Search

Side-arm hydrogenation (SAH) by homogeneous catalysis has extended the reach of the parahydrogen enhanced NMR technique to key metabolites such as pyruvate. However, homogeneous hydrogenation requires rapid separation of the dissolved catalyst and purification of the hyperpolarised species with a purity sufficient for safe in-vivo use. An alternate approach is to employ heterogeneous hydrogenation in a continuous-flow reactor, where separation from the solid catalysts is straightforward. Using a TiO2 -nanorod supported Rh catalyst, we demonstrate continuous-flow parahydrogen enhanced NMR by heterogeneous hydrogenation of a model SAH precursor, propargyl acetate, at a flow rate of 1.5 mL/min. Parahydrogen gas was introduced into the flowing solution phase using a novel tube-in-tube membrane dissolution device. Without much optimization, proton NMR signal enhancements of up to 297 (relative to the thermal equilibrium signals) at 9.4 Tesla were shown to be feasible on allyl-acetate at a continuous total yield of 33 %. The results are compared to those obtained with the standard batch-mode technique of parahydrogen bubbling through a suspension of the same catalyst.

The effects of low-valence dopants on the catalytic properties of samarium oxide xerogel catalysts were investigated in the oxidative coupling of methane (OCM). More specifically, very low concentrations (0.1 and 1.0 % by mol) of transition metal (Ag, Ni, and Cu) and traditional alkali metal (Li and K) dopants were investigated. At these low loadings, it was shown that transition metal dopants have potential to improve the activity and selectivity over an undoped Sm2O3 xerogel, but these dopants can only outperform alkali metal dopants under certain conditions. Even at a concentration of 0.1 mol %, the dopants significantly increased the number of basic sites compared with the pure Sm2O3 xerogel. However, no trend is evident between the number or strength of the basic sites and the activity or selectivity in the methane coupling reaction. The XRD data reveal a lattice expansion upon addition of the low valence dopants, which is consistent with substitutional doping and the formation of oxygen vacancies due to charge compensation. In most cases the majority of the dopant stayed in the lattice during reaction. The dopants were also shown to influence the Sm2O3 structure, and the dopants that were more effective in suppressing the transformation from cubic to monoclinic Sm2O3 in general resulted in the more active and selective catalysts. While the Ag- and Ni-dopants could outperform the alkali metal doped catalysts in narrow temperature ranges, the best performing catalysts were still the K-doped Sm2O3 catalysts, as the 1.0 % K catalyst exhibited the highest activity at the lowest temperature (500 °C) and the 0.1 % K-doped catalyst was the most stable during extended operation. These results indicate that transition metal dopants, at low concentrations, can positively affect the activity and selectivity of a methane coupling catalyst, such as Sm2O3, and suggests that there may be benefits to other OCM catalyst systems from traditionally non-selective dopants, as long as the concentrations are kept very low and stability issues are addressed.

The effects of high-valance dopants on the catalytic properties of samarium oxide xerogels were investigated at concentrations of 0.1 and 1.0 % (by mol) in the oxidative coupling of methane (OCM). Gd, Y, Zr, and V dopants were selected to examine the influence of oxidation states between +3 and +5 on the OCM performance. Even at these low loadings, the high-valance dopants were observed to have a significant impact on the OCM reaction. At the lowest loading, 0.1 mol %, and below 700 °C, all dopants improved the activity over that of the pure Sm2O3 xerogel, and most also improved the selectivity. In particular, the ethylene yield was significantly improved over these catalysts between 500 and 700 °C. However, the stability of the doped catalysts compared to the undoped Sm2O3 xerogel were inferior above 700 °C, and the higher concentration (1.0 mol %) resulted in catalysts with a lower stability. Time-on-stream experiments revealed that the 0.1 mol % high valence dopants improved the stability of the Sm2O3 xerogel at 700 °C. As a result of the higher stability, the doped catalysts retain more of the original specific surface area and appear to stabilize the more active cubic Sm2O3 phase compared with the undoped catalyst. Therefore, the doped catalysts have a higher number of available basic sites during reaction. This study reveals that high-valence dopants have potential to improve the low temperature (500–700 °C) activity of OCM catalysts. However, the concentrations must be kept very low, as dopants that increase the activity and selectivity at concentrations of 0.1 mol % can result in inferior catalysts at 1.0 mol %, and temperatures above 700 °C must be avoided or rapid deactivation can occur.

Oxide supports with well-defined shapes enable investigations on the effects of surface structure on metal-support interactions and correlations to catalytic activity and selectivity. Here, a modified atomic layer deposition technique was developed to achieve ultra-low loadings (8-16 ppm) of Pt on shaped ceria nanocrystals. Using octahedra and cubes, which expose exclusively (111) and (100) surfaces, respectively, the effect of CeO2 surface facet on Pt-CeO2 interactions under reducing conditions was revealed. Strong electronic interactions result in electron-deficient Pt species on CeO2 (111) after reduction, which increased the stability of the atomically dispersed Pt. This afforded significantly higher NMR signal enhancement in parahydrogen-induced polarization experiments compared with the electron-rich platinum on CeO2 (100), and a factor of two higher pairwise selectivity (6.1 %) in the hydrogenation of propene than any previously reported monometallic heterogeneous Pt catalyst.

The effects of transition metal dopants on the catalytic properties of magnesia-supported terbium oxide catalysts in the oxidative coupling of methane were investigated. A few transition metals were included based on their oxidation states relative to Tb in TbO1.81 (where Tb is between Tb3+ and Tb4+). Ni and Fe were selected as low valence dopants (LVDs) and Zr and V were considered high valence dopants (HVDs) in the study. The effect of dopant concentration was also investigated, between 0.2 and 1.0 % by weight (wt%), since high loadings could result in segregation of the transition metal (TM) dopants, while low loadings may not yield observable effects. At the higher loading (1.0 wt%) it appears that the oxidation state relative to that of Tb is of minor importance, and the redox properties of the dopant have a larger impact. At this loading, the best performing catalysts were the Ni- and Zr-doped TbOx/n-MgO, while the selectivities of the Fe- and V-doped catalysts toward OCM products were very low, particularly at the lower temperatures ( The characterization data revealed that even at the low concentration of 0.2 wt%, the transition metals markedly alters the properties of the doped TbOx/n-MgO catalysts. For example, the number of basic sites was significantly increased by adding 0.2 wt% of Fe or Zr to the TbOx/n-MgO catalyst and this led to an improvement in the performance of the catalysts in the OCM reaction. The dopants also affected the oxidation state of Tb in the TbOx lattice as well as the dispersion of TbOx on the surface of the MgO. Both XPS and XRD data indicated that the TbO1.81, which was present on the surface of the fresh catalysts, in most cases was reduced to Tb2O3 during reaction. However, there were no evident trends to indicate that LVDs promote, while HVDs suppress, surface reduction of the host oxide, i.e. TbOx. The lack of trends may in part be due to the presence of the MgO support, which results in a complex catalyst system. Nevertheless, the data revealed that even low loadings of a transition metal dopant can positively affect the activity and selectivity of a TbOx/n-MgO catalyst.

Several novel materials were synthesized by precipitating iron oxide (using the previously optimized 10% Fe loading by weight) onto mixtures of nanoparticle zirconia and ceria to investigate the effects of adding CeO2 to FeOx/ZrO2 materials in the thermochemical water splitting reaction. At water splitting temperatures of 1000 °C (after thermal reduction at 1450 °C), the stability of the CeO2-containing materials was lower than for the FeOx/ZrO2 material, and there was no advantage to adding CeO2 to the FeOx/ZrO2 material. However, when operating at a water splitting (WS) temperature of 1200 °C, the stability increased and the hydrogen production was significantly higher over most materials compared with a water splitting temperature of 1000 °C. At a WS temperature of 1200 °C the FeOx/Zr75Ce25O2 (75% Zr75O2 and 25% CeO2 by weight) and FeOx/Zr50Ce50O2 materials performed slightly better than the FeOx/ZrO2 material, and X-ray photoelectron spectroscopy data revealed that the surface concertation of iron is less important compared with water splitting at 1000 °C. The temperature programmed reduction data indicated that the FeOx-CeO2 interactions were weaker compared with FeOx-ZrO2 interactions, since the FeOx reduction occurred at lower temperatures for the CeO2-containing materials. The weaker interactions can explain why the stability was lower for the materials containing CeO2 (sintering of FeOx was likely more pronounced) The X-ray diffraction data revealed that ZrO2-CeO2 solid solutions formed after activation at 1450 °C and lattice volume calculations indicated that iron did incorporate into the ZrO2-CeO2 matrices. More incorporation was observed after water-splitting at 1200 °C compared with a lower temperature (1000 °C), and likely explains why the materials were more stable during water-splitting at 1200 °C.

The oxidative coupling of methane (OCM) to ethane and ethylene was studied over alkali metal (Li and Na) and alkaline earth metal (Mg and Ca) doped rare earth oxides (Sm2O3, TbOx, PrOy and CeO2) catalysts supported on nanoparticle magnesium oxide (n-MgO). It was found that the Li-TbOx/n-MgO catalyst outperformed all other combinations of rare earth oxide and dopant, in terms of initial activity and selectivity, except at temperatures below 600 °C where the undoped Sm2O3/n-MgO, Ca-Sm2O3/n-MgO and Ca-CeO2/n-MgO catalysts resulted in higher C2+ yields. Li doping was most efficient in yielding C2+ products over all rare earth oxides, followed by doping with Na. These dopants result in the most basic sites on the surface according to CO2 desorption experiments. However, these sites are blocked at lower reaction temperatures, which can explain why most Li- and Na-doped catalysts are not active below 650 °C. Below reaction temperatures of 650 °C moderately basic sites, which bind CO2 up to ∼350 °C, are important and correlate well with OCM activity. Li doping also resulted in the highest ethylene yields, which is a more desirable product than ethane, and the highest ethylene yield (9.7%) was obtained over the Li-CeO2/n-MgO (at 800 °C). However, Li-doped catalysts are not very stable, and the Na-TbOx/n-MgO catalyst therefore outperformed the Li-TbOx/n-MgO catalyst after only a few hours on stream. Removal of both Li and Na during reaction was also confirmed and is probably the reason for the loss in selectivity with time on stream. Tb2O3 is the only TbOx phase observed with XRD on the Na- and Li-doped TbOx/n-MgO catalysts after reaction and this is important since Tb2O3 is likely more selective to C2+ products than TbO1.81 in the OCM reaction.

The pairwise selectivity of hydrogenation is a fundamental quantity in hydrogenation catalysis that underpins the NMR signal enhancement achievable by parahydrogen-induced polarization (PHIP). Herein, we show how the crystal facet dependence of the pairwise selectivity of the semihydrogenation of propyne, when interpreted in the context of recent DFT calculations, can reveal new details about the hydrogenation mechanism. The pairwise selectivity of propyne hydrogenation is strongly shape dependent, which reflects the surface atom arrangements exposed on the different facets of CeO2 nanocrystals. In this first demonstration of a catalyst shape dependence of PHIP, an unprecedented pairwise selectivity of 8.1 % is observed over oxygen-deficient facets of rods, whereas oxygen-rich octahedra facets deliver only 1.6 % selectivity. The PHIP data are consistent with a concerted addition pathway through a six-membered-ring transition state, as predicted in the DFT study.

The effects of strong metal–support interactions (SMSI) on the pairwise selectivity of propene hydrogenation over metal-oxide-supported Ir nanoparticles were investigated using parahydrogen-enhanced NMR spectroscopy. A ∼20-fold increase in the pairwise selectivity was observed following a reduction treatment of the Ir/TiO2 catalyst at 500 °C. Consistent with SMSI, the effects could be completely reversed by oxidation followed by rereduction at 200 °C. Noninteracting supports, such as Al2O3 and SiO2, did not show this behavior. X-ray photoelectron spectroscopy reveals partial reduction of the TiO2 support, and STEM data reveal flattening of Ir particles after high-temperature reduction. The presence of chloride ions during activation was found to further promote pairwise selectivity but only for the Ir/TiO2 catalyst. The results are interpreted in terms of the electronic and possible geometric blocking effects associated with SMSI.

Two-step solar thermochemical water splitting offers a viable solution for employing renewable energy pathways for production of high-purity hydrogen, which can be used in fuel cells or for the production of liquid fuels. One promising thermochemical cycle uses iron oxide as the reactive material to split water in a relatively simple two-step process. However, this cycle requires high temperatures, on the order of 1500 °C. Therefore, to increase stability and thermal shock resistance, iron oxide was deposited onto high-temperature-stable ceramic oxides, such as zirconia (ZrO2) and yttria-stabilized zirconia (YSZ). Nanoparticle zirconia samples (n-ZrO2 and n-YSZ) were used in this study to increase the interactions between the iron oxide and the zirconia, and the resulting materials were compared with an iron oxide supported on a commercially available porous zirconia support (p-ZrO2). After heat treatment, these FeOx/n-ZrO2 and FeOx/n-YSZ materials were surprisingly stable and yielded consistent hydrogen production over several cycles. The hydrogen yield per cycle for the FeOx/n-YSZ material was consistently higher for all iron loadings compared with the FeOx/n-ZrO2 materials, and they both yielded significantly more hydrogen compared to the FeOx supported on the commercial ZrO2 support. Compared with the FeOx/n-YSZ, the iron loading had a larger effect on the hydrogen production over n-ZrO2-supported materials. X-ray diffraction (XRD) data indicated that iron was incorporated into the unit cell lattice of the n-YSZ supported material, which alleviated some of the detrimental iron oxide sintering inherent at the high-temperatures of the thermochemical water splitting reaction, and resulted in a more consistent and higher hydrogen production compared with the FeOx/n-ZrO2 material.

The effects of anatase and rutile titania on PdO/TiO 2 catalysts and how these phases influence the activity in the oxidative coupling of 4-methylpyridine to 4,4′-dimethyl-2,2′-bipyridine were investigated. The anatase and rutile concentrations were systematically varied by heat treating a nearly amorphous nanoparticle titania at temperatures between 105 and 800 °C. X-ray diffraction (XRD) measurements confirmed that the anatase concentration increased with heat treatment temperature until rutile began to form at 600 °C, and at 800 °C rutile TiO 2 was the only phase observed. X-ray photoelectron spectroscopy (XPS) analysis of the catalysts revealed strong PdO-TiO 2 interactions which increased with support treatment temperature. According to XPS, the interactions yielded electron-deficient PdO species (Pd δ+ with δ > 2) and Ti 3+ species (likely TiO 2 with oxygen vacancies) on the surface of the catalysts, as a result of electron transfer from palladium to titania. These PdO-TiO 2 interactions increased with increasing anatase phase in the TiO 2 support and the turnover number (TON: number of product molecules formed per surface palladium site) followed the same trend. The strongest interactions (the largest amounts of Pd δ+ with δ > 2 and Ti 3+ species) were observed for the catalysts supported on rutile TiO 2 , which is also the catalyst with the highest TON, revealing the importance of electron transfer from palladium to titania in this reaction. Despite the high TON for PdO on rutile TiO 2 , the anatase-supported PdO gave a significantly higher product yield due to the very low surface area of rutile TiO 2 .

Intense para-hydrogen-enhanced NMR signals are observed in the hydrogenation of propene and propyne over ceria nanocubes, nano-octahedra, and nanorods. The well-defined ceria shapes, synthesized by a hydrothermal method, expose different crystalline facets with various oxygen vacancy densities, which are known to play a role in hydrogenation and oxidation catalysis. While the catalytic activity of the hydrogenation of propene over ceria is strongly facet-dependent, the pairwise selectivity is low (2.4% at 375 °C), which is consistent with stepwise H atom transfer, and it is the same for all three nanocrystal shapes. Selective semi-hydrogenation of propyne over ceria nanocubes yields hyperpolarized propene with a similar pairwise selectivity of (2.7% at 300 °C), indicating product formation predominantly by a non-pairwise addition. Ceria is also shown to be an efficient pairwise replacement catalyst for propene.

A new sol–gel synthesis route for alumina–samaria mixed aero- and xerogel catalysts based on the so-called epoxide addition method and the use of these systems as catalysts for the oxidative coupling of methane (OCM) is reported. As precursors simple chloride or nitrate salts can be used. The mesoporous materials are X-ray amorphous even after calcination to 800 °C and show an intimate mixing of Al and Sm on the nanoscale. In the case of the xerogels derived from chlorides, C2 yields comparable to pure samaria can be achieved under OCM reaction conditions with 100 % O2 conversion. Even at lower O2 conversions the activity of the xerogel is competitive with a pure samaria reference catalyst taking the lower samaria content of 20 % into account. Accordingly, the approach is suitable to reduce the costs associated with the rare earth oxide. In addition to the preparation of aerogel and xerogel particles, the presented synthesis also allows the fabrication of xerogel films which can be coated on a suitable (monolithic) support. First results of such films are presented.

In this study, one of the most active catalysts in the oxidative coupling of methane (OCM), Mn–Na2WO4/SiO2, was investigated in detail using X-ray diffraction (XRD) and X-ray photoelectron spectroscopy (XPS) to obtain more information about this complex catalyst. A poorly active Mn–Na2WO4/MgO and two reference catalysts, Mn/SiO2 and Na2WO4/SiO2, were also included. XRD reveals the presence of Na2WO4, Mn2O3 and the α-cristobalite phase of SiO2 on the active Mn–Na2WO4/SiO2 catalyst, which all have been identified as very important in the oxidative coupling of methane (OCM). In addition to facilitating the transformation from amorphous to crystalline α-cristobalite SiO2, the presence of Na2WO4 also yields a higher surface concentration of manganese. While this is necessary, it is not sufficient, for a highly active catalyst, as the poorly active Mn–Na2WO4/MgO also has a high manganese surface concentration. However, in contrast to the very stable Mn–Na2WO4/SiO2 catalyst, the MgO support is not able to stabilize the specific Na2WO4 and Mn2O3 phases and a number of NaxWOy compounds with varying Na:W ratios together with MnO2 and mixed manganese compounds are identified by XRD. The broad Na 1s and W 4f XPS peaks support the presence of more than one NaxWOy compound on the surface of the Mn–Na2WO4/MgO catalyst. The fact that Na2WO4 and Mn2O3 are not stable on the MgO-supported catalyst is likely one reason for its poor activity, as the active site is believed to involve a distorted WO4 tetrahedron stabilized by Na (and α-cristobalite SiO2), and accessible Mn3+ in close proximity to the Na2WO4 may be necessary to yield a highly active catalyst. The Na:W surface concentration ratio of the Mn–Na2WO4/SiO2 catalyst is significantly higher (5.8) than the stoichiometric Na:W ratio of Na2WO4 (2.0) revealing a surface enriched in sodium. This is due to some of the sodium interacting closely with the SiO2 as evidenced by shifts in the Na 1s and Si 2p peaks to lower binding energies than expected for SiO2 and Na2WO4. The Na may be important in the activation of oxygen species on the surface toward hydrogen abstraction. The importance of the specific Na2WO4 oxide is further supported by the H4Na2WO6 phase which is observed on the Mn–Na2WO4/SiO2 catalyst after reaction. This phase has not been identified previously on the Mn–Na2WO4/SiO2 catalyst, but corroborates earlier conclusions that Na2WO4 is important and suggests that Na2WO4 is indeed involved in the hydrogen abstraction from methane, the important first step in the oxidative coupling of methane. The strong interactions observed between all the species (Mn, Na, W and Si) on this catalyst may also indicate that the interface between Na2WO4 and Mn2O3 is important in this reaction.

Pairwise and random addition processes are ordinarily indistinguishable in hydrogenation reactions. The distinction becomes important only when the fate of spin correlation matters, such as in parahydrogen-induced polarization (PHIP). Supported metal catalysts were not expected to yield PHIP signals given the rapid diffusion of H atoms on the catalyst surface and in view of the sequential stepwise nature of the H atom addition in the Horiuti-Polanyi mechanism. Thus, it seems surprising that supported metal hydrogenation catalysts can yield detectable PHIP NMR signals. Even more remarkably, supported Pt and Ir nanoparticles are shown herein to catalyze pairwise replacement on propene and 3,3,3-trifluoropropene. By simply flowing a mixture of parahydrogen and alkene over the catalyst, the scalar symmetrization order of the former is incorporated into the latter without a change in molecular structure, producing intense PHIP NMR signals on the alkene. An important indicator of the mechanism of the pairwise replacement is its stereoselectivity, which is revealed with the aid of density matrix spectral simulations. PHIP by pairwise replacement has the potential to significantly diversify the substrates that can be hyperpolarized by PHIP for biomedical utilization.

Parahydrogen induced polarization using heterogeneous catalysis can produce impurity-free hyperpolarized gases and liquids, but the comparatively low signal enhancements and limited scope of substrates that can be polarized pose significant challenges to this approach. This study explores the surface processes affecting the disposition of the bilinear spin order derived from parahydrogen in the hydrogenation of propyne over TiO2-supported Pt nanoparticles. The hyperpolarized adducts formed at low magnetic field are adiabatically transported to high field for analysis by proton NMR spectroscopy at 400 MHz. For the first time, the stereoselectivity of pairwise addition to propyne is measured as a function of reaction conditions. The correlation between partial reduction selectivity and stereoselectivity of pairwise addition is revealed. The systematic trends are rationalized in terms of a hybrid mechanism incorporating non-traditional concerted addition steps and well-established reversible step-wise addition involving the formation of a surface bound 2-propyl intermediate.

Being the shortest unsaturated hydrocarbon, ethylene is a valuable feedstock gas for synthesizing longer chain hydrocarbon products. Using the oxidative coupling of methane to produce ethylene and ethane (C2 products) has been studied extensively over the past few decades. In this work, samaria nanoparticles (NPs) and alumina-supported samaria catalysts were prepared using different methods (water/toluene reverse microemulsion, metal-oleate high temperature decomposition, and incipient-wetness impregnation followed by calcination) and various types of alumina supports (high and low surface area nanoparticle alumina [n-Al2O3(+), n-Al2O3(−)] and a porous gamma alumina support [p-Al2O3]). The highest product yields were obtained over Sm2O3 NPs synthesized using the metal-oleate high temperature decomposition and a nitrate precursor. Using a chloride precursor in the preparation of the Sm2O3 NPs resulted in less active and selective catalysts, and should be avoided. While the C2 selectivities were lower over the Sm2O3/Al2O3 catalysts, the yield per gram of Sm2O3 were higher compared with the Sm2O3 NPs. The best supported catalyst was the Sm2O3/n-Al2O3(−) prepared using incipient wetness impregnation of a nitrate precursor, since this leads to the smallest Sm2O3 particles as well as the largest coverage of the acidic Al2O3 support. XRD analysis revealed that high-temperature calcinations form SmAlO3 on the Al2O3-supported catalysts. While this reduced the near surface Sm2O3 concentration, it had beneficial effects as it also lowered the Al2O3 content. Therefore, Al2O3-supported Sm2O3 warrants further investigation, as surface modifications of the Al2O3 can reduce its acidity and lead to higher C2 selectivities in the oxidative coupling of methane over these catalysts.

Monolithic reactor concepts are currently intensively discussed in the literature. For the oxidative coupling of methane relying on a balance of surface and gas phase reactions, such concepts have previously been claimed to have beneficial effects with respect to obtainable C2 yields. In order to verify the superior performance in the case of a foam catalyst, ceria and samaria foams were fabricated by a direct foaming process. In both cases, mechanically stable, homogeneous open-cell foams were obtained as revealed by 3D magnetic resonance imaging, Hg-porosimetry, and scanning electron microscopy. As a characteristic feature of the introduced foaming methodology the process resulted in bimodal pore size distributions, ensuring low pressure drops on the one hand and sufficiently large surface areas on the other hand. Oxidative coupling of methane was carried out over the samaria foams. It was possible to obtain C2 yields that were indeed higher than those obtained with the samaria powder, in contrast to honeycomb monoliths previously studied in the literature.

4,4′-Dimethyl-2,2′-bipyridine is a useful but expensive chelating agent. Having more efficient routes to the synthesis of this compound would be advantageous to the wide-spread use of this fine chemical. In this work, the effects of adding strongly interacting oxides (ZnO, CeO 2 , and ZrO 2 ) to PdO catalysts supported on high surface area n-Al 2 O 3 (+), n-MgO, and n-TiO 2 prepared via co- and sequential precipitation were investigated. The product yields obtained from these catalysts in the oxidative coupling of 4-methylpyridine are dependent on the additive, the support, and preparation method. Evidently, these are complex catalytic systems in that the PdO–additive and PdO–support interactions must be right to promote product formation while preventing palladium leaching and support or additive migration over the active Pd/PdO sites. Although PdO/n-ZnO catalysts are reasonably active in the coupling reaction, ZnO addition to PdO catalysts supported on n-Al 2 O 3 (+), n-MgO, or n-TiO 2 does not increase the yield in any case. CeO 2 and ZrO 2 can increase the product yields in the reaction depending on the support used. Due to strong PdO–CeO 2 interactions, the addition of CeO 2 in some cases results in CeO x -migration and coverage of active PdO species or disrupts favorable PdO–support interactions leading to Pd leaching. Therefore, ZrO 2 is the better additive with co-precipitated PdO/ZrO 2 /n-Al 2 O 3 (+) consistently producing yields in excess of 3.4 ± 0.1 g/g catalyst which is 36% higher than the 2.5 ± 0.16 g/g catalyst obtained from the PdO/n-Al 2 O 3 (+) (5 wt% Pd), the best catalyst previously reported for this reaction.

A series of ex situ techniques, including scanning electron microscopy, transmission electron microscopy, X-ray photoelectron spectroscopy, X-ray diffraction, and inductively coupled plasma-mass spectroscopy were used to study morphological and chemical changes associated with the aging of a membrane electrode assembly in a polymer-electrolyte-membrane fuel cell. These results were correlated with the results of in situ electrochemical measurements, including measurement of steady-state polarization curves and electrochemical impedance spectroscopy. The results support the premise that the low-frequency inductive features seen in the impedance response provide information useful for understanding phenomena, such as platinum oxidation, that lead to reduction of fuel cell performance. The reduction in electrochemically active surface area, obtained from the high-frequency part of the impedance response, was consistent with the observed agglomeration of platinum particles and platinum oxidation.

Palladium catalysts supported on various metal oxides were characterized using X-ray photoelectron spectroscopy (XPS), X-ray diffraction (XRD) and transmission electron microscopy (TEM) to investigate why these catalysts do not show any correlation between the measured Pd surface area and the catalytic activity for the oxidative coupling of 4-methylpyridine to 4,4′-dimethyl-2,2′-bipyridine. The XPS data confirm the classification of n-Al 2 O 3 (+), n-MgO and p-SiO 2 as non-interacting supports, since the Pd 3d 5/2 binding energy (BE) of palladium on these supports is 336.1 eV, consistent with bulk PdO. In contrast, catalysts supported on p-TiO 2 , n-ZnO, n-ZrO 2 , n-ZrO 2 (CeO 2 ), and n-CeO 2 have Pd 3d 5/2 BEs ranging from 336.3 to 337.4 eV, which reveal varying degrees of metal-support interactions. Metal support interactions leading to electron deficient Pd 2+ species are likely beneficial for the reaction due to a facilitated C–H insertion step. While both the PdO/p-TiO 2 and PdO/n-TiO 2 catalysts have a Pd 3d BE of 336.3 eV, their differences in activity can be attributed to (1) the PdO/n-TiO 2 catalyst as prepared having a significantly higher number of hydroxyl groups on the surface compared with the PdO/p-TiO 2 catalyst, and (2) the p-TiO 2 support being crystalline with an anatase phase, while the n-TiO 2 support is nearly amorphous. The presence of surface hydroxyl groups before reaction could hinder the first C–H activation step, and an anatase phase of the support can result in more favorable palladium-support interactions compared with an amorphous TiO 2 . The XPS data also indicates that while Pd-support interactions are beneficial, very strong interactions, such as in the case of CeO 2 , can lead to migration of the support over Pd, which reduces the Pd surface area and explains the lower than expected activity of the PdO/n-CeO 2 catalyst. On some supports in this study leaching may occur during the reaction, but the characterization data indicate that other factors of catalyst deactivation are more important. XRD reveals that the complete reduction of the PdO particles on the surface is very fast compared to the reaction time. This observation explains why reducible supports with mobile oxygen are beneficial in this reaction. These supports can facilitate the reoxidation of palladium due to strong metal-support interactions. Migration of the support over the active palladium species is another deactivation pathway that appears to be more severe than leaching on these catalysts.

Three Cu/ZnO/ZrO2/Al2O3 methanol reforming catalysts were investigated using X-ray photoelectron spectroscopy (XPS). The catalysts which contained ZrO2 from a monoclinic nanoparticle ZrO2 precursor exhibit both a higher activity toward the methanol steam reforming reaction and a lower CO production rate compared to catalysts composed of an XRD-amorphous ZrO2 produced by impregnation using a Zr(NO3)2 precursor. The presence of a monoclinic phase appears to result in an increased charge transfer between the Zr and Cu species, as evidenced by a relatively electron-rich ZrO2 phase and a partially oxidized Cu species on reduced catalysts. This electron deficient Cu species is more reactive toward the methanol reforming reaction and partially suppresses CO formation through the reverse water gas shift or methanol decomposition reactions.

A number of palladium and palladium oxide on alumina catalysts for the oxidative coupling of 4-methylpyridine were characterized using X-ray photoelectron spectroscopy (XPS), transmission electron microscopy (TEM) and X-ray diffraction (XRD) to obtain more information about the properties that govern these reactions. The most active catalysts, the impregnated and precipitated PdO/n-Al2O3(+), were found to have very broad Pd 3d peaks, suggesting the presence of both PdO2 and Pd0, in addition to PdO, on the surface. This could be due to strong metal–support interactions, which result in more electrophilic palladium or a reduced palladium catalyst that is easier to reoxidize, both of which are expected to result in high catalytic activities. As the PdO is completely reduced to Pd metal on spent catalysts (according to XPS and XRD), facilitated reoxidation could be important in this reaction. While reoxidation is a potential deactivation pathway, carbon deposition is also evident in the XPS spectra and this could block active sites. In contrast, both the XPS and the XRD data indicate that Pd leaching into the reaction solution is not a significant deactivation pathway in the PdO/n-Al2O3(+) catalyst system. The nm-sized fine structure observed in the supports of the PdO/γ-Al2O3 and PdO/n-Al2O3(+), together with their measured activities compared to the poorly active PdO/n-Al2O3(−) catalyst with no such fine structure, supports our hypothesis that low-coordination sites on a support can result in strong metal–support interactions and very active catalysts. Highly active crystalline PdO particles with sizes of 5 nm or below may explain the observed lack of correlation between the measured Pd surface area (which should correlate with the PdO surface area) and the catalytic activity.

The oxidative coupling of 4-methylpyridine to 4,4′-dimethyl-2,2′-bipyridine over palladium oxide is a simple, environmentally friendly, one-step process to produce bipyridines, which are commonly used with transition metal ions to form complexes with interesting properties. However, the reaction is slow and the palladium catalyst deactivates during reaction, which means that catalyst improvements are needed for large-scale production of more economically viable bipyridine products. In this study, a number of metal oxides were investigated as catalyst supports and compared to the best performing catalysts to date, i.e. Pd/C and Pd/n-Al2O3(+). Catalysts supported on several nanoparticle oxides with varying properties as well as some conventional supports were prepared and characterized in an attempt to determine properties that lead to high catalytic activities in the oxidative coupling of 4-methylpyridine. It was found that two general categories of active catalysts can be prepared; (1) palladium supported on very high surface area materials, such as Pd/n-Al2O3(+) and Pd/MgO, and (2) palladium supported on metal oxides known to induce strong palladium-support interactions, e.g. Pd/ZrO2, Pd/(n-ZrO2 + n-CeO2) and Pd/n-ZnO. While there is no simple correlation between the palladium surface area and the catalytic activity, higher palladium dispersions generally gave higher yields compared to lower dispersion catalysts. The results indicate that the reaction is structure sensitive, i.e. not all the palladium on the surface is equivalent and some palladium species are more active than others. The acidic and basic properties of the supports were determined via chemisorption of ammonia and carbon dioxide, respectively. The data indicate that there is no correlation between the acidic or basic sites of the supports and the palladium dispersion or the catalytic activity, although highly acidic or highly basic supports should be avoided as they resulted in lower dispersions than expected from their corresponding surface areas. In terms of economic viability the porous TiO2 support was determined to be the most competitive with the nanoparticle alumina support as it results in a catalyst with comparable yields and is less expensive compared with nanoparticle alumina. The palladium supported on nanoparticle ZrO2 and MgO are also promising catalysts.

Methanol steam reforming was studied over several ZrO2- and CeO2-promoted catalysts prepared by deposition of copper and zinc oxide precursors onto nanoparticle Al2O3 and ZrO2 supports. The catalysts were characterized using temperature programmed reduction (TPR), Brunauer–Emmett–Teller (BET) surface area measurements, N2O decomposition, and X-ray diffraction (XRD) and the results were compared to a commercially available Cu/ZnO/Al2O3 catalyst. It was found that the catalyst activities and CO selectivities are very dependent on both the catalyst preparation method and the ZrO2 precursor. The best performing catalyst was prepared by co-impregnating Cu and ZnO precursors onto a physical mixture of nanoparticle ZrO2 and nanoparticle Al2O3. This catalyst resulted in a highly active reforming catalyst which also suppressed CO production. The turnover frequencies of catalysts containing ZrO2 nanoparticles were notably higher than those of similar catalysts which contained impregnated ZrO2, and dramatically higher than the commercially available reference catalyst. This indicates that the presence of ZrO2 nanoparticles promotes a highly active copper surface. Addition of high surface area alumina was necessary to assure a reasonable Cu surface area. However, the catalytic activities of the catalysts in this study did not correlate with Cu surface area, total surface area, or reduction temperature. It is therefore postulated that the presence of a monoclinic ZrO2 phase, which is a result of using a monoclinic nanoparticle ZrO2 precursor, promotes methanol reforming and also suppresses methanol decomposition, both of which are desired in methanol reforming catalysts.

Methanol steam reforming was studied over several catalysts made by deposition of copper and zinc precursors onto nanoparticle alumina. The results were compared to those of a commercially available copper, zinc oxide and alumina catalyst. Temperature programmed reduction, BET surface area measurements, and N2O decomposition were used to characterize the catalyst surfaces. XRD was used to study the bulk structure of the catalysts, and XPS was used to determine the chemical states of the surface species. The nanoparticle-supported catalysts achieved similar conversions as the commercial reference catalyst but at slightly higher temperatures. However, the nanoparticle-supported catalysts also exhibited a significantly lower CO selectivity at a given temperature and space time than the reference catalyst. Furthermore, the turnover frequencies of the nanoparticle-supported catalysts were higher than that of the commercial catalyst, which means that the activity of the surface copper is higher. It was determined that high alumina concentrations ultimately decrease catalytic activity as well as promote undesirable CH2O formation. The lower catalytic activity may be due to strong Cu-Al2O3 interactions, which result in Cu species which are not easily reduced. Furthermore, the acidity of the alumina support appears to promote CH2O formation, which at low Cu concentrations is not reformed to CO2 and H2. The CO levels present in this study are above what can be explained by the reverse water-gas-shift (WGS) reaction. While coking is not a significant deactivation pathway, migration of ZnO to the surface of the catalyst (or of Cu to the bulk of the catalyst) does explain the permanent loss of catalytic activity. Cu2O is present on the spent nanoparticle catalysts and it is likely that the Cu+/Cu0 ratio is of importance both for the catalytic activity and the CO selectivity.

The C–H activation and C–C coupling of 4-methyl pyridine to 4,4′-dimethyl-2,2′-bipyridine was studied over palladium catalysts supported on nanoparticle alumina [nano-Al 2 O 3 (+)]. Even though Pd/Al 2 O 3 catalysts are traditionally poor catalysts in this reaction, the Pd/nano-Al 2 O 3 (+) catalysts prepared via the precipitation method give the highest yield observed to date for this reaction system. The catalytic activity is very dependent on the catalyst preparation as Pd/nano-Al 2 O 3 (+) catalysts prepared via the wet impregnation method exhibit poor activities in the reaction. Additionally, a catalyst prepared using another nanoparticle alumina with larger particle sizes [nano-Al 2 O 3 (−)] does not have a significant activity. It was shown that using a commercial bimodal γ-Al 2 O 3 support can result in an active catalyst if prepared via the precipitation method, but compared to the best performing catalyst, Pd/nano-Al 2 O 3 (+), the yields obtained from Pd/bimodal-γ-Al 2 O 3 are lower and less reproducible. Pd surface area measurements indicate that the reaction is structure sensitive as there is no correlation between the Pd surface area and the catalytic activity. The reaction is also very sensitive to reactant quality and the 4-methyl pyridine must be distilled over KOH to ensure reproducible yields. Additional experiments indicate that this reaction requires a solid phase for catalysis.

Intense para-hydrogen-enhanced NMR signals are observed in the hydrogenation of propene and propyne over ceria nanocubes, nano-octahedra, and nanorods. The well-defined ceria shapes, synthesized by a hydrothermal method, expose different crystalline facets with various oxygen vacancy densities, which are known to play a role in hydrogenation and oxidation catalysis. While the catalytic activity of the hydrogenation of propene over ceria is strongly facet-dependent, the pairwise selectivity is low (2.4 % at 375 °C), which is consistent with stepwise H atom transfer, and it is the same for all three nanocrystal shapes. Selective semi-hydrogenation of propyne over ceria nanocubes yields hyperpolarized propene with a similar pairwise selectivity of (2.7 % at 300 °C), indicating product formation predominantly by a non-pairwise addition. Ceria is also shown to be an efficient pairwise replacement catalyst for propene.

Electron energy loss spectra (ELS) have been obtained from polycrystalline Cr and Cr2O3 before and after surface reduction by 2 keV Ar+ bombardment. The primary electron energy used in the ELS measurements was systematically varied from 100 to 1150 eV in order to distinguish surface versus bulk loss processes. Two predominant loss features in the ELS spectra obtained from Cr metal at 9.0 and 23.0 eV are assigned to the surface and bulk plasmon excitations, respectively, and a number of other features arising from single electron transitions from both the bulk and surface Cr 3d bands to higher-lying states in the conduction band are also present. The ELS spectra obtained from Cr2O3 exhibit features that originate from both interband transitions and charge-transfer transitions between the Cr and O ions as well as the bulk plasmon at 24.4 eV. The ELS feature at 4.0 eV arises from a charge-transfer transition between the oxygen and chromium ions in the two surface layers beneath the chemisorbed oxygen layer, and the ELS feature at 9.8 eV arises from a similar transition involving the chemisorbed oxygen atoms. The intensity of the ELS peak at 9.8 eV decreases after Ar+ sputtering due to the removal of chemisorbed oxygen atoms. Sputtering also increases the number of Cr2+ states on the surface, which in turn increases the intensity of the 4.0 eV feature. Furthermore, the ELS spectra obtained from the sputtered Cr2O3 surface exhibit features characteristic of both Cr0 and Cr2O3, indicating that Ar+ sputtering reduces Cr2O3. The fact that neither the surface- nor the bulk-plasmon features of Cr0 can be observed in the ELS spectra obtained from sputtered Cr2O3 while the loss features due to Cr0 interband transitions are clearly present indicates that Cr0 atoms form small clusters lacking a bulk metallic nature during Ar+ bombardment of Cr2O3.

Electron energy loss spectra (ELS) obtained from palladium, gold and a palladium–gold (69–31 at.%) alloy as a function of primary beam energy (100–1150 eV) are presented in this study. The spectra are complex, so unambiguous identification of the features is difficult. Nevertheless, some interband transitions and bulk and surface features can be identified. The objective of this study is to test the efficacy for ELS to detect electronic interactions between gold and palladium in the alloy. This was done by comparing the experimental alloy ELS spectra with ELS spectra calculated from the pure metal spectra. These sets of spectra are quite similar indicating that the electronic interaction between alloyed Pd and Au is small. This is expected since both Au and Pd are not very chemically reactive. The small differences apparent in the comparison provide information about the nature of the electronic interaction which is consistent with data obtained in photoemission studies.

Electron energy loss spectroscopy (ELS) spectra have been obtained from polycrystalline Ni, Cr and a Ni–Cr alloy (nichrome) using primary electron energies ranging from 100 to 1150 eV. Loss features due to the surface and bulk plasmon excitations are present in the ELS spectra from each surface as well as a number of smaller features that arise from interband transitions. The ELS spectra obtained from nichrome cannot be reproduced by superposing the pure metal ELS spectra in proportion to the atomic fractions of Ni and Cr in the alloy. The main differences between the calculated and the experimental alloy spectra are an overestimation of the contributions from the plasmons with a concomitant underestimation of the interband transitions in the superimposed spectra. This suggests that the probability for surface plasmon excitation is lower for the Ni–Cr alloy than for the pure metals. The increased contribution from interband transitions results in more Ni-like alloy spectra, since Ni has a higher number of d electrons that can give rise to interband transitions in the energy-loss region under investigation. However, the interband transitions in the alloy are shifted relative to the pure Ni spectra indicating that alloy-specific states have been created. These findings reveal that the electronic structure of the Ni–Cr alloy is altered by chemical bonding interactions between the Ni and Cr atoms demonstrating the usefulness of ELS for this application.

X-ray photoelectron spectroscopy (XPS) and electron energy loss spectroscopy (ELS) have been used to study a sputter-cleaned polycrystalline Co metal foil and CoO plus Co 3 O 4 powders before and after ion bombardment. The ELS data have been collected using primary beam energies ranging from 100 to 1150 eV in order to distinguish between surface and bulk features and to facilitate determination of peak locations. It is shown that ELS spectra from these reference materials are considerably different and should, consequently, allow species identification in future catalyst characterization studies. The main surface features obtained from Co metal are observed at 3.7 and 7.0 eV, while the major bulk energy losses are located at 8.2, 9.8, 23.1, 25.7, and 28.1 eV. Despite the collected data on Co ELS and the electronic structure of Co metal, unambiguous peak assignments are difficult. Interband transitions in the energy loss region under investigation originate from the relatively localized Co 3d filled states, but the limited information about the empty states in the conduction band make specific assignments of the energy loss features in the Co ELS spectra challenging. XPS analysis of as-entered CoO and Co 3 O 4 reveals that both samples consist mainly of Co 3 O 4 , in agreement with a Co 3 O 4 -terminated CoO surface. After prolonged Ar + sputtering both oxides are CoO-like according to high-resolution spectra of the Co 2p binding energy region. In agreement with the XPS data the ELS spectra obtained from as-entered CoO and Co 3 O 4 are very similar. In contrast, ELS spectra obtained from CoO and Co 3 O 4 after prolonged Ar + sputtering are considerably different. The ELS spectra obtained from CoO are not altered significantly with sputtering suggesting that the Co 3 O 4 layer cannot be removed using Ar + bombardment. The sputtered Co 3 O 4 instead gives rise to ELS spectra that appear CoO-like. The major ELS features in the spectra obtained from Co 3 O 4 are located at 2.5, 6.7, 22.1, and 25.0 eV, while those obtained from Ar + -sputtered Co 3 O 4 (CoO) are observed at 7.0, 8.6, 18.3, 21.9, and 36.1 eV. The relative intensities of the features at 6–8 eV and 22–25 eV are considerably different in the two sets of spectra. The intensity ratio of the peaks at 6.7 and 25.0 eV in Co 3 O 4 is significantly higher than the 7.0 to 21.9 eV peak intensity ratio of CoO. Although several of the energy losses in the two spectra overlap and make identification difficult despite the large intensity difference, there are peaks that are specific for each compound, i.e. energy losses at 2.5 and 28.5 eV in Co 3 O 4 and at 36.1 eV in CoO. Consequently, ELS can be used to differentiate the two cobalt oxides and information complementary to XPS data can be obtained from the ELS data.

Electron energy loss spectra (ELS) have been obtained from Ni metal, NiO and NiO after surface reduction by 2 keV Ar + bombardment. Two predominant loss features in the ELS spectra obtained from Ni metal are assigned as the surface and bulk plasmon excitations, and numerous other features arising from single electron transitions from both the bulk and surface Ni 3d bands to higher-lying conduction bands are also present. The surface and bulk ELS features have been unambiguously identified in this study by varying the primary beam energy. The peak assignments are based on a multitude of reported data on the electronic structure of Ni and NiO obtained using several different techniques, ranging from theoretical calculations to optical measurements and previous experimental ELS data. The ELS spectra obtained from NiO are markedly different than the Ni metal spectra as expected. The fact that the NiO spectral features do not vary appreciably with primary beam energy E p indicates that the surface electronic structure is similar to that of the bulk solid. Although most of the loss features from NiO can be attributed to interband transitions originating from the O 2p and Ni 3d states, the major peak centered at a loss energy of about 22.0 eV may be due to a collective excitation of the valence electrons. As expected, the ELS spectra obtained from the sputtered NiO surface exhibit characteristic features from both metallic Ni and NiO, with the metallic features becoming more prominent with increasing sputtering time. The most distinct change in the ELS spectra following reduction is the growth of the surface plasmon peak at 6.0 eV due to the presence of metallic Ni in the near-surface region. The loss features characteristic of bulk Ni metal are not as pronounced in the spectra obtained from the reduced NiO. These observations indicate that the metallic Ni formed during Ar + bombardment is confined to the very near-surface region, and they also demonstrate that ELS is sensitive to differences in the properties of thin metallic films or clusters and bulk Ni metal. The present results demonstrate that ELS can be used to distinguish between metallic nickel and NiO on the sputtered NiO surface.

The controlled deposition of a catalytically active material on a support, resulting in defined particle sizes and shapes, as well as control over the specific surface area of the active material and the support are usually limited when applying conventional catalyst preparation techniques. Flame spray pyrolysis offers the potential to overcome this limitation and is tested for the preparation of Fischer–Tropsch catalysts. Conventional single flame spray pyrolysis and, for the first time, double flame spray pyrolysis are compared for the synthesis of alumina supported cobalt catalysts. In the latter process the metal oxide precursors are combusted individually in two opposing nozzles. The key parameter for defining the final material composition is the intersection distance of the flames, which was systematically varied. The Fischer–Tropsch performance of the Co-based catalysts was studied in a fixed bed reactor at 230 °C and 20 bar. The catalytic results are discussed on the basis of structural characterization of the different catalysts by XRD, BET, TPR, UV–vis and TEM/EF-TEM. While catalysts made by single flame spray pyrolysis were inactive in the Fischer–Tropsch reaction regardless of whether cobalt was subsequently mixed with alumina or the supported catalyst was directly prepared in the flame reactor, the double flame sprayed catalysts showed good catalytic activity. Depending on the intersection distance of the two flames, the formation of cobalt oxide and alumina occurred separately in each flame reactor. In spite of the independent particle growth in the two flames, the double flame reactor geometry lead to good adhesion of the two oxides resulting in good stabilization of cobalt nanoparticles on the alumina support during the Fischer–Tropsch reaction.

The gas-phase decomposition pathways of diethylzinc (DEZn), a common precursor for deposition of Zn-VI compounds, were investigated in detail. The homogeneous thermal decomposition of DEZn in N2 carrier was followed in an impinging-jet, up-flow reactor by Raman scattering. Density Functional Theory calculations were performed to describe the bond dissociation behavior using the model chemistry B3LYP/6-311G(d) to estimate optimal geometries and Raman active vibrational frequencies of DEZn, as well as anticipated intermediates and products. Comparison of the measured DEZn decomposition profile to that predicted by a 2-D hydrodynamic simulation revealed that simple bond dissociation between zinc and carbon atoms is the dominant homogeneous thermal decomposition pathway. The calculations suggest several reactions involving intermediates and Raman scattering experiments confirming the formation of the dimer (ZnC2H5)2. In a different set of experiments, photolysis of DEZn gave evidence for decomposition by beta-hydride elimination. The results suggest that beta-hydride elimination is a minor pathway for the gas-phase homogeneous pyrolysis of diethylzinc. A reasonable transition state during beta-hydride elimination was identified, and the calculated energies and thermodynamic properties support the likelihood of these reaction steps.