Full Paper
(Sigma–Aldrich, 99%), urea (Grꢄnning, 99.5%), Ni(NO3)2·6H2O
(Acros Organics, ꢂ98.5%), stearic acid (Sigma–Aldrich, ꢂ99.5% an-
alytical standard), 1-octadecanol (Sigma–Aldrich, ꢂ99.5% Selecto-
phore), octadecane (Sigma–Aldrich, 99%), heptadecane (Sigma–Al-
drich, 99%), dodecane (Sigma–Aldrich, ꢂ99%, ReagentPlus), pro-
pionic acid (Sigma–Aldrich, ACS grade ꢂ99.5%) were purchased
commercially and were not further purified.
samples were then purged with He for 1 h in order to remove
physisorbed species. After activation, the six samples were heated
from 100–7708C with a rate of 108Cminꢀ1 to desorb NH3 and from
40 to 7008C to remove CO2, and the signals were detected by
a Balzers QME 200 mass spectrometer.
Temperature-programmed reactions (TPR) with H2 were performed
in a packed bed flow reactor equipped with a mass spectrometer.
First, calcined Ni/ZrO2 catalyst (100 mg, 250–400 mm) was activated
in a He flow at 2008C (heating rate of 108Cminꢀ1) for 30 min and
cooled to ambient temperature. The reduction was carried out
Catalyst preparation: Three types of ZrO2 supports were synthe-
sized. Mix-phase ZrO2 was prepared by calcination of Zr(OH)4·H2O
at 4008C in ambient air for 4 h. Monoclinic and tetragonal ZrO2
were prepared by the solvothermal method by mixing ZrO(NO3)2
with water and methanol, respectively.[17] An aqueous or methanol-
ic solution of ZrO(NO3)2 (0.6 molLꢀ1) was added with urea (urea/
Zr=10:1). The solvothermal reaction was performed in a stainless-
steel autoclave with a Teflon liner at 1608C and autogenous pres-
sure for 21 h. After washing five times the precipitate with H2O or
MeOH, it was dried over night at 1108C and then ground and cal-
cined in air at 4008C for 4 h at a heating rate of 28Cminꢀ1 (flow
rate: 100 mLminꢀ1).
from ambient temperature to 8008C (heating rate: 108Cminꢀ1
)
and maintaining 8008C for 30 min in 10% H2/He gas mixture
(2 mLminꢀ1 H2/18 mLminꢀ1 He). The amount of water produced in
the reaction was determined by an online mass spectrometer.
IR spectroscopy of adsorbed propionic acid was performed on
a Bruker VERTEX 70 spectrometer at a resolution of 2 cmꢀ1 with
128 scans in the range of n˜ =400–4000 cmꢀ1. For the measure-
ments, the samples were pressed into self-supporting wafers and
mounted in the sample holder. The ZrO2 samples were activated in
vacuum (p=10ꢀ7 mbar) at 3008C for 1 h. The Ni/ZrO2 catalysts
were activated in H2 at 4008C for 1 h, and then subsequently out-
gassed under vacuum (p=10ꢀ7 mbar) to remove H2 while cooling
to 408C. The adsorption of propionic acid was performed from
0.01 to 0.05 mbar until equilibrium was reached. In addition, the
effect of the temperature was investigated by heating the cell
stepwise up to 2508C. The IR spectra of adsorbed propionic acid
were obtained by subtracting the activated sample, and then were
normalized by the weight of the sample wafer.
The 10 wt% Ni/ZrO2 catalysts were prepared by impregnation.
Ni(NO3)2·6H2O (3.30 g) was dissolved in deionized H2O (5.0 g), and
the resulting solution was added dropwise to the support under
stirring in ambient air. The slurry was further stirred for 4 h, fol-
lowed by drying at 1108C overnight. Subsequently, the ground
solid was calcined in synthetic air (flow rate: 100 mLminꢀ1) at
4508C for 4 h (heating rate: 28Cminꢀ1) and reduced in a H2 flow
(flow rate: 100 mLminꢀ1) at 5008C for 4 h (heating rate: 28Cminꢀ1).
Catalyst characterization: X-ray powder diffraction (XRD) was per-
formed on Philips X’Pert Pro System equipped with a CuKa radia-
tion source (40 kV, 45 mA) with 1.088minꢀ1 in the 2q range of 5–
708. The ratio of monoclinic and tetragonal phases in mix-ZrO2 was
determined by using Equation (6) with the integrated intensities of
The near-edge structure (XANES) and extended X-ray absorption
fine-structure (EXAFS) measurements were performed in the trans-
mission mode at the Pacific Northwest Consortium/X-ray Science
Division (PNC/XSD) bending-magnet beamline at Sector 20 of the
Advanced Photon Source (APS) at Argonne National Laboratory
(ANL). Both Ni (8331.5 eV) and Zr (17995.88 eV) K-edge spectra
were acquired. A combination of monochromator detuning (10%)
and a harmonic rejection mirror placed upstream of the I0 detector
reduced contributions from higher harmonics. A Ni or Zr foil was
placed downstream of the sample cell as a reference to calibrate
the photon energy of each spectrum. Typically, two 15 min scans
(Ni edge spectra) and four 15 min scans (Zr edge spectra) were
averaged to generate the spectra. The catalyst samples were
ground and mixed with boron nitride (catalyst/boron nitride, 20:80
or 5:95 wt% for Ni and Zr edges, respectively), then pressed into
5ꢃ12 mm pellets (80 mg) and mounted onto a multiple sample
holder. The ATHENA software package[19] was used to remove the
background from the c(k) oscillations. The Fourier transform of the
k-space EXAFS data (both real and imaginary parts of c(R)) were
fitted to a theoretical model (FEFF9) calculated by using the ARTE-
MIS software package. A starting point for evaluating the nanopar-
ticle structure was the measurement of reference standards includ-
ing bulk (fcc) NiO, bulk (hcp) a-Ni(OH)2, and bulk (fcc) Ni by using
literature values for their lattice parameters.[20] A combination of
different single and multiple photoelectron scattering paths were
used to fit the first five shells of the NiO, a-Ni(OH)2, and Ni nano-
particles.[21] For samples containing both oxidation states, the
structural parameters were constrained and then the percentage
of each phase was fitted. As a starting point for modeling the ZrO2
nanoparticles, crystalline m-ZrO2 and bulk t-ZrO2 structures derived
from their lattice parameters were used.[22] Single-scattering paths
of Zr and O for the monoclinic and tetragonal phase were fitted
according to Rush et al.[23]
¯
the (111) and (111) reflecting monoclinic and tetragonal XRD pat-
terns, respectively.[18] The (111)m and (111)m reflections for the mon-
¯
oclinic phase are at 2q of 31.4 and 28.38, respectively, whereas the
(111)t reflection from the tetragonal phase is at 2q of 30.48. Accord-
ingly, the actual ratio of monoclinic to tetragonal ZrO2 in the physi-
cally mixed Ni/(m-ZrO2/t-ZrO2) was determined by using Equa-
tion (6) after fitting and integrating the corresponding peaks from
the XRD (Figures S1B and S2 in the Supporting Information).
ꢀ
Ið111ÞmþIð111Þm
ð6Þ
xm
¼
ꢀ
Ið111ÞmþIð111ÞmþIð111Þt
Atomic absorption spectroscopy (AAS) was used to determine the
Ni content of the catalysts with a UNICAM 939 AA spectrometer.
Prior to measurements, the samples were dissolved in boiling con-
centrated hydrofluoric acid.
The BET surface area was determined by adsorption–desorption
with nitrogen at ꢀ1968C by using a Sorptiomatic 1990 series in-
strument. The samples were activated in vacuum at 2508C for 2 h
before measurements.
he EDX mappings were obtained by using a JEM-ARM200CF oper-
ated at 200 KV with an integrated probe aberration (Cs) corrector
and a cold-field emission gun (CFEG) electron source After reduc-
tion the finely ground-powdered catalyst samples were stored and
mounted under an Ar atmosphere.
Temperature-programmed desorption (TPD) of ammonia and
carbon dioxide was carried out in a six-fold parallel reactor system.
The pressed samples (500–710 mm) were firstly activated in a He
flow at 5008C for 1 h and loaded with the adsorbent NH3 or CO2 at
a partial pressure of 1 mbar and 1008C or 408C, respectively. The
For the temperature-programmed isotope (18O2–16O2) exchange of
m- and t-ZrO2, the pelletized supports (100 mg, 500–710 mm) were
Chem. Eur. J. 2014, 20, 1 – 13
11
ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ