ir and at least one hour afterwards. Each re-
sulting colloidal dispersion was macroscopically homogeneous and
transparent with no precipitate.
were individually counted for 600 s using a 22% relative efficiency
ORTEC hyperpure Ge detector housed in a 10 cm Pb cave. The Ge
detector was connected to a PC-based Aptec multichannel analy-
ser card.
Fresh and calcined supported catalysts were analysed by an Ele-
mentar Vario Micro elemental analyser for CHN percentage.
H2–TPR was performed by using an AutoChem 2950 HP device (Mi-
cromeritics). Quantities of approximately 500 mg of the calcined
catalysts were used for the analysis. Samples were initially reduced
in a 50 mLminÀ1 flow of a mixture of 10% H2 in Ar at T=4008C for
1 h, then cooled to ambient temperature and oxidised in
a 50 mLminÀ1 flow of 10% O2 in He at T=4008C for 1 h. The TPR
profiles were collected with a 10 mLminÀ1 mixture of 10% H2 in Ar
from ambient temperature to 4008C with a heating rate of
108CminÀ1. Peak deconvolution was performed by using Origin
software.
CO2–TPD was performed with the same AutoChem 2950 HP
device. Quantities of approximately 250 mg of the similar calcined
catalysts but with higher metal loading (ꢀ1 wt%) were used for
this study because the Al2O3 also adsorbed CO2. Before CO2 ad-
sorption, samples were initially outgassed in Ar at 3008C for 2 h,
and then reduced by a mixture of 10% H2 in Ar at 4008C for 1 h.
After the reduction pretreatment, the active gas was switched to
He and the temperature was reduced to 508C. A mixture of 3%
CO2 in He was passed through the samples at 508C for 1 h with
the flow of 50 mLminÀ1. Physically adsorbed CO2 was then re-
moved by flushing 50 mLminÀ1 of He at 508C for 1 h. The TPD
spectra were collected by flowing 20 mLminÀ1 He from ambient
temperature to 3008C with a heating rate of 108CminÀ1
.
DRIFTS of adsorbed CO was performed with a Nicolet Nexus670
spectrophotometer, at a resolution of 4 cmÀ1in the range 800–
3000 cmÀ1and averaged over 128 scans. The Ni–Ir nanoparticles
were supported on g-Al2O3 with the higher loading of approxi-
mately 1 wt%. Following the same calcination treatment, that is,
2 h at 4008C in air, catalysts were reduced in the AutoChem 2950
with a flow of 50 mLminÀ1 of 10% H2 in Ar from ambient tempera-
ture to 4008C and kept for 1 h. Subsequently, they were allowed
to cool to RT under a flow of Ar and later a mixture of 3% CO in
He with a flow of 50 mLminÀ1 was passed over the samples for ap-
proximately 45 min. CO-adsorbed catalysts were further treated
with a flow of Ar at RT for approximately 30 min to desorb the
physically adsorbed CO; the DRIFTS spectra of the CO-adsorbed
catalysts were recorded immediately. KBr from Sigma–Aldrich was
dried at approximately 2008C overnight and used for collecting
the background spectra. Samples were also diluted with KBr at the
ratio of approximately 20 wt.% before placing in the DRIFTS cell.
The reflectance spectra of the catalysts were transformed to Kubel-
ka–Munk using the OMNIC software. OMNIC and Origin software
were used for further data processing, visualisation and deconvolu-
tion of the spectra.
All synthesised nanoparticles were supported on pre-dried g-Al2O3
by using acetone to precipitate them.[71] Acetone (ꢀ800 mL) was
added to the suspension of nanoparticles (in colloidal solution)
and g-Al2O3, and it was stirred vigorously for 2 h. Then, the super-
natant was removed from the beaker and the resulting powder
was washed several times with acetone and dried at ambient over-
night to form fresh catalysts, which then were calcined in air at
4008C for 2 h to ensure PVP removal.
Catalyst characterisation
Formation of nanoparticles in the colloidal solution was studied by
UV/Vis spectroscopy by using a Varian Cary 50 Scan UV visible
spectrometer with a 1 cm quartz cell.
TEM images of as-synthesised and supported nanoparticles were
recorded with a JEOL JEM2100 TEM device with a LaB6 filament
operating at 200 kV. The system was equipped with a Gatan GIF
Tridiem energy filter with a 2kꢂ2k digital camera. Samples for TEM
analysis were prepared by placing a drop of the colloidal disper-
sion of nanoparticles onto a carbon-coated copper grid, followed
by the evaporation of the solvent in a vacuum oven at 508C. The
average diameter and the standard deviation were calculated by
counting over 200 particles by using the ImageJ software. In the
case of supported catalysts, the sample (0.1 g) was suspended in
ethanol (5 mL) and dispersed in an ultrasonic bath for approxi-
mately 5 min before placing a drop of the suspension on the TEM
grid.
XPS was performed by using a Kratos Axis 165 X-ray photoelec-
tron spectrometer with a mono AlKa source operating at 15 mA
and 14 kV. The XPS analysis was done on pelletised samples that
were calcined in air at 4008C for 2 h followed by the reduction in
hydrogen at 4008C for 1 h. Background subtraction (Shirley-type),
smoothing and peak fitting were performed by using the CasaXPS
software package. All the core-level spectra were corrected with C
1 s at 284.6 eV.
The loading of the Ni and Ir in the calcined catalysts was deter-
mined by neutron activation analysis at the University of Alberta
SLOWPOKE facility. Samples were irradiated in the nuclear reactor
for 600 s following a decay period of ꢁ24 h. Irradiated samples
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2014, 6, 885 – 894 892

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
[9] H. Shi, O. Y. Gutierrez, G. L. Haller, D. Mei, R. Rousseau, J. A. Lercher, J.
Catal. 2013, 297, 70–78.
Catalytic behaviour in indane RO
Catalytic studies were performed at atmospheric pressure in a flow
stainless-steel reactor (length 0.406 m, inner diameter 0.013 m).
Calcined catalysts were diluted with #150 mesh SiC (ꢀ2 g). Cata-
lyst amounts corresponding to 0.8 mg Ir were used for each run
except for monometallic Ni, in which 0.8 mg Ni used. Gas flows
were regulated with calibrated mass-flow controllers (Sierra Instru-
ments). The system was purged with Ar before and after each test.
Catalysts were initially pretreated in situ with 80 mLminÀ1 H2 from
ambient temperature to 3598C for 1 h. Afterwards, the reactor
temperature and the H2 flow rate were changed to 3508C and
120 mLminÀ1, respectively. The catalytic reaction started by bub-
bling hydrogen through the indane bubbler kept at the constant
temperature water bath of 108C that corresponds to 70.8 Pa
indane partial pressure assuming saturation. The actual pressure
was found as 68.4 Pa and indane flow rate of 2.8ꢂ10À5 gminÀ1
was confirmed with GC, which was calibrated with a Gilson HPLC
pump. The reaction products were analysed online every 25 min
with a 430 Varian Gas chromatograph equipped with an FID detec-
tor; the details has been explained previously.[30] Each catalyst was
tested for 200 min on stream and 6 data points were collected at
75, 100, 125, 150, 175 and 200 min. Results were repeatable within
2–5% experimental error. The absence of external and internal
mass transfer limitations was verified experimentally by the
Madon–Boudart method. Within the indicated times on stream,
catalysts did not show significant deactivation. The reaction prod-
ucts were confirmed by an Agilent Technologies 7890 GC coupled
with a 5975C MSD as described previously.[30]
[10] A. Djeddi, I. Fechete, F. Garin, Catal. Commun. 2012, 17, 173–178.
[11] R. Moraes, K. Thomas, S. Thomas, S. V. Donk, G. Grasso, J.-P. Gilson, M.
Houalla, J. Catal. 2012, 286, 62–77.
[12] J.-W. Park, K. Thomas, J. V. Gestel, J.-P. Gilson, C. Collet, J.-P. Dath, M.
Houalla, Appl. Catal. A 2010, 388, 37–44.
[13] S. Dokjampa, T. Rirksomboon, D. T. M. Phuong, D. E. Resasco, J. Mol.
Catal. A 2007, 274, 231–240.
[14] N. Semagina, X. Yin, J. Shen, K. Loganathan, US Patent, No. US-2013-
0248414A1, 2013.
[15] P. T. Do, W. E. Alvarez, D. E. Resasco, J. Catal. 2006, 238, 477–488.
[16] C. G. Walter, B. Coq, F. Figueras, M. Boulet, Appl. Catal. A 1995, 133, 95–
102.
[17] J. G. van Senden, E. H. van Broekhoven, C. T. J. Wreesman, V. Ponec, J.
Catal. 1984, 87, 468–477.
[18] F. Luck, J. L. Schmitt, G. Maire, React. Kinet. Catal. Lett. 1982, 21, 219–
224.
[19] B. Coq, F. Figueras, J. Mol. Catal. A 2001, 173, 117–134.
[20] Y. Wang, N. Toshima, J. Phys. Chem. B 1997, 101, 5301–5306.
[21] N. Toshima, H. Yan, Y. Shiraishi in Metal Nanoclusters in Catalysis and Ma-
terial Science: The Issue of Size Control (Eds.: G. Schmid, N. Toshima, B.
Corain), Elsevier Science, Amesterdam, 2008, pp. 49–75.
[22] G. Schmid, H. West, J. O. Halm, J. O. Bovin, C. Grenthe, Chem. Eur. J.
1996, 2, 1099–1103.
[23] Y. Mizukoshi, T. Fujimoto, Y. Nagata, R. Oshima, Y. Maeda, J. Phys. Chem.
B 2000, 104, 6028–6032.
[24] N. Toshima, Y. Shiraishi, A. Shiotsuki, D. Ikenaga, Y. Wang, Eur. Phys. J. D
2001, 16, 209–212.
[25] A. U. Nilekar, S. Alayoglu, B. Eichhorn, M. Mavrikakis, J. Am. Chem. Soc.
2010, 132, 7418–7428.
[26] N. Toshima, Pure Appl. Chem. 2000, 72, 317–325.
[27] N. Toshima, K. Hirakawa, Polym. J. 1999, 31, 1127–1132.
[28] S. J. Cho, S. K. Kang, Catal. Today 2004, 93, 561–566.
[29] L. D’Souza, S. Sampath, Langmuir 2000, 16, 8510–8517.
[30] H. Ziaei-azad, C.-X. Yin, J. Shen, Y. Hu, D. Karpuzov, N. Semagina, J.
Catal. 2013, 300, 113–124.
The desired single-cleavage products of RO were 2-ethyltoluene
and n-propylbenzene. Main dealkylation products were o-xylene,
ethylbenzene, toluene and benzene. Compounds with GC reten-
tion times less than that of benzene were named “lights”. The
same main products were reported in ring opening of indane over
Pt–Ir catalysts.[7]
[31] W.-R. Lee, M. G. Kim, J.-R. Choi, J.-I. Park, S. J. Ko, S. J. Oh, J. Cheon, J.
Am. Chem. Soc. 2005, 127, 16090–16097.
[32] S. Bhlapibul, K. Pruksathorn, P. Piumsomboon, Renewable Energy 2012,
41, 262–266.
[33] Y. Zhao, X. Yang, J. Tian, F. Wang, L. Zhan, Int. J. Hydrogen Energy 2010,
35, 3249–3257.
[34] K. Nagaveni, A. Gayen, G. N. Subbanna, M. S. Hegde, J. Mater. Chem.
2002, 12, 3147–3151.
Selectivity values are reported not on mole, but on a mass basis
(because up to nine moles of methane may be produced from one
mole of indane). Response factors of different products with re-
spect to indane were found by performing GC flame ionisation de-
tector calibration with reference gas cylinders.
[35] C.-H. Jun, Y. J. Park, Y.-R. Yeon, J.-R. Choi, W.-R. Lee, S.-J. Ko, J. Cheon,
Chem. Commun. 2006, 1619–1621.
[36] L. Durꢄn Pachꢅn, M. B. Thathagar, F. Hartl, G. Rothenberg, Phys. Chem.
Chem. Phys. 2006, 8, 151–157.
Acknowledgements
[37] Z. D. Wei, Y. C. Feng, L. Li, M. J. Liao, Y. Fu, C. X. Sun, Z. G. Shao, P. K.
Shen, J. Power Sources 2008, 180, 84–91.
Financial support from the Imperial Oil–Alberta Ingenuity Center
for Oil Sands Innovation at the University of Alberta is gratefully
acknowledged.
[38] S. K. Singh, Q. Xu, Chem. Commun. 2010, 46, 6545–6547.
[39] S. Link, M. A. El-Sayed, J. Phys. Chem. B 1999, 103, 8410–8426.
[40] J. Lapujoulade, K. S. Neil, Surf. Sci. 1973, 35, 288–301.
[41] Y.-J. Huang, S. C. Fung, W. E. Gates, G. B. Mcvicker, J. Catal. 1989, 118,
192–202.
[42] S. Subramanian, J. A. Schwarz, Appl. Catal. 1991, 74, 65–81.
[43] J. Zielinski, J. Mol. Catal. 1993, 83, 197–206.
[44] U. Nylꢃn, J. F. Delgado, S. Jaras, M. Boutonnet, Appl. Catal. A 2004, 262,
189–200.
[45] J. Hagen, Industrial Catalysis: A practical approach, Wiley-VCH, Wein-
heim, 2006.
[46] H.-S. Roh, K.-W. Jun, W.-S. Dong, J.-S. Chang, S.-E. Park, Y.-I. Joe, J. Mol.
Catal. A 2002, 181, 137–142.
Keywords: alloys · cleavage reactions · iridium · nickel ·
nanoparticles
[1] G. B. McVicker et al., J. Catal. 2002, 210, 137–148 see Supporting Infor-
mation.
[2] R. C. Santana, P. T. Do, M. Santikunaporn, W. E. Alvarez, J. D. Taylor, E. L.
Sughrue, D. E. Resasco, Fuel 2006, 85, 643–656.
[3] W. E. Alvarez, D. E. Resasco, J. Catal. 1996, 164, 467–476.
[4] D. Teschner, K. Matusek, Z. Paal, J. Catal. 2000, 192, 335–343.
[47] M. Cai, J. Wen, W. Chu, X. Cheng, Z. Li, J. Nat. Gas Chem. 2011, 20, 318–
324.
ˇ
[5] D. Kubicka, M. Kangas, N. Kumar, M. Tiitta, M. Lindblad, D.Yu. Murzin,
Top. Catal. 2010, 53, 1438–1445.
[48] U. Nylꢃn, B. Pawelec, M. Boutonnet, J. L. G. Fierro, Appl. Catal. A 2006,
299, 14–29.
[49] N. Sheppard, C. D. L. Cruz, Catal. Today 2001, 70, 3–13.
[50] G. B. McVicker, R. T. K. Baker, R. L. Garten, E. L. Kugler, J. Catal. 1980, 65,
207–220.
[6] A. Djeddi, I. Fechete, F. Garin, Appl. Catal. A 2012, 413–414, 340–349.
[7] U. Nylꢃn, L. Sassu, S. Melis, S. Jaras, M. Boutonnet, Appl. Catal. A 2006,
299, 1–13.
[8] H. Du, C. Fairbridge, H. Yang, Z. Ring, Appl. Catal. A 2005, 294, 1–21.
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2014, 6, 885 – 894 893

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
[51] F. Solymosi, J. Rasko, J. Catal. 1980, 62, 253–263.
[52] G. Poncelet, M. A. Centeno, R. Molina, Appl. Catal. A 2005, 288, 232–
242.
[53] M. Mihaylov, T. Tsoncheva, K. Hadjiivanov, J. Mater. Sci. 2011, 46, 7144–
7151.
[54] P. Samoila, M. Boutzeloit, C. Especel, F. Epron, P. Marecot, J. Catal. 2010,
276, 237–248.
[55] O. Roduner, E. Terekhina, J. Phys. Chem. C 2012, 116, 6973–6979.
[56] J. F. Moulder, W. F. Stickle, P. E. Sobol, K. D. Bomben, Handbook of X-ray
Photoelectron Spectroscopy, Ulvac-Phi, Chigasaki and Physical Electron-
ics, Minnesota, 1995.
[63] S. C. Yang, N. Chen, P. Nash, Ir-Ni Phase Diagram, ASM Alloy Phase Dia-
grams Center, P. Villars, editor-in-chief; H. Okamoto and K. Cenzual, sec-
tion editors; http://www1.asminternational.org/AsmEnterprise/APD,
ASM International, Materials Park, OH, 2006.
[64] H. B. Liu, U. Pal, J. A. Ascencio, J. Phys. Chem. C 2008, 112, 19173–
19177.
[65] R. Huang, Y.-H. Wen, Z.-Z. Zhu, S.-G. Sun, J. Phys. Chem. C 2012, 116,
8664–8671.
[66] T. Miyake, M. Teranishi, Chem. Mater. 1998, 10, 594–600.
[67] R. M. Rioux, H. Songa, M. Grass, S. Habas, K. Niesz, J. D. Hoefelmeyer, P.
Yang, G. A. Somorjai, Top. Catal. 2006, 39, 167–174.
[57] R. Nyholm, A. Berndtsson, N. Martensson, J. Phys. C 1980, 13, L1091–
L1096.
[58] M. Wu, D. M. Hercules, J. Phys. Chem. 1979, 83, 2003–2008.
[59] B. Moraweck, A. J. Renouprez, E. K. Htil, R. Baudoing-Savois, J. Phys.
Chem. 1993, 97, 4288–4292.
[68] R. Van Hardeveld, F. Hartog, Surf. Sci. 1969, 15, 189–230.
[69] A. Chhabra, M.Sc. Thesis, University of Alberta, 2012.
[70] G. G. Couto, J. J. Klein, W. H. Schreiner, D. H. Mosca, A. J. A. de Oliveira,
A. J. G. Zarbin, J. Colloid Interface Sci. 2007, 311, 461–468.
[71] X. Yan, H. Liu, K. Y. Liew, J. Mater. Chem. 2001, 11, 3387–3391.
[60] R. L. Garten, J. H. Sinfelt, J. Catal. 1980, 62, 127–139.
[61] V. I. Severing, A. V. Tseplayaeva, N. E. Khandamirova, Y. A. Priselkov, N. A.
Chernova, l. V. Golubtsov, V. B. Lukyanov, Mendeleev Commun. 1992, 2,
97–100.
Received: October 6, 2013
Revised: November 25, 2013
Published online on January 21, 2014
[62] R. F. Hampson, Jr., R. F. Walker, J. Res. Natl. Bur. Stand. Sect. A 1961, 65,
289–295.
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2014, 6, 885 – 894 894