Final Report Summary - CHIPCAT (Design of Thin-Film Nanocatalysts for On-Chip Fuel Cell Technology)
Executive Summary:
The project chipCAT aimed at the knowledge-driven development of a novel type of thin-film catalysts for silicon-based “on-chip” micro fuel cells (u-FCs). Combining fundamental surface-science, model catalysis, and first-principles computational studies, led to a detailed understanding of the surface chemistry on complex nanostructured catalysts. This microscopic-level understanding was used to tailor active sites and their mutual interplay at the nanoscale in order to maximize activity and selectivity and to reduce deactivation and poisoning.
Starting from new oxide-based materials with minor demand for noble metals we explored new material concepts. Target structures defined by fundamental research were then transferred to standard FC and u-FC catalysis using advanced thin-film preparation techniques, such as magnetron sputtering and vacuum deposition. Atomic-level control, using a broad spectrum of surface spectroscopies, in-situ/operando spectroscopies and microscopies with atomic resolution provided many channels of communication between fundamental research, applied research and technology development. Using modern microtechnologies, the novel tailor-made catalyst materials were integrated into working FC test devices. Prototype devices were fabricated for performance tests, the results of which provided basis for drawing up a process for laboratory-scale production of u-FC batches, as well as feedback to fundamental research.
Surface science, model catalysis, state-of-the-art theory, thin-film-technology, applied heterogeneous catalysis, and microtechnology were connected in an interdisciplinary effort aiming at the development of a new generation of metal-oxide FC catalysts with improved performance and stability. This project allowed dramatic reduction of critical materials use in fuel cell anodes and an incremental one in cathodes. The on-chip fuel cell prototypes also opened the pathway to a novel and potentially useful energy storage technology for mobile devices.
The project RTD was organized in work packages 2 through 7 thus: Fundamental knowledge was being pursued in WP 2 (theoretical and computational) and WPs 3 through 5 (experimental). A vigorous exchange of findings took place between these work packages; WP3 with its focus on model experiments in particular formed a meeting point and bridge between theory and the rest of experimental work, WP4 acted as a bridge from experimental work of fundamental research to the application driven WPs 6 and 7. These latter WPs covered the pursuit of the set practical goals of the project: In WP6 we scrutinized new materials for catalytic properties in thin film forms, created with (mainly) vacuum physical methods and with nanofabrication we pursued their absolute power density (W/cm2) and specific power density (W/mgPt cm2), WP7 was dedicated to the complete design of microfluidic fuel cells (micro-FC) ready for integration in silicon wafer chips along with integrated circuits. The thin film catalysts refined in WP6 were to be used as low-Platinum catalysts in PEMFC with the usual “sandwich” construction and also in the novel on-chip micro-FC coming out of WP7. In both cases the ambition was to produce a laboratory prototype with measurable and stable power output, ideally in a competitive range of power densities.
Project Context and Objectives:
The chipCAT project (FP7 GA no. 310191) was established by research groups with successful record in fundamental research of catalysts for proton exchange membrane fuel cells (PEMFC) and partners from related industrial fields. Our principal aim was to continue the promising explorations of the missing fundamental knowledge in catalysis and, at the same time, utilize this in knowledge-based development of technologies for cheaper, more durable and more mobile industrial-level fuel cells.
The motivation for our work was to contribute, with our expert knowledge and facilities, to making substantially more affordable proton-exchange membrane fuel cells, which were staying at high prices owing to the amounts of Platinum, expensive catalyst production and FC assembly and limited lifetime of the catalysts. As the envisaged new thin-film catalyst technology would be compatible with integrated circuit devices in Silicon wafer chips, we proposed to make an attempt at application of the catalyst films in making a prototype of and on-chip microfluidic fuel cell.
These were the principal objectives:
• By way of model experimental and theoretical (computational) physics uncover the precise mechanisms behind the function of the Platinum-based low-noble metal content catalysts, especially in combination with reducible oxides;
• Transfer of the patented low-noble metal content catalyst preparation method from model to practicable industrial scale;
• Explore a selection of other low-Platinum or precious metal-free catalysts for their suitability for deposition by the developed physical method and in planar geometry.
• Develop a physical thin-film catalyst preparation method compatible with the standard planar processes in semiconductor device industry;
• Develop a complete functional prototype of a microfluidic PEMFC device;
The consortium Coordinated by the surface physics group at Charles University in Prague (CUP) engaged computational modelling, model and applied experiments and technology and device development to achieve the objectives.
Project Results:
Convention notwithstanding, the articles published by the project are cited/listed in the approximate order of their publication, not the appearance order. Full list of peer-review publications is appended to this section.
*Computational modelling
Work Package 2 covered theoretical and computational research of materials relevant to hydrogen fuel cell catalysis. Throughout the project this WP communicated vigorously with WP3 (model experimental) and in its most prominent results also with WPs 4 and 5 (applied experimental). The theory groups produced 26 articles on their own and, remarkably, 14 articles together with the experimental counterparts in various configurations.
Two theory groups were engaged in this work package: one of prof. Konstantin Neyman at the University of Barcelona (UBCN) and the other of prof. Stefano Fabris at the Italian National Research Council (CNR). These two teams tackled the catalytic materials using different, mutually complementary methods. The UBCN team worked on the basis of standalone nanocrystals (nanoparticles), their facet properties and interactions with atoms and ions. Their work spanned several areas: metal nanocrystallites (elemental, alloys) and metal oxide nanocrystallites (e.g. ceria) pure and mixed with metal atoms (e.g. Platinum). The CNR team dealt with interactions (e.g. water oxidation, formation and activity of metal nanoparticles) on metal oxide surfaces and metal/metal-oxide interfaces, with the optional accounting for finite pressure of water vapor on the surface (effectively studying a water-ceria interface).
Overview
The chief subject of fundamental research in the project was the understanding anode (hydrogen splitting) and cathode (oxygen reduction) reactions in PEMFC in the context of nanoparticular and atomically dispersed catalytic metals in reducible oxide matrix. The theoretical and collaborative research led to a substantial breakthrough in understanding of the atomically dispersed Pt-CeO2 system [7], when the project identified stable sites – {100} nanopockets – on ceria nanoparticle surface, where a single Pt atom was stabilized and, importantly, in 2+ oxidation state. This ionic state was already suspect from experimental work to being key to the material’s catalytic function. The theoretical achievement stemmed from a detailed study of the ceria nanoparticles alone [16,20] and the nanoparticle-approach-driven discovery in [7] prompted further advances purely theoretical [14,32,37,40,61,62,68,69] and in theory-experiment collaboration [47,56].
Another notable advance [58] was made later in the project by the slab-surface-based research which, in coupling with model experiment, showed that adsorption sites on naturally stepped flat surface of ceria offers site analogous to the {100} nanopockets that act the same way. This was a strong evidence that the atomic dispersion of Pt on ceria is not only possible, but occurs naturally at a rate that can be practically harnessed, just as had been indicated by the works of applied experimental research [23,25,47,64].
Nanoparticles of elemental metals and binary alloys
Platinum and other metal nanoclusters were a matter of attention as well as elemental metals [5,6,8,12,18,35,59,67] and binaries (alloys, core-shell) arrangements [34,38]. As it transpired that the atomically dispersed Pt-CeO2 system is suitable for anodes, but not efficient in the cathode-side oxygen reduction reaction (ORR), these studies of Pt and binary nanoparticles fuelled progress in the understanding the cathode mechanisms and developing the optimal catalysts for ORR.
The theoretical research of electronic structure of bimetallic alloy nanoparticles [38] was another major advance made by the team of prof. Konstantin Neyman and a concept of alloy nanoparticle computational screening for catalytic properties derived from that work was filed, in the last year of the project, for patenting in the U.S. as the most likely market for such technology.
Throughout, there was a persistent question of particle-support interaction in the metal/oxide arrangement, as it is well known that, depending on the interaction strength, properties of the nanoparticles can be influenced or even determined by this interaction. The effort to determine the strength of the interaction gave an important result [59], where model experiments heavily backed by theory produced a method to determine the amount of charge transfer between particle and support as a function of particle size.
Ceria and atomically dispersed metals
Platinum in the form of nanoclusters supported by ceria was studied in parallel [14,66] so as to reveal the dynamics of coalescence of Pt atoms and to understand, why in some scenarios the dispersed Pt atoms do not tend to form stable clusters. As the coalescence of Pt is among the leading causes of Pt catalyst ageing and fuel cell degradation, these researches were of supreme importance and tied the molecular dynamic simulations to experiments both in model (ultra-high vacuum) and in operando (ambient vapour pressure, electrolyte) conditions.
In the rank of dispersed metals (atomically or in nanoclusters) in reducible oxide matrix there was an amount of necessary research done apart from the Pt-CeO2 combination. Other metal atoms were tested with respect to the {100} nanopocket site and the results of these calculations were summarized in [61].
In parallel non-Platinum atoms and clusters were studied, supported by ceria and other oxides, in theory and in experiments: Au/CeO2 [3], Au/TiO2 [13], Rh/CeO2 [54], Pt/MgO, Pd/MgO [12] and others. These researches had multiple motivation: the fundamental understanding, the practice-aware push for minimizing or zeroing the need for precious metals in hydrogen fuel cells, and another practical effort to find ways to reduce the catalysts’ susceptibility to poisoning.
Other materials
While most of the project addressed materials for hydrogen fuel cells, a dedicated attention was given to materials for hydrocarbon oxidation in the latter half of the project [46,50,65,70,71] such as palladium and cobalt oxide. These are of interest, for example, for anodes of direct methanol fuel cells (DMFC), which are a major “mainstream” alternative to hydrogen fuel cells (HFC). As the HFC advances of the project were satisfactory throughout chipCAT, however, there was no material utilization of the new methanol-oxidation-related knowledge in the project.
*Model catalysts
Work Package 3 was the project hub for fundamental experimental research. Prof. Jörg Libuda at the Friedrich-Alexander University of Erlangen-Nuremberg (FAU) led this WP and his group made much of the experimental work on model materials in UHV and in operando conditions. Their focus was on model materials in planar geometry used for revealing mechanisms of adsorption and surface reactions of H2, O2, CO (i.e. molecules most relevant for hydrogen PEMFC) on the novel materials. In model studies, other small test molecules (methanol, ethanol, acetic acid, ethylene glycol,...) were also used in probing the material properties with controlled adsorption/desorption and surface sensitive spectroscopies: This was mainly a broad range of photoelectron spectroscopies in UHV and using synchrotron radiation, and a number of infrared spectroscopy methods using existing and newly installed instrumentation (in-situ/operando IRAS, phase-modulation IRAS, operando microreactor with ATR and IRAS mode).
One major line of research was on the different phases of ceria [2], Platinum [11] and atomically dispersed metals and metal nanoclusters on ceria. It was the multiple synchrotron radiation photoelectron spectroscopy experiments with model Pt/CeO2 and Pt-CeO2 materials (the different notation refers to Pt nanoclusters on ceria and Pt atomically dispersed on (and in) ceria, respectively) that were the first to reveal the presence and significance of the Pt2+ ionic state in ceria matrix [19]. The same type of experiments gave a substantial basis to deciphering the mechanisms subsequently [7,31,49,55,60].
In parallel, properties of similar ceria-based materials were studied in response to the call for finding the limit for reducing Pt (and Pt group metal) content, if it existed, and to make the Pt-based catalysts more resistant to poisoning by contaminants, mainly by CO. In the latter line of research it was tin (Sn) that received substantial attention as a promising additive to reduce the catalysts’ susceptibility to poisoning by CO.
As described in detail in [17], Sn retains the oxidation state +2 which is associated with the formation of a mixed Pt-Sn-CeO2 oxide. At high Sn concentrations, we detected a small amount of metallic Sn0. The formation of the Sn-CeO2 mixed oxide is accompanied by an increase of the RER associated with the increase of the Ce3+/Ce4+ ratio in the film. The nature of the Pt* species at 72 eV is not entirely clear. Earlier we assigned the corresponding species at 72 eV on Pt-CeO2 films to PtO. The intensity of the corresponding Pt* peak increases during annealing to 700 K at the expense of Pt2+. The rather small binding energy shift of Pt* peak (~0.4 eV) upon annealing suggests high stability of the PtO species.
Interestingly, annealing of the film also triggers the decrease of the metallic Sn0 which is accompanied by the increase of Sn2+ intensity. The decrease of the Sn0 is accompanied by a slight decrease of the RER. This suggests that metallic Sn0 is converted into Sn2+ followed by its diffusion into the bulk of Pt-Sn-CeO2 oxide during annealing. In conclusion, the Sn atoms were shown to be rather mobile, able to form a complex (Sn,Ce)O2 oxide as well as Pt-Sn binary nanoclusters (alloys and core-shell) and to transition between the states [17,24,45,62].
In the second period work continued on the Pt-CeO2 and Pt-(Sn,Ce)O2 systems, as mentioned above with references to publications. In addition, the search for alternative catalytic materials meant that some elements (Pd, Ni [55]) were tested in conjunction with ceria in order to find the possibility to disperse them atomically like Platinum and to reveal their function in such state.
In as much as Pd is concerned, it was concluded that one characteristic difference between the two metals is the lower thermal stability of Pd2+ sites as compared to Pt2+. This difference was associated with the fact that Pt2+ is preferentially stabilized at the square oxygen pocket sites at the surface, whereas Pd can be stabilized both at the surface and in the form of a solid solution in the bulk. In contrast to the Pt−CeO2, the Pd−CeO2 films maintain a considerably higher degree of reduction at even lower dopant concentration. The higher degree of reduction in the Pd−CeO2 films is likely caused by the formation of the PdO phase. The charge transfer between Pd and Ce4+ is accompanied by the withdrawal of lattice oxygen and growth of PdO particles. Consequently, the decomposition of PdO during annealing results in re-oxidation of Ce3+.
We studied the Ni 2p5/2 spectra obtained from 13% Ni−CeO2. The main peak at 855.0-855.3 eV is associated with Ni2+ ions. Additionally, two satellite features I and II emerge at 857.0 eV and 861.1 eV, respectively. Satellite I arises from the non-local screening effects and, therefore, is sensitive to the degree of Ni dispersion in CeO2. In particular, the low intensity of the satellite I is consistent with the high degree of Ni dispersion in CeO2 film. Satellite II emerges due to a shake-up process associated with the O 2p→Ni 3p charge-transfer multielectron excitation during ionization of Ni 2p core level. Annealing of Ni−CeO2 results in a decrease of both satellites I and II which is consistent with enhanced diffusion of Ni2+ towards the bulk. The Ni−CeO2 films retain a low degree of reduction, similar to that detected for the Pt−CeO2 mixed oxide. Unlike the Pd−CeO2, we identified only one phase associated with Ni2+ in CeO2. Since Ni content in 13% Ni−CeO2 is close to the solubility limit, it is possible that the presence of satellites I and II indicate the existence of small amount of a NiO phase. The disappearance of the satellites I and II from the Ni 2p spectra suggests diffusion of Ni2+ into the substrate and, most likely, decomposition of the NiO phase during annealing in UHV. This observation suggests that a single phase associated with a solid solution of Ni2+ is formed for 13% Ni−CeO2 above 600 K.
In summary, Pt2+, Pd2+, and Ni2+ were found to be stabilized in atomically dispersed form on CeO2 in the limit of low concentration. However, the thermal stabilities of Pd-CeO2 and Ni−CeO2 are compromised by the diffusion of Pd2+ and Ni2+ into the bulk upon annealing above 600 K. This is in contrast to the Pt-CeO2 which is stable at these conditions.
As an alternative material for hydrogen dissociation, cobalt oxide Co3O4 was also studied as model material of Co3O4 thin films on Ir(100) substrate under UHV FAU [70,71]. The films of Co3O4(111), stoichiometric and with surplus Co in atomic and nanocluster form, were characterized structurally and studied using IRAS with adsorption of CO and O2. By this latter method [70] the adsorption sites were revealed for low coverage (surface Co2+ sites), high coverage (sites between Co2+ ions) and bridging carbonate species associated with crystal defects.
In the second half of the project a substantial line of research focused on the cathode-side reactions and materials, namely binary nanoparticles of Pt-Ni and Pt-Co alloys. The Pt-Co nanoparticles were studied under UHV conditions and their behaviour under annealing was tracked [39] in HAXPES. It was concluded that Pt shell tends to form on the surface and there it prefers to occupy edge and vertex positions, while Co atoms linger on the facet insides. Subsurface layer consisted predominantly of Co atoms and the core remained Pt-Co, so that the discovered preferred arrangement was PtYCo1-Y-core@Co-rich-subsurface@Pt-rich_shell.
It had been known from before that in electrochemical environment (in electrolyte under potential cycling) the Ni and Co atoms are leached out of the particles. This was the case in several kinds of particles, including those with core-Pt_shell structure. Pt-Ni thin films were cycled in electrochemical AFM with increasing cycling potential limit, which was alternated with surface morphology imaging.
Strong dissolution effects result in a fast degradation of Pt-Ni catalyst upon cycling above the upper potential limit of 1.2 V. The degradation mechanism involves oxidation of Pt-shell followed by leaching of Ni core.
Dynamic changes of the surface sites were also monitored, this by means of IR spectroscopy using CO molecule as a probe. It was found that migration of Ni atoms from the bulk to the surface/subsurface region results in the formation of Pt-Ni species that influences the adsorption properties of CO molecule. Upon cycling to higher potentials, these species dissolve recovering the composition of Pt-rich shell.
*Thin films
Large surfaces
Two tasks of Work Package 4 were started at the commencement of the project and they concerned the refinement of the thin film growth techniques for uniform coverage of large (several centimetres squared) substrates and the first use of ALD as an alternative to magnetron sputtering. Most of the research in WPs 4 and 5 was carried out by the teams of Charles University in Prague (Coordinator) led by prof. Vladimír Matolín and of the University of Burgundy in Dijon led by prof. Sylvie Bourgeois and prof. Bruno Domenichini.
In scope of the first task an in communication with WP 6 NafionTM and AquivionTM membranes were etched by oxygen plasma. In AFM micrographs a desirable roughening was observed without any obvious degradation that might signify a compromised membrane function. These membranes, coated with 10nm and 30nm thick layers of Pt were passed on to WP 6 for testing in MEAs. Those tests, however, brought no significant improvement of performance or increase in active surface; the only clear change was a decrease in membrane resistance. It had been observed before that layers of Platinum oxide (PtOx) subsequently reduced by Hydrogen or CO exposure have noticeably rougher morphology, even at room temperature. This could be a fuel cell friendly “activation procedure”, but layers created by reduction of PtOx performed badly in FC tests.
Ceria
Structure of the 2 µgPt/cm2 catalyst film deposited on nano GDL (nGDL) was observed in STEM and HRTEM micrographs. Catalyst coating has striped character, which is due to porosity of the carbon substrate showing the channels etched by oxygen plasma. The channels are filled with the nanocrystalline catalyst exhibiting particles 2 – 3 nm large. HRTEM shows presence of two types of nanoparticles of cerium oxide, CeO2 and Ce2O7, respectively. We compared the catalyst structure before and after hydrogen annealing simulating working conditions of the PEMFC anode. Before hydrogen annealing, CeO2 and Ce7O12 crystallites were found, whereas after annealing Ce7O12 are predominant after annealing. Cerium carbide (CeC2) particles were also observed at the interface with the nGDL layer.
Nanostructured catalyst films in fuel cells
Sputtering Pt-CeOx onto nGDL represented a leap-wise advance in surface area of the NCF. PEMFC tests with Pt-CeOx anodes showed that among them existed some optimal, relatively tiny loading of 2µg/cm2, which produced biggest power density. It was remarkable that the 2µg Pt-CeOx anode gave out power at 90% of the reference, while its Pt loading was 1000 times (three orders of magnitude) lower. In another test 2µg/cm2 of Pt was deposited onto nGDL without CeOx and the power output was 5.5 times lower than that of Pt-CeOx with the same loading.
Examination of the low-loading Pt-CeOx material in model planar geometry in Work Package 3 showed that (up to a certain relative amount of Pt) the hydrogen exposure shifts all Platinum in Pt-CeOx into the Pt2+ ionic state. At higher Pt ratio the Pt exists in Pt2+ ionic state as well as Pt0 metallic state (i.e. clusters).
Electrochemical cycling is a standard test of catalyst durability and we subjected our Pt-CeOx layers to it. One cycle consisted of a sequence of stable power delivery, short circuit and open circuit. The graph shows a remarkable endurance of the material, which indicates that the Pt2+ ionic state is very stable and the dispersed atoms do not have a tendency to coalesce and form metal clusters.
As a material for the cathode side of the PEMFC (and the ORR) the standard material, Pt-Co alloy was tested. Apart from a fundamental investigation in WP 3 [39] the Pt-Co was deposited on nGDL at different initial ratios and used as a cathode catalyst in PEMFC tests and, having shown stable enough, was used in the first PEMFC with NCF on both electrodes. For a broad range of Pt relative amount (from 100% to 40%) on the cathode the power outputs are very similar. The loadings (measured by quantitative XPS) were 2 µg/cm2 on the anode and 48 µg/cm2 on the cathode. Taking the power density of 0.120 W/cm2 into account this fully thin-film PEMFC exhibits specific power (Pt-mass activity) of 2.4 kW/gPt. Endurance test of this PEMFC showed a relatively large variation in power density during the test but also a very small decrease after 3000 cycles. In the view of known Cobalt leaching from Pt-Co during cycling and assuming that it happens on the cathode of the present PEMFC we infer that it is mainly the dispersion of nanoparticles in the nGDL that stands in the way of coalescence-driven ageing (i.e. decrease of power).
Carbon nitride interlayers [26,44]
In the second period the sufficient porosity of the NCF was still a major concern, although nGDL as a substrate provide a major advancement. The next one was achieved by the discovery that during deposition of ceria (or Pt-CeOx) onto carbonaceous substrates by reactive sputtering in Oxygen plasma the substrate is etched by the plasma simultaneously with the material deposition.
As the deposition initially produces disconnected islands, which consequently shadow the substrate material beneath while the rest is etched away, an irregular array of (Pt-)CeOx pillars. This process can be tuned by parameters of the plasma discharge to some extent, but more importantly by the composition of the substrate itself. It was found that the optimal efficiency was achieved using carbon nitrate (CNx) layers for substrate, which can be prepared by reactive carbon sputtering in Nitrogen plasma. Upon deposition of the Pt-CeOx in Oxygen plasma almost all the CNx material is removed and catalyst-coated nanowires remain. Pt concentration had no influence on the as-prepared morphology of the catalyst. This method was stacked on the nGDL as a primary substrate, so that NCF layers with several “levels” of pore sizes resulted and such structure improved the active surface size still further.
Ceria and ALD/CVD
The ceria has been subject to a thorough examination in this WP in order to find the optimal method for its deposition. Apart from the different magnetron sputtering recipes, growth via ALD/CVD was tested.
The PVD produced mode compact layers than ALD/CVD. In context with experience from fuel cell tests of these layers (in this WP and in WP 6) nanoporosity on the scale given by the ALD is “invisible” in a fuel cell and the layers seem compact from that perspective. On the other hand, ALD has been successfully tested for coating the walls of microfluidic channels (for use in WP7) with high aspect ratio.
Tungsten carbide and oxide
For testing non-Platinum materials as catalysts candidates tungsten carbide deposition methods were explored [30]. The WC layers obtained in the first period were rather compact.
Tungsten carbide and oxide films were fabricated on carbon membrane in the second period by direct liquid injection CVD (DLI-CVD). After deposition, the carbon membrane is not degraded and the deposit consists of crystallized needles with a diameter around 5 – 7 nm. An envelope is visible around all the needles as a light grey contour. The contrast suggests that the envelope consists of amorphous carbon. Interplane distances measured on high-resolution images are consistent with tungsten trioxide WO3 crystallized in a monoclinic structure.
After the first period, more experimenting remained to be done. In the second period, however, having experimented with tungsten oxide (WO3) and tungsten oxycarbide (WCO2) in planar systems and deposited on GDL it was concluded that this class of materials is not catalytically active. Only WO3 remained in the scope of the project as an alternative to ceria – a matrix in which catalyst (e.g. Platinum) atoms can be dispersed.
*Characterisation of materials
The Work Package 5 was dedicated to the standard characterization of novel materials’ structure, chemical states (electronic structure) and integral behaviour under operando, typically electrochemical conditions. We focused on materials that were not much simplified for the measurements and analysis. In contrast with model experimental work of WP3 we researched the characteristics of materials close to the form in which they would be used as catalysts in fuel cells.
The first period had it planned to investigate Pt and Pt-Ru nanostructured NCF directly on an “inert” substrate and on ceria, which was of great promise as an “active” substrate already. The Pt-Ru alloy had been put on hold early on; its chief benefit was to be the comparatively higher CO poisoning resistance with respect to Pt. The Pt-CeO2 exhibited similar property, as shown below, so the Pt-Ru alloy lost practical interest and remained a largely unexplored branch.
Comparing TF catalysts and Pt(111) reference by cyclic voltammetry, we observed characteristic changes of the COt band with respect to both the shape and the band position. On Pt(111) the signal is located at 2064 cm-1, whereas on the Pt doped CeO2 samples a red-shift was observed. The corresponding bands are observed at 2056 cm-1 for high Pt concentration and 2023 cm-1 for the low Pt concentration. Whereas the band position is characteristic for metallic Pt, the red shift can be associated with a different morphology of the Pt nanoparticles. Red-shifted features are characteristically observed for CO adsorption at low-coordinated Pt sites. We conclude that for the Pt-CeO2 sample with low Pt concentration, a fraction of the Pt is reduced to Pt0 and stabilized in form of very small nanoparticles. This hypothesis is supported by the observation that the CO signal becomes significantly broader at low Pt concentration. The CO signal on the pure Pt thin film shows a slight blue-shift (2074 cm-1) with respect to the Pt(111) reference. Possible explanations involve the enhanced activity for methanol oxidation at defect sites (formation of CO at more negative electrode potential as also indicated by the shape of the absorption band), co-adsorption of electron withdrawing species and/or and increased CO density on the rougher surface of the magnetron-sputtered film.
The diffuse reflection IR Fourier transform spectroscopy (DRIFTS) was employed to investigate the composition and electronic state of Pt and Pt/CeO2 thin films. Specifically, we studied spectra obtained after dosing CO on pure Pt and Pt/CeO2 thin films prepared on Macor® ceramics by magnetron sputtering. A spectrum of the clean samples after O2 oxidation at 573 K for 15 min has been used as a reference. After CO dosage, the Pt sample (top spectrum) shows a single signal at 2081 cm-1 which can be attributed to CO bound linearly on metallic platinum (Pt0) atoms. The lack of a signal at 1840 cm-1, which is characteristic for CO bridging two Pt atoms, suggests that the dispersion of the Pt film is very high. The samples of Pt co-deposited with CeO2 show no CO-adsorption signals in this range after dosage of CO (spectrum not shown). After annealing the sample in a H2 stream for 1 h at 773 K and subsequent CO dosage, a signal at the same frequency as for the Pt sample appears, indicating the presence of metallic Pt. With increasing reduction temperature, another signal at higher wavenumbers (2161 cm-1) arises, which can be assigned to CO adsorbed linearly on cationic platinum species (Pt2+). The intensity of this new feature increases with progressing reduction temperature. Finally, after reduction at 1017 K no adsorbed CO can be found on the sample anymore.
A considerable number of samples of Pt-CeO2 thin films were made throughout the project by various methods and characterized using electrons microscopies. Highly instructive among the analyses was the HRTEM. TEM micrographs of Pt-CeOx layers prepared using the DLI-CVD process were obtained and from them the films were deemed “nanoporous”. For this characterization the films had been made on flat substrates – this contrasts with efforts made in WPs 4 and 6 that had a great drive for increasing the NCF active surface. However, here the aim was the efficient HRTEM characterization made easier at a greater density of nanocrystallites in each HRTEM lamella. Moreover, the micro-FC arrangement worked out in WP 7 specifically needed knowledge about growth and composition of materials prepared on flat surfaces like Silicon chips.
HRTEM imaging in conjunction with synchrotron radiation XPS provided major evidence in the progress of understanding of the Pt dispersion in ceria. In the second period of the project, the {100} nanopockets on ceria nanoparticles [7] had already been identified as sites stabilizing Pt atoms in the 2+ ionic state. We had obtained HRTEM micrographs of CeO2 nanoparticles (roughly 10 nm in diameter) in Pt-CeOx film and it was clear from this image (and several like it) that the {100} sites are exposed on ceria NP surfaces in no small numbers. Interestingly, though, small amounts of Platinum deposited on the ceria, clearly present there according to XPS, were not detectable in HRTEM. More precisely, for low Pt dosages the EDX measurements in the STEM mode detected uniform presence of Pt signal throughout the layer, but no local aggregations (clusters). The self-same samples characterized using SRPES exhibited presence of Pt almost exclusively in the Pt2+ state, whereas those with greatest total Pt/Ce ratio showed both Pt2+ (ionic) and Pt0 (metallic) species.
The SR XPS measurements with lower incident photon energy have shown that the relative Pt concentration is 5 times higher than the one obtained with conventional XPS. One should point that in this case, the analysis thickness can be decreased below the size of a single particle. As the Pt relative content strongly increases when the probing depth decreases from the entire particle to its topmost atomic layers only, we can conclude that the spatial distribution of Pt through the particle is not homogeneous and, more precisely, that Pt is localized at the surface of the ceria particles instead of in their volume. It was concluded that in low dosages Pt disperses atomically about the ceria nanocrystallite surfaces and are stabilized (trapped) in the {100} nanopockets. With increasing Pt dosage the nanopockets become saturated, after which the Pt atoms start forming metallic nanoclusters. These conclusions eventually led to the realization that the {100} sites are quite abundant on compact ceria thin films as well, namely at step edges, and that low dosages of Pt are dispersed atomically on flat ceria by decorating the step edges [58].
While standing as the only satisfactory hypothesis consistent with the data and supported by DFT calculations, the idea of Pt dispersed atomically on ceria still lacked a conclusive proof in the form of images, where one could “see” the single Pt atom. Indeed identification of a single atom is a major challenge, because the methods that offer the best way to resolve individual atoms (STM, AFM) have very limited capacity to identify those atoms as to their chemical element, not to mention their oxidation state. However, shortly before the project’s end the team got to characterize the Pt-CeO2 samples on one of the world’s highest-performance HRTEM, the NION SuperSTEM at the Trinity College of Dublin.
Having located a ceria nanoparticle oriented with its (100) plane in the image plane the intensity line profile was measured along the atom rows. In these line profiles certain brighter points were identified and using the theory of electron scattering they were ruled, with a great degree of confidence, as being the visualization of Pt single atoms localized in surface {100} nanopocket sites.
We compared formation of Pt structures for different Pt loadings and our observations were consistent with expectations based on the atomic dispersion model: At Pt loadings lower than 10 µg dispersed atoms were observed, occasional Pt dimers (two Pt atoms in neighboring surface nanopockets) and rather rare disordered clusters of a few Pt atoms. As the loading surpassed 10 µg he Pt clusters occurred in much greater abundance, consistently with the saturation of the {100} nanopockets, which accommodate the single Pt2+ ions.
*Thin catalyst layers evaluation and application testing
Fuel cell testing was carried out by the German industrial SME SolviCore GmbH & KG, Hanau, in the second half of the project the testing was largely taken over by the Charles University team. Work Package 6 progressed to a large degree with the aid of the know-how of Solvicore(transformed into Greenerity GmbH by the time of writing this report) and the reports were labelled confidential. This report, therefore summarizes information reduced so as not to infringe that protection.
The NCF layers tested in this WP were prepared mostly by the Charles University team and the Czech SME LET optomechanika Praha Ltd.
Benchmarking
WP6 started with sharing the (public) information on standards of automotive industry and goals of fuel cell performance, setting up protocols to be used in fuel cell testing of catalysts generated by the project, and preparing and testing a benchmark membrane assembly (MEA) prepared so as to be comparable with those using the nanocatalyst films (NCF). As the second step, MEAs with NCF on the anode side were assembled and tested. The conclusion of these tests was that the critical weakness of the NCF is their compactness, i.e. a vanishing porosity. It was the porosity, therefore, that needed to be addressed primarily.
Testing strategies to enlarge surface area
Next, the WP proceeded by testing MEAs made with NCF deposited on variously prepared membranes. The MEAs were mostly assembled with the NCF catalyst on the anode and a well-reproducible state-of-the-art (i.e. “standard” for the purposes of comparison within the project) catalyst on the cathode side. One MEA was assembled with NCF on both sides.
It was concluded that ion-etching of membranes reduces its internal resistance, but the etching did not have a marked effect on the compactness (low surface area) of the NCF. The universally low surface areas of NCF (the EPSA parameter) indicated that nanoporosity identified elsewhere in the project (see Work Package 5) was not of the scale that can be exploited in the electrochemical environment. Results obtained for the NCF-NCF MEA, when normalized to Cathode loading (intrinsically much smaller than for the “standard”), showed some a certain promise, provided the NCF could be prepared with greater EPSA.
Tests of variously modified NCF have shown that the electrochemical surface area (ECA in m² g-1), of sputtered platinum electrodes onto edged membranes could not be increased by the desired factor of 10 (i.e. from 6 m² g-1 to 60 m² g-1). Indeed, independently from the NCF thickness, 10 or 20nm, the surface area led to same EPSA of around 2 cm²Pt cmMEA-2. This indicates that the electrodes are two-dimensional (compact) and not three-dimensional (sufficiently porous). A certain porosity was observed in SEM and AFM images, but it seems that the pore sizes are too small for protons and gas molecules to penetrate the layer sufficiently. The porosity of the NCF still remained to be increased.
The following approaches were proposed to help increase the porosity and, therefore, the active platinum surface:
1. Nano GDL – the possibility of direct sputtering of platinum on GDL (nanoGDL) was to be further explored and exploited
2. CCM based on a coated carbon/ionomer layer with sputtered platinum – the carbon+ionomer+sputtered_Platinum layer could be applied either directly to the membrane, or to inert substrate, from which the layer would be transferred onto the membrane
3. Carbon powder sputtered with platinum – either using high surface area carbon particles or carbon nanotubes
4. Sputtering polymer – attempting to sputter polymer onto the membrane with the expectation that increased surface would result, Pt would be sputtered onto the membrane thus prepared
The strategy primarily used was sputtering Pt onto a carbon ionomer layer (c-i-l) in several thicknesses and used as anodes. In MEA they were again coupled with standard cathodes. At the nominal Pt layer thickness of 30nm the Pt/c-i-l NCF anode had similar performance to the benchmark; this was the first observation of so close a result, within a few percent of power at voltages up to 1.6 V. This best anode had ECA 7 m2 gPt-1 and EPSA 4 cm2Pt cm-2geo , which was not an improvement.
Tests performed on anodes with a ceria interlayer (Pt/CeOx/c-i-l) gave worse results, the main effect of CeOx presence being the increased resistance of the MEA. On the other hand, testing of the layers on CO poisoning was done using the CO stripping method and it showed no CO adsorption for low Pt loadings. At large Pt loadings some CO capture was observed, which was attributed to presence of Pt0 clusters as opposed to atomically dispersed Pt.
Pt/c-i-l NCF were tested on the cathode side as well. The loadings were a factor 4-10 lower than that of standard cathodes and the NCF cathode MEAs exhibited systematically lower cell voltages, bigger Tafel slopes and a big variation indicated that the NCF suffered from mass transport issues at currents even as low as 0.2 A/cm2. In other words, the insufficient NCF porosity persisted as the chief obstacle. Pt/CeOx NCF deposited by glancing angle deposition (GLAD) showed some improvement in performance (and porosity).
Toward fully thin-film PEMFCs
In the third period the MEA assembly and testing for the project has almost completely migrated to the Coordinator CUP), thanks to the built-up FC testing facility and the adoption of knowledge and industrial standards for the testing. Indeed, CUP progressed with its own instrument development so that alongside the NCF technology the CUP team began to market their own fuel cell test stations in a commercial venture (LEANCAT s.r.o. a spin-out from CUP in partnership with industry).
The chipCAT Pt-doped ceria catalyst research was concluded by preparing thin film anodes with almost negligible content of platinum, while achieving industrial standards in performance and durability. The Pt-ceria catalysts, which had been introduced as an innovative material due to very high specific power, were continuously improved during the project and represent one of its best achievements.
During the first three years of the project we repeatedly saw that further improvement of the catalyst performance concerning its power density can be obtained by tuning the catalyst film substrate porosity, which would lead to the increase of the catalyst active surface [23,25]. Additionally, we focused our effort onto tuning the hydrophilicity/hydrophobicity ratio by preparing our own nano GDL, to which we switched from the commercial one used previously used (Sigracet GDL).
The final catalyst preparation steps were:
- GDL coating by carbon NPs dissolved in ionomer and PTFE solutions;
- reactive deposition of the CNx GDL interlayer;
- Pt-ceria TF deposition.
CNx interlayer is increasing the efficiency of oxygen plasma etching. Efficiency of the CNx film etching was well observable in case of the same process applied on flat CNx layer deposited on a Si wafer.
Fuel cell testing using single cell of 4.5 cm2 of MEA active surface was performed for different Pt loadings from 1 µg of Pt per 1 cm2 of the catalyst geometrical surface. The test was performed by using commercial cathode. The results show that the maximum power density increases with Pt loading and approaches asymptotically the maximum value of 1.4 W/cm2 for 10 µg Pt/cm2.
By comparing with similar results presented in the progress report for 2nd period where we have shown results obtained without using a CNx interlayer we can see the same asymptotic behaviour, however in previous case the maximum power density was 0.6 W/cm2.
Thanks to a new fuel cell testing laboratory at CUP, we performed accelerated durability tests of 10µg and 4µg Pt samples using two different testing protocols, applied simultaneously to three identical samples. Both tests have shown high stability of Pt-ceria anodes with performance decrease smaller than 10%.
Beside development of an efficient Pt-ceria TF anode an effort has been made in development of a new TF cathode in order to develop full thin film MEA and/or efficient micro-fuel cells with optimised electrodes. We proposed a new procedure based on preparation of a nanoporous substrate by reactive deposition of carbon nitride and cerium oxide on “homemade” ionomer+PTFE+C NPs coated GDL. Then the substrate was coated by metallic Pt thin film and a MEA made with it was tested. The stable MEA revealed 0.45 W/cm2 giving specific power of 9 kW/g Pt.
These results have prompted tests of NCF PEMFC with alternative materials, which showed promise in theory, model and applied research of WPs 2 through 5. While these tests are not concluded yet, there have been promising data collected, for example, on atomically dispersed Pt in WO3 matrix.
*Chip design, microfabrication
Most of the work related to the on-chip micro-fuel cells was carried out by the Italian SME ThinderNIL s.r.l. In accordance with the Work Package 7 plan, in the first Period work started on the design of the microfluidic chips to form the base of the micro-fuel cells. This was done by means of finite element calculations. Although delayed to the second period, the modelling was carried out and took into account practical problems and workarounds discovered in parallel.
Basic concept
In an on-chip microfluidic fuel cell the anode and cathode catalysts would be placed in close-lying microchannels etched into one face of a Silicon wafer (chip). This concept assumed that the catalysts would cover the channel walls and be contacted by thin film electrodes so that the barrier between the two channels is reliably insulating. The proton-exchange membrane (PEM) would be hot-pressed onto the top of the chip and during function protons generated in the anode (fuel) channel would migrate laterally towards the cathode, where they would recombine with oxygen into water. Water is to be carried away by the cathode channel as exhaust.
Lithography patterns
The realization of the desired on-chip devices required choosing channel patterns to test, developing the recipes for channel lithography and for the depositions of electrodes and catalysts and producing the necessary set of lithography masks. There were several patterns made in the masks: A simple one with only one linear section of close-lying channels and a complex one with a large number (ca. 20) of interdigitated channels for fuel and oxygen. The simple one was used in the first proof of principle devices, whereas the latter was later deemed non-functional and replaced by one, where the pair of channels meanders many times. This latter type’s advantage is the lack of dead ends.
Practical tests of hot-press fixing of the PEM to the Silicon chip showed several difficulties. A major one to overcome was the fact that the PEM (NafionTM or similar material) dilates considerably, when soaking with water molecules in preparation for function. Hydration is essential for the membrane function, which is why the working gasses are wetted en route to the PEMFC by passing through bubblers. However, the expanding soaked membrane tends to squeeze into the channels and block them. To overcome this problem an array of micro-pillars was left inside the channels. They were dense enough to prevent the channel blockage by membrane but sparse enough to not impede the flow of the fuel and oxygen through the channels.
A “pioneer” microchannel cell
As the design, lithography work and tests went on, the planar fuel cell concept was tested with a “macro” model. There the channels had larger dimensions, were etched in glass blocks and did not require micrometre precision for machining, etching or for thin film depositions. Designed and manufactured between the teams of Charles University and of LET optomechanika Praha.
The parallel channels in the “pioneer” planar cell were separated by 100 µm, 1 mm wide and 100 µm deep. Length of parallel part of channels was 10 mm. Both anode and cathode channels were coated with 20 nm thick Pt catalyst film, the electrical contact of electrodes were ensured by the thin film of gold evaporated on the cell glass plate. The Nafion NR-212 membrane was used as electrolyte. The planar fuel cell was tested using hydrogen and oxygen gas at room temperature and atmospheric pressure and it did produce a standard obtained polarisation curve.
Finite-element simulations
The device design simulations already included the pillars inside the microchannels. For the simulation to was decided to assume the micro-FC device working as a direct methanol fuel cell (DMFC). In other words, the finite elements computation solved Navier-Stokes equation for incompressible fluids in laminar flow, where the appropriateness of the latter assumption (laminar flow) was tested by enumeration of Reynolds number. The highest estimate of Re was ~0.08 and, therefore, the solutions of NS equation can lead to realistic solutions.
Laminar flow of methanol was simulated using COMSOL software, first in an empty channel with dimensions being 200 µm wide, 30 µm deep and 400 µm long. The same computation was then performed with the channel filled with square pillars of 10 μm arranged in a square array of 30 μm period, separated from the walls of 20 μm. The pressure drop over 400 um was 3.5 Pa in the former case and 13.2 Pa in the latter. In this regime the pressure drop is proportional to the viscosity and linearly dependent on the fluid speed. Therefore, a single simulation done for a given value of the viscosity and given fluid speed is used to generate the entire set of data for all viscosities and fluid speeds of interest.
Assuming 50% FC efficiency and targeting 10 mW/cm2 output, the pressure drop along a channel was below 2 bar/m at for both geometries (without and with pillars) at all methanol concentrations from 2M up. The longest meander considered in the project for an on-chip micro-FC is 80 cm long. It is indicated that the available flow will be able to supply enough fuel for the target output power.
Chip interface platform
After making the microfluidic devices, depositing the electrodes and catalysts and pressing the PEM on, it was necessary to connect it to a fuel-cell testing infrastructure – a platform interfacing the gas leads and electronics with the microfluidic device. The platform design was made already in the first period, although it was utilized only in the second.
A guiding principle adopted in the design of the chip holder was to reduce the sources of possible contaminations and, thus, poisoning of the fuel cell. For this reason all materials were selected for their high chemical and thermal resistance. The lower base plate and the fluid pipes are in AISI 316 stainless steel and are pressure-joined by accurate matching the holes diameters and without any glue. These couplings have been tested to be leak-free up to 7.5 bar of N2 pressure.
Along with the platform, a punch press was also designed and manufactured for reproducible perforation of the membranes, where fuel and oxygen inlets/outlets had to connect.
Sealing
µ-FC were sealed with Nafion PEM, by hot embossing with multiple Pulsed-NIL pulses. The process consisted of pressing the Nafion membrane with a flat stamp with integrated heater, and applying a series of strong heating pulse of 50 µs of 10 J/cm2/pulse to the stamp.
Leak tests
In the first test, a microscope glass slide has been cut to the same dimensions of a chip (22 x 22 mm²) and placed inside the holder. Then compressed nitrogen with a pressure of 7.5 bar has been supplied to all four pipes and then isolated from the supply line by a manual valve. The pressure of each branch of the system was monitored for 3 days without any detectable drop of pressure.
A functional test was performed by assembling a chip in the holder and applying nitrogen to a pressure of up to 4 bar to only one of the four pipes while the open ends of the other three pipes were let out into beakers with water in order to detect even very small flows of gas. A considerable flow of gas (1 bubble of about 4 mm³ per second) was detected at the exit of the first circuit, while no bubbles were detected at either end of the other circuit. Then the exit of the first circuit was closed with a manometer and we waited until the same pressure of the inlet was reached. Again, no bubbles were detected at the ends of the other circuit. The same procedure was applied to the second circuit with the same results. Thus we considered the test as passed.
Dielectric test
The second test aimed at checking the electrical insulation of the four cables with respect to each other and to the chip holder body. Being coaxial cables, we tested separately the inner conductor and the external shield. A tension up to 200 V dc has been applied across to each combination of single conductor couples with no detectable current flow.
Pulsed nanoimprint lithography fabrication process
The technology of the Pulsed-NIL has been used in two specific steps of the µ-FC fabrication process, namely in the patterning of the microfluidic channels in parallel for all the 12 chips of a 4” wafer, and for the sealing, chip by chip, of the proton exchange membrane (Nafion) onto the patterned substrate of the chips. The patterning of the microfluidic circuits of the chip is obtained by Pulsed-NIL process with the Pulsed-NIL process.
On-chip fuel cell measurements
The platform was afterwards used for measurements of several chips of the simple channel design and one resulted in a stable output with power maximum around 8µW. However low the power, this was noted as a successful step on the way to the on-chip micro-FC.
Pilot production line proposal
Table: Throughput of the different process steps for the fabrication of the µ-FCs. The table shows the throughput at the present stage, and provides the estimation for the production in a Pilot Line.
Process step Technology Present Throughput Projected Throughput in a Pilot Line
Microfluidic chip Patterning Pulsed-NIL 14 w/h 120 w/h
Pattern transfer into silicon ICP-RIE (Bosh process) 3 w/h 6-8 w/h
Deposition of Pt Catalyst Sputtering 6 w/h 60 w/h
Patterning of the metal layer PhLith or Pulsed-NIL 15 w/h 120 w/h
Sealing Pulsed-NIL 20 chips/h 120 w/h
It is clear that the main bottleneck of the fabrication process, both at the present stage of the development and for an estimated pilot production is the process of deep etching of the microfluidic structures into silicon. Therefore, in order to have all steps at the same throughput level, in production more ICP-RIE system should be running in parallel to advance the batch at comparable speed with the other steps. However, this has a disadvantage, represented by the high cost of modern ICP-RIE systems, and the cost of a process involving vacuum systems.
An alternative to the high investment and running costs and the low throughput of ICP-RIE systems, would be the fabrication of the microfluidic base of the µ-FCs chips in high-melting and high useful operating temperature thermoplastic materials, such as Polyether ether ketone (PEEK), Polyethersulphone (PES), Polyphenylene sulfide (PPS). Making µ-FCs by processes like injection molding, hot embossing, or Pulsed-NIL would have the advantage of higher productivity and lower costs of equipment, of the row materials, and running costs.
* Conclusion
The chipCAT project produced 71 articles published in peer-reviewed journals, 30 of them being collaborative articles between the Consortium’s academic research groups, and with the knowledge embodied there the Consortium reached both applied goals of the project: a) Thin-film catalysts for both electrodes with very small amounts of Platinum were developed into nanostructured layers that in proton exchange membrane fuel cells near the performance goals of the automotive industry (0.5 – 1 W/cm2) at a greatly reduced Platinum loading. b) Proof-of-concept prototypes of on-chip micro-fuel cells were developed by realization of several of their key enabling ideas, they were assembled and successfully tested. The teams involved consider the project completely successful, the primary cause of which was the balance between fundamental and applied research within.
*Publications
[1] Martin Dubau, Jaroslava Lavková, Ivan Khalakhan, Stanislav Haviar, Valerie Potin, Vladimír Matolín, and Iva Matolínová, Preparation of Magnetron Sputtered Thin Cerium Oxide Films with a Large Surface on Silicon Substrates Using Carbonaceous Interlayers, ACS Appl. Mater. Interfaces, 6 (2), 2013, 1213-1218, 10.1021/am4049546
[2] Vitalii Stetsovych, Federico Pagliuca, Filip Dvořák, Tomáš Duchoň, Mykhailo Vorokhta, Marie Aulická, Jan Lachnitt, Stefan Schernich, Iva Matolínová, Kateřina Veltruská, Tomáš Skála, Daniel Mazur, Josef Mysliveček, Jörg Libuda, and Vladimír Matolín, Epitaxial Cubic Ce2O3 Films via Ce–CeO2 Interfacial Reaction, J. Phys. Chem. Letters, 4 (6), 2013, 866-871, 10.1021/jz400187j
[3] Ghosh P., M. Farnesi Camellone, and S. Fabris, Fluxionality of Au clusters at ceria surfaces during CO oxidation: relationships among reactivity, size, cohesion, and surface defects from DFT simulations, J. Phys. Chem. Letters, 4 (14), 2013, 2256-2263, 10.1021/jz4009079
[4] Karolina Kwapien, Simone Piccinin, and Stefano Fabris, Energetics of Water Oxidation Catalyzed by Cobalt Oxide Nanoparticles: Assessing the Accuracy of DFT and DFT+U Approaches Against Coupled Cluster Methods, J. Phys. Chem. Letters, 4 (24), 2013, 4223-4230, 10.1021/jz402263d
[5] Paul Jennings, Hristiyan A. Aleksandrov, Konstantin M. Neyman, and Roy L. Johnston, A DFT study of oxygen dissociation on platinum based nanoparticles, Nanoscale, 6, 2013, 1153-1165, 10.1039/C3NR04750D
[6] Sergey M. Kozlov, Hristiyan A. Aleksandrov, Jacek Goniakowski, and Konstantin M. Neyman, Effect of MgO(100) support on structure and properties of Pd and Pt nanoparticles with 49-155 atoms, J. Chem. Phys, 139, 2013, 084701, 10.1063/1.4817948
[7] Bruix A., Y. Lykhach, I. Matolínová, A. Neitzel, T. Skála, N. Tsud, M. Vorokhta, V. Stetsovych, K. Ševčíková, J. Mysliveček, K. C. Prince, S. Bruyère, V. Potin, F. Illas, V. Matolín, J. Libuda, and K. M. Neyman, Maximum Noble Metal Efficiency in Catalytic Materials: Atomically Dispersed Surface Platinum, Angewandte Chemie Intl. Ed., 53 (39), 2014, 10525-10530, 10.1002/anie.201402342
[8] Hristiyan A. Aleksandrov, Sergey M. Kozlov, Swetlana Schauermann, Georgi N. Vayssilov, Konstantin M. Neyman, How Absorbed Hydrogen Affects the Catalytic Activity of Transition Metals, Angewandte Chemie Intl. Ed., 53 (49), 2014, 13371-13375, 10.1002/anie.201405738
[9] Yaroslava Lykhach, Armin Neitzel, Klára Ševčíková, Viktor Johánek, Nataliya Tsud, Tomáš Skála, Kevin C. Prince, Vladimír Matolín, and Jörg Libuda, The Mechanism of Hydrocarbon Oxygenate Reforming: C—C Bond Scission, Carbon Formation, and Noble-Metal-Free Oxide Catalysts, Chem. Sus. Chem., 7 (1), 2014, 77-81, 10.1002/cssc.201301000
[10] Sevcikova, K; Nehasil, V; Zahoranova, T; Vorokhta, M; Tsud, N; Yoshikawa, H; Kobata, M; Kobayashi, K ; Matolin, V, The effect of the substrate on thermal stability of CeOx and Rh–Ce–O thin films, Surf. Interface Anal., 46 (10-11), 2014, 980-983, 10.1002/sia.5503
[11] Armin Neitzel , Yaroslava Lykhach , Viktor Johánek , Nataliya Tsud , Tomáš Skála , Kevin C. Prince , Vladimír Matolín , and Jörg Libuda, Role of Oxygen in Acetic Acid Decomposition on Pt(111), J. Phys. Chem. C, 118 (26), 2014, 14316-14325, 10.1021/jp502017t
[12] Sergey M. Kozlov, Hristiyan A. Aleksandrov, Konstantin M. Neyman, Adsorbed and Subsurface Absorbed Hydrogen Atoms on Bare and MgO(100)-Supported Pd and Pt Nanoparticles, J. Phys. Chem. C, 118 (28), 2014, 15242-15250, 10.1021/jp502575a
[13] Farnesi Camellone M., D Marx, Nature and role of activated molecular oxygen species at the gold/titania interface in the selective oxidation of alcohols, J. Phys. Chem. C, 118, 2014, 20989-21000, 10.1021/jp5060233
[14] Fabio R. Negreiros, Stefano Fabris, Role of Cluster Morphology in the Dynamics and Reactivity of Subnanometer Pt Clusters Supported on Ceria Surfaces, J. Phys. Chem. C, 118 (36), 2014, 21014-21020, 10.1021/jp506404z
[15] Ma C., S. Piccinin, S. Fabris, Interface structure and reactivity of water-oxidation Ru-polyoxometalate catalysts on functionalized graphene electrodes, Phys. Chem. Chem. Phys., 16, 2014, 5333-5341, 10.1039/C3CP54943G
[16] Sergey M. Kozlov , Konstantin M. Neyman, O vacancies on steps on the CeO2(111) surface, Physical Chemistry Chemical Physics, Vol. 16/Issue 17, 2014, 7823-7829, 10.1039/c4cp00136b
[17] Neitzel A., Y. Lykhach, T. Skála, N. Tsud, V. Johánek, M. Vorokhta, K. C. Prince, V. Matolín, J. Libuda, Hydrogen activation on Pt-Sn nanoalloys supported on mixed Sn-Ce oxide films, Physical Chemistry Chemical Physics, 16, 2014, 13209-13219, 10.1039/C4CP01632G
[18] Paul C. Jennings, Hristiyan A. Aleksandrov, Konstantin M. Neyman, Roy L. Johnston, DFT studies of oxygen dissociation on the 116-atom platinum truncated octahedron particle, Physical Chemistry Chemical Physics, 16 (48), 2014, 26539-26545, 10.1039/c4cp02147a
[19] Armin Neitzel , Yaroslava Lykhach , Tomáš Skála , Nataliya Tsud , Mykhailo Vorokhta , Daniel Mazur , Kevin C. Prince , Vladimír Matolín , Jörg Libuda, Surface sites on Pt–CeO2 mixed oxide catalysts probed by CO adsorption: a synchrotron radiation photoelectron spectroscopy study, Physical Chemistry Chemical Physics, 16 (45), 2014, 24747-24754, 10.1039/c4cp03346a
[20] Mahasin Alam Sk, Sergey M. Kozlov, Kok Hwa Lim, Annapaola Migani, Konstantin M. Neyman, Oxygen vacancies in self-assemblies of ceria nanoparticles, J. of Materials Chemistry A, 2 (43), 2014, 18329-18338, 10.1039/C4TA02200A
[21] Stanislav Haviar, Martin Dubau, Jaroslava Lavková, Ivan Khalakhan, Valérie Potin, Vladimír Matolín, Iva Matolínová, Investigation of Growth Mechanism of Thin Sputtered Cerium Oxide Films on Carbon Substrates, Science of Advanced Materials, 6 (6), 2014, 1278-1285, 10.1166/sam.2014.1905
[22] Firas Faisal, Arafat Toghan, Ivan Khalakhan, Mykhailo Vorokhta, Vladimír Matolín, Jörg Libuda, Characterization of thin CeO2 films electrochemically deposited on HOPG, Applied Surface Science, 350, 2015, 142-148, 10.1016/j.apsusc.2015.01.198
[23] Fiala R., M. Václavů, A. Rednyk, I. Khalakhan, J. Lavková, V. Potin, I. Matolínová, V. Matolín, Pt–CeOx thin film catalysts for PEMFC, Catalysis Today, 240 Part B, 2015, 236-241, 10.1016/j.cattod.2014.03.069
[24] Beran, J; Matolin, V; Masek, K, RHEED structural study of the novel tin-cerium oxide catalyst, Ceramics Intl, 41 (3) Part B, 2015, 4946-4652, 10.1016/j.ceramint.2014.12.057
[25] Fiala R., M. Václavů, M. Vorokhta, I. Khalakhan, J. Lavková, V. Potin, I. Matolínová, V. Matolín, Proton exchange membrane fuel cell made of magnetron sputtered Pt–CeOx and Pt–Co thin film catalysts, Journal of Power Sources, 273, 2015, 105-109, 10.1016/j.jpowsour.2014.08.093
[26] Potin V., J. Lavkova, S. Bourgeois, M. Dubau, I. Matolinova and V. Matolin, Structural and chemical characterization of cerium oxide thin layers grown on silicon substrate, Materials Today: Proceedings, 2 (1), 2015, 101-107, 10.1016/j.matpr.2015.04.014
[27] Avril L., N. Zanfoni, P. Simon, L. Imhoff, S. Bourgeois, B. Domenichini, MOCVD growth of porous cerium oxide thin films on silicon substrate, Surface and Coatings Technology, 280, 2015, 148-153, 10.1016/j.surfcoat.2015.07.055
[28] Zanfoni N., L. Avril, L. Imhoff, B. Domenichini, S. Bourgeois, Direct liquid injection chemical vapor deposition of platinum doped cerium oxide thin films, Thin Solid Films, 589, 2015, 246-251, 10.1016/j.tsf.2015.05.037
[29] Avril L., S. Bourgeois, P. Simon, B. Domenichini, N. Zanfoni, F. Herbst, L. Imhoff, Nanostructured Pt–TiO2 composite thin films obtained by direct liquid injection metal organic chemical vapor deposition: Control of chemical state by X-ray photoelectron spectroscopy, Thin Solid Films, 591B, 2015, 237-244, 10.1016/j.tsf.2015.06.007
[30] Nazon J., M. Herbst, M.C. Marco de Lucas, S. Bourgeois, B. Domenichini, WC-based thin films obtained by reactive radio-frequency magnetron sputtering using W target and methane gas, Thin Solid Films, 591B, 2015, 119-125, 10.1016/j.tsf.2015.08.035
[31] Armin Neitzel, Yaroslava Lykhach, Viktor Johánek, Nataliya Tsud, Tomáš Skála, Kevin C. Prince, Vladimír Matolín, Jörg Libuda, Decomposition of Acetic Acid on Model Pt/CeO2 Catalysts: The Effect of Surface Crowding, J. Phys. Chem. C, 119 (24), 2015, 13721-13734, 10.1021/acs.jpcc.5b03079
[32] Negreiros F.R. M. Farnesi Camellone, S. Fabris, Effects of thermal fluctuations on the hydroxylation and reduction of ceria surfaces by molecular H2, J. Phys. Chem. C, 119 (37), 2015, 21567-21573, 10.1021/acs.jpcc.5b07030
[33] Lucie Szabová, Yoshitaka Tateyama, Vladimír Matolín, Stefano Fabris, Water Adsorption and Dissociation at Metal-Supported Ceria Thin Films: Thickness and Interface-Proximity Effects Studied with DFT+U Calculations, J. Phys. Chem. C, 119 (5), 2015, 2537-2544, 10.1021/jp5109152
[34] Paul C. Jennings, Hristiyan A. Aleksandrov, Konstantin M. Neyman, Roy L. Johnston, O2 Dissociation on M@Pt Core−Shell Particles for 3d, 4d, and 5d Transition Metals, J. Phys. Chem. C, 119 (20), 2015, 11031-11041, 10.1021/jp511598e
[35] Sergey M. Kozlov, Hristiyan A. Aleksandrov, Konstantin M. Neyman, Energetic Stability of Absorbed H in Pd and Pt Nanoparticles in a More Realistic Environment, J. Phys. Chem. C, 119 (9), 2015, 5180-5186, 10.1021/jp513022m
[36] Jaroslava Lavková, Ivan Khalakhan, Mykhailo Chundak, Mykhailo Vorokhta, Valerie Potin, Vladimír Matolín, Iva Matolínová, Growth and composition of nanostructured and nanoporous cerium oxide thin films on a graphite foil, Nanoscale, 7, 2015, 4038-4047, 10.1039/C4NR06550F
[37] Sergey M. Kozlov, Ilker Demiroglu, Konstantin M. Neyman, Stefan T. Bromley, Reduced ceria nanofilms from structure prediction, Nanoscale, 7 (10), 2015, 4361-4366, 10.1039/C4NR07458K
[38] Sergey M. Kozlov, Gábor Kovács, Riccardo Ferrando, Konstantin M. Neyman, How to determine accurate chemical ordering in several nanometer large bimetallic crystallites from electronic structure calculations, Chemical Science, 6, 2015, 3868-3880, 10.1039/C4SC03321C
[39] Gábor Kovács, Sergey M. Kozlov, Iva Matolínová, Mykhailo Vorokhta, Vladimír Matolín, Konstantin M. Neyman, Revealing chemical ordering in Pt–Co nanoparticles using electronic structure calculations and X-ray photoelectron spectroscopy, Physical Chemistry Chemical Physics, 17, 2015, 28298-28310, 10.1039/C5CP01070E
[40] Hristiyan A. Aleksandrov, Konstantin M. Neyman, Georgi N. Vayssilov, The structure and stability of reduced and oxidized mononuclear platinum species on nanostructured ceria from density functional modelling, Physical Chemistry Chemical Physics, 17 (22), 2015, 14551-14560, 10.1039/C5CP01685A
[41] Nguyen M.T. M. Farnesi Camellone, R. Gebauer, On the electronic, structural, and thermodynamic properties of Au supported on α-Fe2O3 surfaces and their interaction with CO, J. Chem. Phys., 143, 2015, 034704, 10.1063/1.4926835
[42] Piccinin S., S. Fabris, Water oxidation by Ru-polyoxometalate catalysts: Overpotential dependency on the number and charge of the metal centers, Inorganics, 3, 2015, 374-387, 10.3390/inorganics3030374
[43] Myslivecek, J; Matolin, V; Matolinova, I, Heteroepitaxy of Cerium Oxide Thin Films on Cu(111), Materials, 8 (9), 2015, 6346-6359, 10.3390/ma8095307
[44] Khalakhan, I; Fiala, R; Lavková, J; Kúš, P; Ostroverkh, A; Václavů, M; Vorokhta, M; Matolínová, I; Matolín, V, Candle Soot as Efficient Support for Proton Exchange Membrane Fuel Cell Catalyst, Fuel Cells, 16 (5), 2016, 652-655, 10.1002/fuce.201600016
[45] Neitzel A., G. Kovács, Y. Lykhach, S.M. Kozlov, N. Tsud, T. Skála, M. Vorokhta, V. Matolín, K.M. Neyman, J. Libuda, Atomic ordering and Sn segregation in Pt-Sn nanoalloys supported on CeO2 thin films, Top. Catal., online only, 2016, , 10.1007/s11244-016-0709-5
[46] Monai M.; Montini, T; Melchionna, M; Duchoň, T; Kúš, P; Tsud, N; Prince, KC; Matolin, V; Gorte, RJ; Fornasiero, P, Phosphorus poisoning during wet oxidation of methane over Pd@CeO2/graphite model catalysts, Appl. Catal. B-Environ., 197, 2016, 271-279, 10.1016/j.apcatb.2015.10.001
[47] Fiala R., Figueroba, A; Bruix, A; Vaclavu, M; Rednyk, A; Khalakhan, I; Vorokhta, M; Lavkova, J; Illas, F; Potin, V; Matolinova, I; Neyman, KM; Matolin, V, High efficiency of Pt2+-CeO2 novel thin film catalyst as anode for proton exchange membrane fuel cells, Appl. Catal. B-Environ., 197, 2016, 262-270, 10.1016/j.apcatb.2016.02.036
[48] Vorokhta M., I. Khalakhan, M. Václavu, G. Kovács, S.M. Kozlov, P. Kúš, T. Skála, N. Tsud, J. Lavková, V. Potin, I. Matolínová, K.M. Neyman, V. Matolín, Surface composition of magnetron sputtered Pt-Co thin film catalyst for proton exchange membrane fuel cells, Appl. Surf. Sci., 365, 2016, 245-251, 10.1016/j.apsusc.2016.01.004
[49] Neitzel A., V. Johánek, Y. Lykhach, T. Skála, N. Tsud, M. Vorokhta, V. Matolín, J. Libuda, Reduction of Pt2+ species in model Pt—CeO2 fuel cell catalysts upon reaction with methanol, Appl. Surf. Sci., 387, 2016, 674-681, 10.1016/j.apsusc.2016.06.156
[50] Wolfbeisser A., G. Kovács, S.M. Kozlov, K. Föttinger, J. Bernardi, B. Klötzer, K.M. Neyman, G. Rupprechter. , Surface composition changes of CuNi-ZrO2 catalysts during methane decomposition: An operando NAP-XPS and density functional study, Catal. Today, online only, 2016, , 10.1016/j.cattod.2016.04.022
[51] Khalakhan I., M. Vorokhta, M. Václavů, B. Šmíd, J. Lavková, I. Matolínová, R. Fiala, N. Tsud, T. Skála, V. Matolín, In-situ electrochemical atomic force microscopy study of aging of magnetron sputtered Pt-Co nanoalloy thin films during accelerated degradation test, Electrochim. Acta, 211, 2016, 52-58, 10.1016/j.electacta.2016.06.035
[52] Rednyk A; Johanek, V; Khalakhan, I; Dubau, M; Vorokhta, M; Matolin, V, Methanol oxidation on sputter-coated platinum oxide catalysts, Int. J. Hydrog. Energy, 41 (1), 2016, 265-275, 10.1016/j.ijhydene.2015.09.147
[53] Simon P., N. Zanfoni, L. Avril, J. Lavkova, I. Matolinova, L. Imhoff, V. Potin, B. Domenichini, S. Bourgeois, Observation of surface reduction in porous ceria thin film grown on graphite foil substrate, Materials Today: Proceedings, 3 (8), 2016, 2772-2779, 10.1016/j.matpr.2016.06.026
[54] Ševčíková K., L. Szabová, M. Kettner, P. Homola, N. Tsud, S. Fabris, V. Matolín, V. Nehasil, Experimental and theoretical study on the electronic interaction between Rh adatoms and CeOx substrate in dependence on a degree of cerium oxide reduction, J. Phys. Chem C, 120, 2016, 5468-5476, 10.1021/acs.jpcc.5b11431
[55] Neitzel A., A. Figueroba, Y. Lykhach, T. Skála, M. Vorokhta, N. Tsud, S. Mehl, K. Ševčíková, K.C. Prince, K.M. Neyman, V. Matolín, J. Libuda, Atomically Dispersed Pd, Ni and Pt Species in Ceria-Based Catalysts: Principal Differences in Stability and Reactivity, J. Phys. Chem. C, 120, 2016, 9852-9862, 10.1021/acs.jpcc.6b02264
[56] Brummel O., F. Waidhas, F. Faisal, R. Fiala, M. Vorokhta, I. Khalakhan, M. Dubau, A. Figueroba, G. Kovács, H.A. Aleksandrov, G.N. Vayssilov, S.M. Kozlov, K.M. Neyman, V. Matolín, J. Libuda, Stabilization of Small Platinum Nanoparticles on Pt–CeO2 Thin Film Electrocatalysts During Methanol Oxidation, J. Phys. Chem. C, 120, 2016, 19723-19736, 10.1021/acs.jpcc.6b05962
[57] Ostroverkh, A; Johánek, V; Kúš, P; Šedivá, R; Matolín, V, Efficient Ceria–Platinum Inverse Catalyst for Partial Oxidation of Methanol, Langmuir, 32 (25), 2016, 6297–6309, 10.1021/acs.langmuir.6b01316
[58] Dvorak, F; Camellone, MF; Tovt, A; Tran, N-D; Negreiros, FR; Vorokhta, M; Skala, T; Matolinova, I; Myslivecek, J; Matolin, V; Fabris, S, Creating single-atom Pt-ceria catalysts by surface step decoration, Nat. Commun., 7, 2016, 10801 (8 pages), 10.1038/ncomms10801
[59] Yaroslava Lykhach, Sergey M. Kozlov, Tomáš Skála, Andriy Tovt, Vitaliy Stetsovych, Nataliya Tsud, Filip Dvořák, Viktor Johánek, Armin Neitzel, Josef Mysliveček, Stefano Fabris, Vladimír Matolín, Konstantin M. Neyman, Jörg Libuda, Counting electrons on supported nanoparticles, Nature Materials, 15, 2016, 284-288, 10.1038/NMAT4500
[60] Lykhach Y., A. Figueroba, M.F. Camellone, A. Neitzel, T. Skála, F.R. Negreiros, M. Vorokhta, N. Tsud, K.C. Prince, S. Fabris, K. Neyman, V. Matolín, J. Libuda, Reactivity of Atomically Dispersed Pt2+ Species towards H2: Model Pt−CeO2 Fuel Cell Catalyst, Phys. Chem. Chem. Phys., 18, 2016, 7672-7679, 10.1039/C6CP00627B
[61] Figueroba A., G. Kovács, A. Bruix, K.M. Neyman, Towards stable single-atom catalysts: strong binding of atomically dispersed transition metals on the surface of nanostructured ceria, Catal. Sci. Technol., 6, 2016, 6806-6813, 10.1039/C6CY00294C
[62] Neitzel A., G. Kovács, Y. Lykhach, N. Tsud, S.M. Kozlov, T. Skála, M. Vorokhta, V. Matolín, K.M. Neyman, J. Libuda, Steering the formation of supported Pt-Sn nanoalloys by reactive metal-oxide interaction, RSC Adv., 6, 2016, 85688-85697, 10.1039/C6RA18801J
[63] Viktor Johánek, Anna Ostroverkh, Roman Fiala, Andrii Rednyk, and Vladimír Matolín, Mass Spectrometry of Polymer Electrolyte Membrane Fuel Cells, Journal of Analytical Methods in Chemistry, 2016 online, 2016, 1-9, 10.1155/2016/6097285
[64] Khalakhan, I; Lavková, J; Matolínová, I; Vorokhta, M; Potin, V; Kúš, P; Václavů, M; Maraloiu, V-A; Kuncser, A-C; Matolín, V, Electrochemically shape-controlled transformation of magnetron sputtered platinum films into platinum nanostructures enclosed by high-index facets, Surf & Coat Technol, 309, 2017, 6-11, 10.1016/j.surfcoat.2016.11.017
[65] Sergey M. Kozlov, Konstantin M. Neyman, Insights from methane decomposition on nanostructured palladium, J. of Catalysis, 337, 2016, 111-121,10.1016/j.jcat.2016.02.010
[66] Camellone, MF; Ribeiro, FN; Szabova, L; Tateyama, Y; Fabris, S, Catalytic Proton Dynamics at the Water/Solid Interface of Ceria-Supported Pt Clusters, J. Am. Chem. Soc., 138 (36), 2016, 11560-11567, 10.1021/jacs.6b03446
[67] Sergey M. Kozlov, Konstantin M. Neyman, Effects of electron transfer in model catalyst composed of Pt nanoparticles on CeO2(111) surface, J. of Catalysis, 344, 2016, 207-5015, 10.1016/j.jcat.2016.10.014
[68] Aleksandrov H.A. K.M. Neyman, K.I. Hadjiivanov, G.N. Vayssilov, Can the state of platinum species be unambiguously determined by the stretching frequency of adsorbed CO probe molecule?, Phys. Chem. Chem. Phys., 18, 2016, 22108-22121, 10.1039/C6CP03988J
[69] Bruix A., Neyman K.M. Modeling ceria-based nanomaterials for catalysis and related applications, Catal. Lett., 164, 2016, 2053-2080, 10.1007/s10562-016-1799-1
[70] Ferstl, P.; Mehl, S.; Arman, M. A.; Schuler, M.; Toghan, A.; Laszlo, B.; Lykhach, Y.; Brummel, O.; Lundgren, E.; Knudsen, J; Hammer, L; Schneider, MA; Libuda, J, Adsorption and Activation of CO on Co3O4(111) Thin Films, J. Phys. Chem. C, 119 (29), 2015, 16688-16699, 10.1021/acs.jpcc.5b04145
[71] Mehl, S.; Ferstl, P.; Schuler, M.; Toghan, A.; Brummel, O.; Hammer, L.; Schneider, M. A.; Libuda, J, Thermal evolution of cobalt deposits on Co3O4(111): atomically dispersed cobalt, two-dimensional CoO islands, and metallic Co nanoparticles, Phys. Chem. Chem. Phys., 17 (36), 2015, 23538-23546, 10.1039/c5cp03922c
Potential Impact:
The chipCAT project has advanced the field of heterogeneous catalysis in several ways, all fundamental ones and most of the applied ones are published in peer-reviewed journals and will impact the field in the standard process of scientific verification, discourse and adoption. Among the fundamental results we consider most prominent the conclusive evidence of single-atom catalyst actual existence and function. This realization shifts the paradigm of catalysis and may have manifold effect in all parts of the catalytic domain.
The achievement of developing a class of low-Platinum content catalysts suitable for PEMFC anodes is likely to be appreciated, gradually, as hydrogen fuel cell cars become more in mainstream demand: Saving Platinum will be of essence, because its global supplies do not cover the potential demand that would be created by “all cars on HFCs”. Currently the pricing of HFC cars starts around €500 000, which means (a) that the global fleet of HFC cars can be relatively tiny in size, so the global amount of Platinum will not figure, (b) that the current price of Platinum per car (ca. €1000) does not figure much in the total car price.
At present, the more attractive aspects of the nanocatalyst technology are the “dry” physical deposition of the layers and the remarkable durability of the membrane assemblies. These could bring the catalysts into mass production and use, because they both translate into reduced total cost of the technology in the end-user products. The industry as a whole may not grasp the technology yet, though: Currently, only anodes can be produced by the “dry” physical deposition route, while cathodes still need “wet” steps to achieve the industry target power outputs. Hence, companies working with “fully wet-route” catalyst layers are more likely to stick to their production lines. Once the properly efficient cathodes can be manufactured by physical deposition – if it happens – we expect to observe a massive shift in the industrial approach.
The newly developed fuel cell testing stations marketed as products constitute a minor addition to the global market in itself. Considering their performance, though, they have a great potential of becoming the test instrument of choice in laboratory and development environment. The first items were sold to synchrotron beamline laboratories to control fuel cells in operando analyses. Since the testing stations were developed with operando analyses in mind and seeing how operando spectroscopies and microscopies are in vogue in current world of catalysis research, we expect the product to help accelerate the field considerably.
The on-chip micro-FC technology has passed the proof of concept stage and has been brought to TRL 3-4. It is likely to be developed further and eventually become an alternative in mobile energy generation. Presently it still needs to pass practical hurdles to reach the stage, when tangible impact can be foreseen.
List of Websites:
URL: http://www.chipcat.eu
Corresponding contact for the project:
prof. RNDr. Vladimír Matolín, DrSc., matolin@mbox.troja.mff.cuni.cz +420-95155-2323
The project chipCAT aimed at the knowledge-driven development of a novel type of thin-film catalysts for silicon-based “on-chip” micro fuel cells (u-FCs). Combining fundamental surface-science, model catalysis, and first-principles computational studies, led to a detailed understanding of the surface chemistry on complex nanostructured catalysts. This microscopic-level understanding was used to tailor active sites and their mutual interplay at the nanoscale in order to maximize activity and selectivity and to reduce deactivation and poisoning.
Starting from new oxide-based materials with minor demand for noble metals we explored new material concepts. Target structures defined by fundamental research were then transferred to standard FC and u-FC catalysis using advanced thin-film preparation techniques, such as magnetron sputtering and vacuum deposition. Atomic-level control, using a broad spectrum of surface spectroscopies, in-situ/operando spectroscopies and microscopies with atomic resolution provided many channels of communication between fundamental research, applied research and technology development. Using modern microtechnologies, the novel tailor-made catalyst materials were integrated into working FC test devices. Prototype devices were fabricated for performance tests, the results of which provided basis for drawing up a process for laboratory-scale production of u-FC batches, as well as feedback to fundamental research.
Surface science, model catalysis, state-of-the-art theory, thin-film-technology, applied heterogeneous catalysis, and microtechnology were connected in an interdisciplinary effort aiming at the development of a new generation of metal-oxide FC catalysts with improved performance and stability. This project allowed dramatic reduction of critical materials use in fuel cell anodes and an incremental one in cathodes. The on-chip fuel cell prototypes also opened the pathway to a novel and potentially useful energy storage technology for mobile devices.
The project RTD was organized in work packages 2 through 7 thus: Fundamental knowledge was being pursued in WP 2 (theoretical and computational) and WPs 3 through 5 (experimental). A vigorous exchange of findings took place between these work packages; WP3 with its focus on model experiments in particular formed a meeting point and bridge between theory and the rest of experimental work, WP4 acted as a bridge from experimental work of fundamental research to the application driven WPs 6 and 7. These latter WPs covered the pursuit of the set practical goals of the project: In WP6 we scrutinized new materials for catalytic properties in thin film forms, created with (mainly) vacuum physical methods and with nanofabrication we pursued their absolute power density (W/cm2) and specific power density (W/mgPt cm2), WP7 was dedicated to the complete design of microfluidic fuel cells (micro-FC) ready for integration in silicon wafer chips along with integrated circuits. The thin film catalysts refined in WP6 were to be used as low-Platinum catalysts in PEMFC with the usual “sandwich” construction and also in the novel on-chip micro-FC coming out of WP7. In both cases the ambition was to produce a laboratory prototype with measurable and stable power output, ideally in a competitive range of power densities.
Project Context and Objectives:
The chipCAT project (FP7 GA no. 310191) was established by research groups with successful record in fundamental research of catalysts for proton exchange membrane fuel cells (PEMFC) and partners from related industrial fields. Our principal aim was to continue the promising explorations of the missing fundamental knowledge in catalysis and, at the same time, utilize this in knowledge-based development of technologies for cheaper, more durable and more mobile industrial-level fuel cells.
The motivation for our work was to contribute, with our expert knowledge and facilities, to making substantially more affordable proton-exchange membrane fuel cells, which were staying at high prices owing to the amounts of Platinum, expensive catalyst production and FC assembly and limited lifetime of the catalysts. As the envisaged new thin-film catalyst technology would be compatible with integrated circuit devices in Silicon wafer chips, we proposed to make an attempt at application of the catalyst films in making a prototype of and on-chip microfluidic fuel cell.
These were the principal objectives:
• By way of model experimental and theoretical (computational) physics uncover the precise mechanisms behind the function of the Platinum-based low-noble metal content catalysts, especially in combination with reducible oxides;
• Transfer of the patented low-noble metal content catalyst preparation method from model to practicable industrial scale;
• Explore a selection of other low-Platinum or precious metal-free catalysts for their suitability for deposition by the developed physical method and in planar geometry.
• Develop a physical thin-film catalyst preparation method compatible with the standard planar processes in semiconductor device industry;
• Develop a complete functional prototype of a microfluidic PEMFC device;
The consortium Coordinated by the surface physics group at Charles University in Prague (CUP) engaged computational modelling, model and applied experiments and technology and device development to achieve the objectives.
Project Results:
Convention notwithstanding, the articles published by the project are cited/listed in the approximate order of their publication, not the appearance order. Full list of peer-review publications is appended to this section.
*Computational modelling
Work Package 2 covered theoretical and computational research of materials relevant to hydrogen fuel cell catalysis. Throughout the project this WP communicated vigorously with WP3 (model experimental) and in its most prominent results also with WPs 4 and 5 (applied experimental). The theory groups produced 26 articles on their own and, remarkably, 14 articles together with the experimental counterparts in various configurations.
Two theory groups were engaged in this work package: one of prof. Konstantin Neyman at the University of Barcelona (UBCN) and the other of prof. Stefano Fabris at the Italian National Research Council (CNR). These two teams tackled the catalytic materials using different, mutually complementary methods. The UBCN team worked on the basis of standalone nanocrystals (nanoparticles), their facet properties and interactions with atoms and ions. Their work spanned several areas: metal nanocrystallites (elemental, alloys) and metal oxide nanocrystallites (e.g. ceria) pure and mixed with metal atoms (e.g. Platinum). The CNR team dealt with interactions (e.g. water oxidation, formation and activity of metal nanoparticles) on metal oxide surfaces and metal/metal-oxide interfaces, with the optional accounting for finite pressure of water vapor on the surface (effectively studying a water-ceria interface).
Overview
The chief subject of fundamental research in the project was the understanding anode (hydrogen splitting) and cathode (oxygen reduction) reactions in PEMFC in the context of nanoparticular and atomically dispersed catalytic metals in reducible oxide matrix. The theoretical and collaborative research led to a substantial breakthrough in understanding of the atomically dispersed Pt-CeO2 system [7], when the project identified stable sites – {100} nanopockets – on ceria nanoparticle surface, where a single Pt atom was stabilized and, importantly, in 2+ oxidation state. This ionic state was already suspect from experimental work to being key to the material’s catalytic function. The theoretical achievement stemmed from a detailed study of the ceria nanoparticles alone [16,20] and the nanoparticle-approach-driven discovery in [7] prompted further advances purely theoretical [14,32,37,40,61,62,68,69] and in theory-experiment collaboration [47,56].
Another notable advance [58] was made later in the project by the slab-surface-based research which, in coupling with model experiment, showed that adsorption sites on naturally stepped flat surface of ceria offers site analogous to the {100} nanopockets that act the same way. This was a strong evidence that the atomic dispersion of Pt on ceria is not only possible, but occurs naturally at a rate that can be practically harnessed, just as had been indicated by the works of applied experimental research [23,25,47,64].
Nanoparticles of elemental metals and binary alloys
Platinum and other metal nanoclusters were a matter of attention as well as elemental metals [5,6,8,12,18,35,59,67] and binaries (alloys, core-shell) arrangements [34,38]. As it transpired that the atomically dispersed Pt-CeO2 system is suitable for anodes, but not efficient in the cathode-side oxygen reduction reaction (ORR), these studies of Pt and binary nanoparticles fuelled progress in the understanding the cathode mechanisms and developing the optimal catalysts for ORR.
The theoretical research of electronic structure of bimetallic alloy nanoparticles [38] was another major advance made by the team of prof. Konstantin Neyman and a concept of alloy nanoparticle computational screening for catalytic properties derived from that work was filed, in the last year of the project, for patenting in the U.S. as the most likely market for such technology.
Throughout, there was a persistent question of particle-support interaction in the metal/oxide arrangement, as it is well known that, depending on the interaction strength, properties of the nanoparticles can be influenced or even determined by this interaction. The effort to determine the strength of the interaction gave an important result [59], where model experiments heavily backed by theory produced a method to determine the amount of charge transfer between particle and support as a function of particle size.
Ceria and atomically dispersed metals
Platinum in the form of nanoclusters supported by ceria was studied in parallel [14,66] so as to reveal the dynamics of coalescence of Pt atoms and to understand, why in some scenarios the dispersed Pt atoms do not tend to form stable clusters. As the coalescence of Pt is among the leading causes of Pt catalyst ageing and fuel cell degradation, these researches were of supreme importance and tied the molecular dynamic simulations to experiments both in model (ultra-high vacuum) and in operando (ambient vapour pressure, electrolyte) conditions.
In the rank of dispersed metals (atomically or in nanoclusters) in reducible oxide matrix there was an amount of necessary research done apart from the Pt-CeO2 combination. Other metal atoms were tested with respect to the {100} nanopocket site and the results of these calculations were summarized in [61].
In parallel non-Platinum atoms and clusters were studied, supported by ceria and other oxides, in theory and in experiments: Au/CeO2 [3], Au/TiO2 [13], Rh/CeO2 [54], Pt/MgO, Pd/MgO [12] and others. These researches had multiple motivation: the fundamental understanding, the practice-aware push for minimizing or zeroing the need for precious metals in hydrogen fuel cells, and another practical effort to find ways to reduce the catalysts’ susceptibility to poisoning.
Other materials
While most of the project addressed materials for hydrogen fuel cells, a dedicated attention was given to materials for hydrocarbon oxidation in the latter half of the project [46,50,65,70,71] such as palladium and cobalt oxide. These are of interest, for example, for anodes of direct methanol fuel cells (DMFC), which are a major “mainstream” alternative to hydrogen fuel cells (HFC). As the HFC advances of the project were satisfactory throughout chipCAT, however, there was no material utilization of the new methanol-oxidation-related knowledge in the project.
*Model catalysts
Work Package 3 was the project hub for fundamental experimental research. Prof. Jörg Libuda at the Friedrich-Alexander University of Erlangen-Nuremberg (FAU) led this WP and his group made much of the experimental work on model materials in UHV and in operando conditions. Their focus was on model materials in planar geometry used for revealing mechanisms of adsorption and surface reactions of H2, O2, CO (i.e. molecules most relevant for hydrogen PEMFC) on the novel materials. In model studies, other small test molecules (methanol, ethanol, acetic acid, ethylene glycol,...) were also used in probing the material properties with controlled adsorption/desorption and surface sensitive spectroscopies: This was mainly a broad range of photoelectron spectroscopies in UHV and using synchrotron radiation, and a number of infrared spectroscopy methods using existing and newly installed instrumentation (in-situ/operando IRAS, phase-modulation IRAS, operando microreactor with ATR and IRAS mode).
One major line of research was on the different phases of ceria [2], Platinum [11] and atomically dispersed metals and metal nanoclusters on ceria. It was the multiple synchrotron radiation photoelectron spectroscopy experiments with model Pt/CeO2 and Pt-CeO2 materials (the different notation refers to Pt nanoclusters on ceria and Pt atomically dispersed on (and in) ceria, respectively) that were the first to reveal the presence and significance of the Pt2+ ionic state in ceria matrix [19]. The same type of experiments gave a substantial basis to deciphering the mechanisms subsequently [7,31,49,55,60].
In parallel, properties of similar ceria-based materials were studied in response to the call for finding the limit for reducing Pt (and Pt group metal) content, if it existed, and to make the Pt-based catalysts more resistant to poisoning by contaminants, mainly by CO. In the latter line of research it was tin (Sn) that received substantial attention as a promising additive to reduce the catalysts’ susceptibility to poisoning by CO.
As described in detail in [17], Sn retains the oxidation state +2 which is associated with the formation of a mixed Pt-Sn-CeO2 oxide. At high Sn concentrations, we detected a small amount of metallic Sn0. The formation of the Sn-CeO2 mixed oxide is accompanied by an increase of the RER associated with the increase of the Ce3+/Ce4+ ratio in the film. The nature of the Pt* species at 72 eV is not entirely clear. Earlier we assigned the corresponding species at 72 eV on Pt-CeO2 films to PtO. The intensity of the corresponding Pt* peak increases during annealing to 700 K at the expense of Pt2+. The rather small binding energy shift of Pt* peak (~0.4 eV) upon annealing suggests high stability of the PtO species.
Interestingly, annealing of the film also triggers the decrease of the metallic Sn0 which is accompanied by the increase of Sn2+ intensity. The decrease of the Sn0 is accompanied by a slight decrease of the RER. This suggests that metallic Sn0 is converted into Sn2+ followed by its diffusion into the bulk of Pt-Sn-CeO2 oxide during annealing. In conclusion, the Sn atoms were shown to be rather mobile, able to form a complex (Sn,Ce)O2 oxide as well as Pt-Sn binary nanoclusters (alloys and core-shell) and to transition between the states [17,24,45,62].
In the second period work continued on the Pt-CeO2 and Pt-(Sn,Ce)O2 systems, as mentioned above with references to publications. In addition, the search for alternative catalytic materials meant that some elements (Pd, Ni [55]) were tested in conjunction with ceria in order to find the possibility to disperse them atomically like Platinum and to reveal their function in such state.
In as much as Pd is concerned, it was concluded that one characteristic difference between the two metals is the lower thermal stability of Pd2+ sites as compared to Pt2+. This difference was associated with the fact that Pt2+ is preferentially stabilized at the square oxygen pocket sites at the surface, whereas Pd can be stabilized both at the surface and in the form of a solid solution in the bulk. In contrast to the Pt−CeO2, the Pd−CeO2 films maintain a considerably higher degree of reduction at even lower dopant concentration. The higher degree of reduction in the Pd−CeO2 films is likely caused by the formation of the PdO phase. The charge transfer between Pd and Ce4+ is accompanied by the withdrawal of lattice oxygen and growth of PdO particles. Consequently, the decomposition of PdO during annealing results in re-oxidation of Ce3+.
We studied the Ni 2p5/2 spectra obtained from 13% Ni−CeO2. The main peak at 855.0-855.3 eV is associated with Ni2+ ions. Additionally, two satellite features I and II emerge at 857.0 eV and 861.1 eV, respectively. Satellite I arises from the non-local screening effects and, therefore, is sensitive to the degree of Ni dispersion in CeO2. In particular, the low intensity of the satellite I is consistent with the high degree of Ni dispersion in CeO2 film. Satellite II emerges due to a shake-up process associated with the O 2p→Ni 3p charge-transfer multielectron excitation during ionization of Ni 2p core level. Annealing of Ni−CeO2 results in a decrease of both satellites I and II which is consistent with enhanced diffusion of Ni2+ towards the bulk. The Ni−CeO2 films retain a low degree of reduction, similar to that detected for the Pt−CeO2 mixed oxide. Unlike the Pd−CeO2, we identified only one phase associated with Ni2+ in CeO2. Since Ni content in 13% Ni−CeO2 is close to the solubility limit, it is possible that the presence of satellites I and II indicate the existence of small amount of a NiO phase. The disappearance of the satellites I and II from the Ni 2p spectra suggests diffusion of Ni2+ into the substrate and, most likely, decomposition of the NiO phase during annealing in UHV. This observation suggests that a single phase associated with a solid solution of Ni2+ is formed for 13% Ni−CeO2 above 600 K.
In summary, Pt2+, Pd2+, and Ni2+ were found to be stabilized in atomically dispersed form on CeO2 in the limit of low concentration. However, the thermal stabilities of Pd-CeO2 and Ni−CeO2 are compromised by the diffusion of Pd2+ and Ni2+ into the bulk upon annealing above 600 K. This is in contrast to the Pt-CeO2 which is stable at these conditions.
As an alternative material for hydrogen dissociation, cobalt oxide Co3O4 was also studied as model material of Co3O4 thin films on Ir(100) substrate under UHV FAU [70,71]. The films of Co3O4(111), stoichiometric and with surplus Co in atomic and nanocluster form, were characterized structurally and studied using IRAS with adsorption of CO and O2. By this latter method [70] the adsorption sites were revealed for low coverage (surface Co2+ sites), high coverage (sites between Co2+ ions) and bridging carbonate species associated with crystal defects.
In the second half of the project a substantial line of research focused on the cathode-side reactions and materials, namely binary nanoparticles of Pt-Ni and Pt-Co alloys. The Pt-Co nanoparticles were studied under UHV conditions and their behaviour under annealing was tracked [39] in HAXPES. It was concluded that Pt shell tends to form on the surface and there it prefers to occupy edge and vertex positions, while Co atoms linger on the facet insides. Subsurface layer consisted predominantly of Co atoms and the core remained Pt-Co, so that the discovered preferred arrangement was PtYCo1-Y-core@Co-rich-subsurface@Pt-rich_shell.
It had been known from before that in electrochemical environment (in electrolyte under potential cycling) the Ni and Co atoms are leached out of the particles. This was the case in several kinds of particles, including those with core-Pt_shell structure. Pt-Ni thin films were cycled in electrochemical AFM with increasing cycling potential limit, which was alternated with surface morphology imaging.
Strong dissolution effects result in a fast degradation of Pt-Ni catalyst upon cycling above the upper potential limit of 1.2 V. The degradation mechanism involves oxidation of Pt-shell followed by leaching of Ni core.
Dynamic changes of the surface sites were also monitored, this by means of IR spectroscopy using CO molecule as a probe. It was found that migration of Ni atoms from the bulk to the surface/subsurface region results in the formation of Pt-Ni species that influences the adsorption properties of CO molecule. Upon cycling to higher potentials, these species dissolve recovering the composition of Pt-rich shell.
*Thin films
Large surfaces
Two tasks of Work Package 4 were started at the commencement of the project and they concerned the refinement of the thin film growth techniques for uniform coverage of large (several centimetres squared) substrates and the first use of ALD as an alternative to magnetron sputtering. Most of the research in WPs 4 and 5 was carried out by the teams of Charles University in Prague (Coordinator) led by prof. Vladimír Matolín and of the University of Burgundy in Dijon led by prof. Sylvie Bourgeois and prof. Bruno Domenichini.
In scope of the first task an in communication with WP 6 NafionTM and AquivionTM membranes were etched by oxygen plasma. In AFM micrographs a desirable roughening was observed without any obvious degradation that might signify a compromised membrane function. These membranes, coated with 10nm and 30nm thick layers of Pt were passed on to WP 6 for testing in MEAs. Those tests, however, brought no significant improvement of performance or increase in active surface; the only clear change was a decrease in membrane resistance. It had been observed before that layers of Platinum oxide (PtOx) subsequently reduced by Hydrogen or CO exposure have noticeably rougher morphology, even at room temperature. This could be a fuel cell friendly “activation procedure”, but layers created by reduction of PtOx performed badly in FC tests.
Ceria
Structure of the 2 µgPt/cm2 catalyst film deposited on nano GDL (nGDL) was observed in STEM and HRTEM micrographs. Catalyst coating has striped character, which is due to porosity of the carbon substrate showing the channels etched by oxygen plasma. The channels are filled with the nanocrystalline catalyst exhibiting particles 2 – 3 nm large. HRTEM shows presence of two types of nanoparticles of cerium oxide, CeO2 and Ce2O7, respectively. We compared the catalyst structure before and after hydrogen annealing simulating working conditions of the PEMFC anode. Before hydrogen annealing, CeO2 and Ce7O12 crystallites were found, whereas after annealing Ce7O12 are predominant after annealing. Cerium carbide (CeC2) particles were also observed at the interface with the nGDL layer.
Nanostructured catalyst films in fuel cells
Sputtering Pt-CeOx onto nGDL represented a leap-wise advance in surface area of the NCF. PEMFC tests with Pt-CeOx anodes showed that among them existed some optimal, relatively tiny loading of 2µg/cm2, which produced biggest power density. It was remarkable that the 2µg Pt-CeOx anode gave out power at 90% of the reference, while its Pt loading was 1000 times (three orders of magnitude) lower. In another test 2µg/cm2 of Pt was deposited onto nGDL without CeOx and the power output was 5.5 times lower than that of Pt-CeOx with the same loading.
Examination of the low-loading Pt-CeOx material in model planar geometry in Work Package 3 showed that (up to a certain relative amount of Pt) the hydrogen exposure shifts all Platinum in Pt-CeOx into the Pt2+ ionic state. At higher Pt ratio the Pt exists in Pt2+ ionic state as well as Pt0 metallic state (i.e. clusters).
Electrochemical cycling is a standard test of catalyst durability and we subjected our Pt-CeOx layers to it. One cycle consisted of a sequence of stable power delivery, short circuit and open circuit. The graph shows a remarkable endurance of the material, which indicates that the Pt2+ ionic state is very stable and the dispersed atoms do not have a tendency to coalesce and form metal clusters.
As a material for the cathode side of the PEMFC (and the ORR) the standard material, Pt-Co alloy was tested. Apart from a fundamental investigation in WP 3 [39] the Pt-Co was deposited on nGDL at different initial ratios and used as a cathode catalyst in PEMFC tests and, having shown stable enough, was used in the first PEMFC with NCF on both electrodes. For a broad range of Pt relative amount (from 100% to 40%) on the cathode the power outputs are very similar. The loadings (measured by quantitative XPS) were 2 µg/cm2 on the anode and 48 µg/cm2 on the cathode. Taking the power density of 0.120 W/cm2 into account this fully thin-film PEMFC exhibits specific power (Pt-mass activity) of 2.4 kW/gPt. Endurance test of this PEMFC showed a relatively large variation in power density during the test but also a very small decrease after 3000 cycles. In the view of known Cobalt leaching from Pt-Co during cycling and assuming that it happens on the cathode of the present PEMFC we infer that it is mainly the dispersion of nanoparticles in the nGDL that stands in the way of coalescence-driven ageing (i.e. decrease of power).
Carbon nitride interlayers [26,44]
In the second period the sufficient porosity of the NCF was still a major concern, although nGDL as a substrate provide a major advancement. The next one was achieved by the discovery that during deposition of ceria (or Pt-CeOx) onto carbonaceous substrates by reactive sputtering in Oxygen plasma the substrate is etched by the plasma simultaneously with the material deposition.
As the deposition initially produces disconnected islands, which consequently shadow the substrate material beneath while the rest is etched away, an irregular array of (Pt-)CeOx pillars. This process can be tuned by parameters of the plasma discharge to some extent, but more importantly by the composition of the substrate itself. It was found that the optimal efficiency was achieved using carbon nitrate (CNx) layers for substrate, which can be prepared by reactive carbon sputtering in Nitrogen plasma. Upon deposition of the Pt-CeOx in Oxygen plasma almost all the CNx material is removed and catalyst-coated nanowires remain. Pt concentration had no influence on the as-prepared morphology of the catalyst. This method was stacked on the nGDL as a primary substrate, so that NCF layers with several “levels” of pore sizes resulted and such structure improved the active surface size still further.
Ceria and ALD/CVD
The ceria has been subject to a thorough examination in this WP in order to find the optimal method for its deposition. Apart from the different magnetron sputtering recipes, growth via ALD/CVD was tested.
The PVD produced mode compact layers than ALD/CVD. In context with experience from fuel cell tests of these layers (in this WP and in WP 6) nanoporosity on the scale given by the ALD is “invisible” in a fuel cell and the layers seem compact from that perspective. On the other hand, ALD has been successfully tested for coating the walls of microfluidic channels (for use in WP7) with high aspect ratio.
Tungsten carbide and oxide
For testing non-Platinum materials as catalysts candidates tungsten carbide deposition methods were explored [30]. The WC layers obtained in the first period were rather compact.
Tungsten carbide and oxide films were fabricated on carbon membrane in the second period by direct liquid injection CVD (DLI-CVD). After deposition, the carbon membrane is not degraded and the deposit consists of crystallized needles with a diameter around 5 – 7 nm. An envelope is visible around all the needles as a light grey contour. The contrast suggests that the envelope consists of amorphous carbon. Interplane distances measured on high-resolution images are consistent with tungsten trioxide WO3 crystallized in a monoclinic structure.
After the first period, more experimenting remained to be done. In the second period, however, having experimented with tungsten oxide (WO3) and tungsten oxycarbide (WCO2) in planar systems and deposited on GDL it was concluded that this class of materials is not catalytically active. Only WO3 remained in the scope of the project as an alternative to ceria – a matrix in which catalyst (e.g. Platinum) atoms can be dispersed.
*Characterisation of materials
The Work Package 5 was dedicated to the standard characterization of novel materials’ structure, chemical states (electronic structure) and integral behaviour under operando, typically electrochemical conditions. We focused on materials that were not much simplified for the measurements and analysis. In contrast with model experimental work of WP3 we researched the characteristics of materials close to the form in which they would be used as catalysts in fuel cells.
The first period had it planned to investigate Pt and Pt-Ru nanostructured NCF directly on an “inert” substrate and on ceria, which was of great promise as an “active” substrate already. The Pt-Ru alloy had been put on hold early on; its chief benefit was to be the comparatively higher CO poisoning resistance with respect to Pt. The Pt-CeO2 exhibited similar property, as shown below, so the Pt-Ru alloy lost practical interest and remained a largely unexplored branch.
Comparing TF catalysts and Pt(111) reference by cyclic voltammetry, we observed characteristic changes of the COt band with respect to both the shape and the band position. On Pt(111) the signal is located at 2064 cm-1, whereas on the Pt doped CeO2 samples a red-shift was observed. The corresponding bands are observed at 2056 cm-1 for high Pt concentration and 2023 cm-1 for the low Pt concentration. Whereas the band position is characteristic for metallic Pt, the red shift can be associated with a different morphology of the Pt nanoparticles. Red-shifted features are characteristically observed for CO adsorption at low-coordinated Pt sites. We conclude that for the Pt-CeO2 sample with low Pt concentration, a fraction of the Pt is reduced to Pt0 and stabilized in form of very small nanoparticles. This hypothesis is supported by the observation that the CO signal becomes significantly broader at low Pt concentration. The CO signal on the pure Pt thin film shows a slight blue-shift (2074 cm-1) with respect to the Pt(111) reference. Possible explanations involve the enhanced activity for methanol oxidation at defect sites (formation of CO at more negative electrode potential as also indicated by the shape of the absorption band), co-adsorption of electron withdrawing species and/or and increased CO density on the rougher surface of the magnetron-sputtered film.
The diffuse reflection IR Fourier transform spectroscopy (DRIFTS) was employed to investigate the composition and electronic state of Pt and Pt/CeO2 thin films. Specifically, we studied spectra obtained after dosing CO on pure Pt and Pt/CeO2 thin films prepared on Macor® ceramics by magnetron sputtering. A spectrum of the clean samples after O2 oxidation at 573 K for 15 min has been used as a reference. After CO dosage, the Pt sample (top spectrum) shows a single signal at 2081 cm-1 which can be attributed to CO bound linearly on metallic platinum (Pt0) atoms. The lack of a signal at 1840 cm-1, which is characteristic for CO bridging two Pt atoms, suggests that the dispersion of the Pt film is very high. The samples of Pt co-deposited with CeO2 show no CO-adsorption signals in this range after dosage of CO (spectrum not shown). After annealing the sample in a H2 stream for 1 h at 773 K and subsequent CO dosage, a signal at the same frequency as for the Pt sample appears, indicating the presence of metallic Pt. With increasing reduction temperature, another signal at higher wavenumbers (2161 cm-1) arises, which can be assigned to CO adsorbed linearly on cationic platinum species (Pt2+). The intensity of this new feature increases with progressing reduction temperature. Finally, after reduction at 1017 K no adsorbed CO can be found on the sample anymore.
A considerable number of samples of Pt-CeO2 thin films were made throughout the project by various methods and characterized using electrons microscopies. Highly instructive among the analyses was the HRTEM. TEM micrographs of Pt-CeOx layers prepared using the DLI-CVD process were obtained and from them the films were deemed “nanoporous”. For this characterization the films had been made on flat substrates – this contrasts with efforts made in WPs 4 and 6 that had a great drive for increasing the NCF active surface. However, here the aim was the efficient HRTEM characterization made easier at a greater density of nanocrystallites in each HRTEM lamella. Moreover, the micro-FC arrangement worked out in WP 7 specifically needed knowledge about growth and composition of materials prepared on flat surfaces like Silicon chips.
HRTEM imaging in conjunction with synchrotron radiation XPS provided major evidence in the progress of understanding of the Pt dispersion in ceria. In the second period of the project, the {100} nanopockets on ceria nanoparticles [7] had already been identified as sites stabilizing Pt atoms in the 2+ ionic state. We had obtained HRTEM micrographs of CeO2 nanoparticles (roughly 10 nm in diameter) in Pt-CeOx film and it was clear from this image (and several like it) that the {100} sites are exposed on ceria NP surfaces in no small numbers. Interestingly, though, small amounts of Platinum deposited on the ceria, clearly present there according to XPS, were not detectable in HRTEM. More precisely, for low Pt dosages the EDX measurements in the STEM mode detected uniform presence of Pt signal throughout the layer, but no local aggregations (clusters). The self-same samples characterized using SRPES exhibited presence of Pt almost exclusively in the Pt2+ state, whereas those with greatest total Pt/Ce ratio showed both Pt2+ (ionic) and Pt0 (metallic) species.
The SR XPS measurements with lower incident photon energy have shown that the relative Pt concentration is 5 times higher than the one obtained with conventional XPS. One should point that in this case, the analysis thickness can be decreased below the size of a single particle. As the Pt relative content strongly increases when the probing depth decreases from the entire particle to its topmost atomic layers only, we can conclude that the spatial distribution of Pt through the particle is not homogeneous and, more precisely, that Pt is localized at the surface of the ceria particles instead of in their volume. It was concluded that in low dosages Pt disperses atomically about the ceria nanocrystallite surfaces and are stabilized (trapped) in the {100} nanopockets. With increasing Pt dosage the nanopockets become saturated, after which the Pt atoms start forming metallic nanoclusters. These conclusions eventually led to the realization that the {100} sites are quite abundant on compact ceria thin films as well, namely at step edges, and that low dosages of Pt are dispersed atomically on flat ceria by decorating the step edges [58].
While standing as the only satisfactory hypothesis consistent with the data and supported by DFT calculations, the idea of Pt dispersed atomically on ceria still lacked a conclusive proof in the form of images, where one could “see” the single Pt atom. Indeed identification of a single atom is a major challenge, because the methods that offer the best way to resolve individual atoms (STM, AFM) have very limited capacity to identify those atoms as to their chemical element, not to mention their oxidation state. However, shortly before the project’s end the team got to characterize the Pt-CeO2 samples on one of the world’s highest-performance HRTEM, the NION SuperSTEM at the Trinity College of Dublin.
Having located a ceria nanoparticle oriented with its (100) plane in the image plane the intensity line profile was measured along the atom rows. In these line profiles certain brighter points were identified and using the theory of electron scattering they were ruled, with a great degree of confidence, as being the visualization of Pt single atoms localized in surface {100} nanopocket sites.
We compared formation of Pt structures for different Pt loadings and our observations were consistent with expectations based on the atomic dispersion model: At Pt loadings lower than 10 µg dispersed atoms were observed, occasional Pt dimers (two Pt atoms in neighboring surface nanopockets) and rather rare disordered clusters of a few Pt atoms. As the loading surpassed 10 µg he Pt clusters occurred in much greater abundance, consistently with the saturation of the {100} nanopockets, which accommodate the single Pt2+ ions.
*Thin catalyst layers evaluation and application testing
Fuel cell testing was carried out by the German industrial SME SolviCore GmbH & KG, Hanau, in the second half of the project the testing was largely taken over by the Charles University team. Work Package 6 progressed to a large degree with the aid of the know-how of Solvicore(transformed into Greenerity GmbH by the time of writing this report) and the reports were labelled confidential. This report, therefore summarizes information reduced so as not to infringe that protection.
The NCF layers tested in this WP were prepared mostly by the Charles University team and the Czech SME LET optomechanika Praha Ltd.
Benchmarking
WP6 started with sharing the (public) information on standards of automotive industry and goals of fuel cell performance, setting up protocols to be used in fuel cell testing of catalysts generated by the project, and preparing and testing a benchmark membrane assembly (MEA) prepared so as to be comparable with those using the nanocatalyst films (NCF). As the second step, MEAs with NCF on the anode side were assembled and tested. The conclusion of these tests was that the critical weakness of the NCF is their compactness, i.e. a vanishing porosity. It was the porosity, therefore, that needed to be addressed primarily.
Testing strategies to enlarge surface area
Next, the WP proceeded by testing MEAs made with NCF deposited on variously prepared membranes. The MEAs were mostly assembled with the NCF catalyst on the anode and a well-reproducible state-of-the-art (i.e. “standard” for the purposes of comparison within the project) catalyst on the cathode side. One MEA was assembled with NCF on both sides.
It was concluded that ion-etching of membranes reduces its internal resistance, but the etching did not have a marked effect on the compactness (low surface area) of the NCF. The universally low surface areas of NCF (the EPSA parameter) indicated that nanoporosity identified elsewhere in the project (see Work Package 5) was not of the scale that can be exploited in the electrochemical environment. Results obtained for the NCF-NCF MEA, when normalized to Cathode loading (intrinsically much smaller than for the “standard”), showed some a certain promise, provided the NCF could be prepared with greater EPSA.
Tests of variously modified NCF have shown that the electrochemical surface area (ECA in m² g-1), of sputtered platinum electrodes onto edged membranes could not be increased by the desired factor of 10 (i.e. from 6 m² g-1 to 60 m² g-1). Indeed, independently from the NCF thickness, 10 or 20nm, the surface area led to same EPSA of around 2 cm²Pt cmMEA-2. This indicates that the electrodes are two-dimensional (compact) and not three-dimensional (sufficiently porous). A certain porosity was observed in SEM and AFM images, but it seems that the pore sizes are too small for protons and gas molecules to penetrate the layer sufficiently. The porosity of the NCF still remained to be increased.
The following approaches were proposed to help increase the porosity and, therefore, the active platinum surface:
1. Nano GDL – the possibility of direct sputtering of platinum on GDL (nanoGDL) was to be further explored and exploited
2. CCM based on a coated carbon/ionomer layer with sputtered platinum – the carbon+ionomer+sputtered_Platinum layer could be applied either directly to the membrane, or to inert substrate, from which the layer would be transferred onto the membrane
3. Carbon powder sputtered with platinum – either using high surface area carbon particles or carbon nanotubes
4. Sputtering polymer – attempting to sputter polymer onto the membrane with the expectation that increased surface would result, Pt would be sputtered onto the membrane thus prepared
The strategy primarily used was sputtering Pt onto a carbon ionomer layer (c-i-l) in several thicknesses and used as anodes. In MEA they were again coupled with standard cathodes. At the nominal Pt layer thickness of 30nm the Pt/c-i-l NCF anode had similar performance to the benchmark; this was the first observation of so close a result, within a few percent of power at voltages up to 1.6 V. This best anode had ECA 7 m2 gPt-1 and EPSA 4 cm2Pt cm-2geo , which was not an improvement.
Tests performed on anodes with a ceria interlayer (Pt/CeOx/c-i-l) gave worse results, the main effect of CeOx presence being the increased resistance of the MEA. On the other hand, testing of the layers on CO poisoning was done using the CO stripping method and it showed no CO adsorption for low Pt loadings. At large Pt loadings some CO capture was observed, which was attributed to presence of Pt0 clusters as opposed to atomically dispersed Pt.
Pt/c-i-l NCF were tested on the cathode side as well. The loadings were a factor 4-10 lower than that of standard cathodes and the NCF cathode MEAs exhibited systematically lower cell voltages, bigger Tafel slopes and a big variation indicated that the NCF suffered from mass transport issues at currents even as low as 0.2 A/cm2. In other words, the insufficient NCF porosity persisted as the chief obstacle. Pt/CeOx NCF deposited by glancing angle deposition (GLAD) showed some improvement in performance (and porosity).
Toward fully thin-film PEMFCs
In the third period the MEA assembly and testing for the project has almost completely migrated to the Coordinator CUP), thanks to the built-up FC testing facility and the adoption of knowledge and industrial standards for the testing. Indeed, CUP progressed with its own instrument development so that alongside the NCF technology the CUP team began to market their own fuel cell test stations in a commercial venture (LEANCAT s.r.o. a spin-out from CUP in partnership with industry).
The chipCAT Pt-doped ceria catalyst research was concluded by preparing thin film anodes with almost negligible content of platinum, while achieving industrial standards in performance and durability. The Pt-ceria catalysts, which had been introduced as an innovative material due to very high specific power, were continuously improved during the project and represent one of its best achievements.
During the first three years of the project we repeatedly saw that further improvement of the catalyst performance concerning its power density can be obtained by tuning the catalyst film substrate porosity, which would lead to the increase of the catalyst active surface [23,25]. Additionally, we focused our effort onto tuning the hydrophilicity/hydrophobicity ratio by preparing our own nano GDL, to which we switched from the commercial one used previously used (Sigracet GDL).
The final catalyst preparation steps were:
- GDL coating by carbon NPs dissolved in ionomer and PTFE solutions;
- reactive deposition of the CNx GDL interlayer;
- Pt-ceria TF deposition.
CNx interlayer is increasing the efficiency of oxygen plasma etching. Efficiency of the CNx film etching was well observable in case of the same process applied on flat CNx layer deposited on a Si wafer.
Fuel cell testing using single cell of 4.5 cm2 of MEA active surface was performed for different Pt loadings from 1 µg of Pt per 1 cm2 of the catalyst geometrical surface. The test was performed by using commercial cathode. The results show that the maximum power density increases with Pt loading and approaches asymptotically the maximum value of 1.4 W/cm2 for 10 µg Pt/cm2.
By comparing with similar results presented in the progress report for 2nd period where we have shown results obtained without using a CNx interlayer we can see the same asymptotic behaviour, however in previous case the maximum power density was 0.6 W/cm2.
Thanks to a new fuel cell testing laboratory at CUP, we performed accelerated durability tests of 10µg and 4µg Pt samples using two different testing protocols, applied simultaneously to three identical samples. Both tests have shown high stability of Pt-ceria anodes with performance decrease smaller than 10%.
Beside development of an efficient Pt-ceria TF anode an effort has been made in development of a new TF cathode in order to develop full thin film MEA and/or efficient micro-fuel cells with optimised electrodes. We proposed a new procedure based on preparation of a nanoporous substrate by reactive deposition of carbon nitride and cerium oxide on “homemade” ionomer+PTFE+C NPs coated GDL. Then the substrate was coated by metallic Pt thin film and a MEA made with it was tested. The stable MEA revealed 0.45 W/cm2 giving specific power of 9 kW/g Pt.
These results have prompted tests of NCF PEMFC with alternative materials, which showed promise in theory, model and applied research of WPs 2 through 5. While these tests are not concluded yet, there have been promising data collected, for example, on atomically dispersed Pt in WO3 matrix.
*Chip design, microfabrication
Most of the work related to the on-chip micro-fuel cells was carried out by the Italian SME ThinderNIL s.r.l. In accordance with the Work Package 7 plan, in the first Period work started on the design of the microfluidic chips to form the base of the micro-fuel cells. This was done by means of finite element calculations. Although delayed to the second period, the modelling was carried out and took into account practical problems and workarounds discovered in parallel.
Basic concept
In an on-chip microfluidic fuel cell the anode and cathode catalysts would be placed in close-lying microchannels etched into one face of a Silicon wafer (chip). This concept assumed that the catalysts would cover the channel walls and be contacted by thin film electrodes so that the barrier between the two channels is reliably insulating. The proton-exchange membrane (PEM) would be hot-pressed onto the top of the chip and during function protons generated in the anode (fuel) channel would migrate laterally towards the cathode, where they would recombine with oxygen into water. Water is to be carried away by the cathode channel as exhaust.
Lithography patterns
The realization of the desired on-chip devices required choosing channel patterns to test, developing the recipes for channel lithography and for the depositions of electrodes and catalysts and producing the necessary set of lithography masks. There were several patterns made in the masks: A simple one with only one linear section of close-lying channels and a complex one with a large number (ca. 20) of interdigitated channels for fuel and oxygen. The simple one was used in the first proof of principle devices, whereas the latter was later deemed non-functional and replaced by one, where the pair of channels meanders many times. This latter type’s advantage is the lack of dead ends.
Practical tests of hot-press fixing of the PEM to the Silicon chip showed several difficulties. A major one to overcome was the fact that the PEM (NafionTM or similar material) dilates considerably, when soaking with water molecules in preparation for function. Hydration is essential for the membrane function, which is why the working gasses are wetted en route to the PEMFC by passing through bubblers. However, the expanding soaked membrane tends to squeeze into the channels and block them. To overcome this problem an array of micro-pillars was left inside the channels. They were dense enough to prevent the channel blockage by membrane but sparse enough to not impede the flow of the fuel and oxygen through the channels.
A “pioneer” microchannel cell
As the design, lithography work and tests went on, the planar fuel cell concept was tested with a “macro” model. There the channels had larger dimensions, were etched in glass blocks and did not require micrometre precision for machining, etching or for thin film depositions. Designed and manufactured between the teams of Charles University and of LET optomechanika Praha.
The parallel channels in the “pioneer” planar cell were separated by 100 µm, 1 mm wide and 100 µm deep. Length of parallel part of channels was 10 mm. Both anode and cathode channels were coated with 20 nm thick Pt catalyst film, the electrical contact of electrodes were ensured by the thin film of gold evaporated on the cell glass plate. The Nafion NR-212 membrane was used as electrolyte. The planar fuel cell was tested using hydrogen and oxygen gas at room temperature and atmospheric pressure and it did produce a standard obtained polarisation curve.
Finite-element simulations
The device design simulations already included the pillars inside the microchannels. For the simulation to was decided to assume the micro-FC device working as a direct methanol fuel cell (DMFC). In other words, the finite elements computation solved Navier-Stokes equation for incompressible fluids in laminar flow, where the appropriateness of the latter assumption (laminar flow) was tested by enumeration of Reynolds number. The highest estimate of Re was ~0.08 and, therefore, the solutions of NS equation can lead to realistic solutions.
Laminar flow of methanol was simulated using COMSOL software, first in an empty channel with dimensions being 200 µm wide, 30 µm deep and 400 µm long. The same computation was then performed with the channel filled with square pillars of 10 μm arranged in a square array of 30 μm period, separated from the walls of 20 μm. The pressure drop over 400 um was 3.5 Pa in the former case and 13.2 Pa in the latter. In this regime the pressure drop is proportional to the viscosity and linearly dependent on the fluid speed. Therefore, a single simulation done for a given value of the viscosity and given fluid speed is used to generate the entire set of data for all viscosities and fluid speeds of interest.
Assuming 50% FC efficiency and targeting 10 mW/cm2 output, the pressure drop along a channel was below 2 bar/m at for both geometries (without and with pillars) at all methanol concentrations from 2M up. The longest meander considered in the project for an on-chip micro-FC is 80 cm long. It is indicated that the available flow will be able to supply enough fuel for the target output power.
Chip interface platform
After making the microfluidic devices, depositing the electrodes and catalysts and pressing the PEM on, it was necessary to connect it to a fuel-cell testing infrastructure – a platform interfacing the gas leads and electronics with the microfluidic device. The platform design was made already in the first period, although it was utilized only in the second.
A guiding principle adopted in the design of the chip holder was to reduce the sources of possible contaminations and, thus, poisoning of the fuel cell. For this reason all materials were selected for their high chemical and thermal resistance. The lower base plate and the fluid pipes are in AISI 316 stainless steel and are pressure-joined by accurate matching the holes diameters and without any glue. These couplings have been tested to be leak-free up to 7.5 bar of N2 pressure.
Along with the platform, a punch press was also designed and manufactured for reproducible perforation of the membranes, where fuel and oxygen inlets/outlets had to connect.
Sealing
µ-FC were sealed with Nafion PEM, by hot embossing with multiple Pulsed-NIL pulses. The process consisted of pressing the Nafion membrane with a flat stamp with integrated heater, and applying a series of strong heating pulse of 50 µs of 10 J/cm2/pulse to the stamp.
Leak tests
In the first test, a microscope glass slide has been cut to the same dimensions of a chip (22 x 22 mm²) and placed inside the holder. Then compressed nitrogen with a pressure of 7.5 bar has been supplied to all four pipes and then isolated from the supply line by a manual valve. The pressure of each branch of the system was monitored for 3 days without any detectable drop of pressure.
A functional test was performed by assembling a chip in the holder and applying nitrogen to a pressure of up to 4 bar to only one of the four pipes while the open ends of the other three pipes were let out into beakers with water in order to detect even very small flows of gas. A considerable flow of gas (1 bubble of about 4 mm³ per second) was detected at the exit of the first circuit, while no bubbles were detected at either end of the other circuit. Then the exit of the first circuit was closed with a manometer and we waited until the same pressure of the inlet was reached. Again, no bubbles were detected at the ends of the other circuit. The same procedure was applied to the second circuit with the same results. Thus we considered the test as passed.
Dielectric test
The second test aimed at checking the electrical insulation of the four cables with respect to each other and to the chip holder body. Being coaxial cables, we tested separately the inner conductor and the external shield. A tension up to 200 V dc has been applied across to each combination of single conductor couples with no detectable current flow.
Pulsed nanoimprint lithography fabrication process
The technology of the Pulsed-NIL has been used in two specific steps of the µ-FC fabrication process, namely in the patterning of the microfluidic channels in parallel for all the 12 chips of a 4” wafer, and for the sealing, chip by chip, of the proton exchange membrane (Nafion) onto the patterned substrate of the chips. The patterning of the microfluidic circuits of the chip is obtained by Pulsed-NIL process with the Pulsed-NIL process.
On-chip fuel cell measurements
The platform was afterwards used for measurements of several chips of the simple channel design and one resulted in a stable output with power maximum around 8µW. However low the power, this was noted as a successful step on the way to the on-chip micro-FC.
Pilot production line proposal
Table: Throughput of the different process steps for the fabrication of the µ-FCs. The table shows the throughput at the present stage, and provides the estimation for the production in a Pilot Line.
Process step Technology Present Throughput Projected Throughput in a Pilot Line
Microfluidic chip Patterning Pulsed-NIL 14 w/h 120 w/h
Pattern transfer into silicon ICP-RIE (Bosh process) 3 w/h 6-8 w/h
Deposition of Pt Catalyst Sputtering 6 w/h 60 w/h
Patterning of the metal layer PhLith or Pulsed-NIL 15 w/h 120 w/h
Sealing Pulsed-NIL 20 chips/h 120 w/h
It is clear that the main bottleneck of the fabrication process, both at the present stage of the development and for an estimated pilot production is the process of deep etching of the microfluidic structures into silicon. Therefore, in order to have all steps at the same throughput level, in production more ICP-RIE system should be running in parallel to advance the batch at comparable speed with the other steps. However, this has a disadvantage, represented by the high cost of modern ICP-RIE systems, and the cost of a process involving vacuum systems.
An alternative to the high investment and running costs and the low throughput of ICP-RIE systems, would be the fabrication of the microfluidic base of the µ-FCs chips in high-melting and high useful operating temperature thermoplastic materials, such as Polyether ether ketone (PEEK), Polyethersulphone (PES), Polyphenylene sulfide (PPS). Making µ-FCs by processes like injection molding, hot embossing, or Pulsed-NIL would have the advantage of higher productivity and lower costs of equipment, of the row materials, and running costs.
* Conclusion
The chipCAT project produced 71 articles published in peer-reviewed journals, 30 of them being collaborative articles between the Consortium’s academic research groups, and with the knowledge embodied there the Consortium reached both applied goals of the project: a) Thin-film catalysts for both electrodes with very small amounts of Platinum were developed into nanostructured layers that in proton exchange membrane fuel cells near the performance goals of the automotive industry (0.5 – 1 W/cm2) at a greatly reduced Platinum loading. b) Proof-of-concept prototypes of on-chip micro-fuel cells were developed by realization of several of their key enabling ideas, they were assembled and successfully tested. The teams involved consider the project completely successful, the primary cause of which was the balance between fundamental and applied research within.
*Publications
[1] Martin Dubau, Jaroslava Lavková, Ivan Khalakhan, Stanislav Haviar, Valerie Potin, Vladimír Matolín, and Iva Matolínová, Preparation of Magnetron Sputtered Thin Cerium Oxide Films with a Large Surface on Silicon Substrates Using Carbonaceous Interlayers, ACS Appl. Mater. Interfaces, 6 (2), 2013, 1213-1218, 10.1021/am4049546
[2] Vitalii Stetsovych, Federico Pagliuca, Filip Dvořák, Tomáš Duchoň, Mykhailo Vorokhta, Marie Aulická, Jan Lachnitt, Stefan Schernich, Iva Matolínová, Kateřina Veltruská, Tomáš Skála, Daniel Mazur, Josef Mysliveček, Jörg Libuda, and Vladimír Matolín, Epitaxial Cubic Ce2O3 Films via Ce–CeO2 Interfacial Reaction, J. Phys. Chem. Letters, 4 (6), 2013, 866-871, 10.1021/jz400187j
[3] Ghosh P., M. Farnesi Camellone, and S. Fabris, Fluxionality of Au clusters at ceria surfaces during CO oxidation: relationships among reactivity, size, cohesion, and surface defects from DFT simulations, J. Phys. Chem. Letters, 4 (14), 2013, 2256-2263, 10.1021/jz4009079
[4] Karolina Kwapien, Simone Piccinin, and Stefano Fabris, Energetics of Water Oxidation Catalyzed by Cobalt Oxide Nanoparticles: Assessing the Accuracy of DFT and DFT+U Approaches Against Coupled Cluster Methods, J. Phys. Chem. Letters, 4 (24), 2013, 4223-4230, 10.1021/jz402263d
[5] Paul Jennings, Hristiyan A. Aleksandrov, Konstantin M. Neyman, and Roy L. Johnston, A DFT study of oxygen dissociation on platinum based nanoparticles, Nanoscale, 6, 2013, 1153-1165, 10.1039/C3NR04750D
[6] Sergey M. Kozlov, Hristiyan A. Aleksandrov, Jacek Goniakowski, and Konstantin M. Neyman, Effect of MgO(100) support on structure and properties of Pd and Pt nanoparticles with 49-155 atoms, J. Chem. Phys, 139, 2013, 084701, 10.1063/1.4817948
[7] Bruix A., Y. Lykhach, I. Matolínová, A. Neitzel, T. Skála, N. Tsud, M. Vorokhta, V. Stetsovych, K. Ševčíková, J. Mysliveček, K. C. Prince, S. Bruyère, V. Potin, F. Illas, V. Matolín, J. Libuda, and K. M. Neyman, Maximum Noble Metal Efficiency in Catalytic Materials: Atomically Dispersed Surface Platinum, Angewandte Chemie Intl. Ed., 53 (39), 2014, 10525-10530, 10.1002/anie.201402342
[8] Hristiyan A. Aleksandrov, Sergey M. Kozlov, Swetlana Schauermann, Georgi N. Vayssilov, Konstantin M. Neyman, How Absorbed Hydrogen Affects the Catalytic Activity of Transition Metals, Angewandte Chemie Intl. Ed., 53 (49), 2014, 13371-13375, 10.1002/anie.201405738
[9] Yaroslava Lykhach, Armin Neitzel, Klára Ševčíková, Viktor Johánek, Nataliya Tsud, Tomáš Skála, Kevin C. Prince, Vladimír Matolín, and Jörg Libuda, The Mechanism of Hydrocarbon Oxygenate Reforming: C—C Bond Scission, Carbon Formation, and Noble-Metal-Free Oxide Catalysts, Chem. Sus. Chem., 7 (1), 2014, 77-81, 10.1002/cssc.201301000
[10] Sevcikova, K; Nehasil, V; Zahoranova, T; Vorokhta, M; Tsud, N; Yoshikawa, H; Kobata, M; Kobayashi, K ; Matolin, V, The effect of the substrate on thermal stability of CeOx and Rh–Ce–O thin films, Surf. Interface Anal., 46 (10-11), 2014, 980-983, 10.1002/sia.5503
[11] Armin Neitzel , Yaroslava Lykhach , Viktor Johánek , Nataliya Tsud , Tomáš Skála , Kevin C. Prince , Vladimír Matolín , and Jörg Libuda, Role of Oxygen in Acetic Acid Decomposition on Pt(111), J. Phys. Chem. C, 118 (26), 2014, 14316-14325, 10.1021/jp502017t
[12] Sergey M. Kozlov, Hristiyan A. Aleksandrov, Konstantin M. Neyman, Adsorbed and Subsurface Absorbed Hydrogen Atoms on Bare and MgO(100)-Supported Pd and Pt Nanoparticles, J. Phys. Chem. C, 118 (28), 2014, 15242-15250, 10.1021/jp502575a
[13] Farnesi Camellone M., D Marx, Nature and role of activated molecular oxygen species at the gold/titania interface in the selective oxidation of alcohols, J. Phys. Chem. C, 118, 2014, 20989-21000, 10.1021/jp5060233
[14] Fabio R. Negreiros, Stefano Fabris, Role of Cluster Morphology in the Dynamics and Reactivity of Subnanometer Pt Clusters Supported on Ceria Surfaces, J. Phys. Chem. C, 118 (36), 2014, 21014-21020, 10.1021/jp506404z
[15] Ma C., S. Piccinin, S. Fabris, Interface structure and reactivity of water-oxidation Ru-polyoxometalate catalysts on functionalized graphene electrodes, Phys. Chem. Chem. Phys., 16, 2014, 5333-5341, 10.1039/C3CP54943G
[16] Sergey M. Kozlov , Konstantin M. Neyman, O vacancies on steps on the CeO2(111) surface, Physical Chemistry Chemical Physics, Vol. 16/Issue 17, 2014, 7823-7829, 10.1039/c4cp00136b
[17] Neitzel A., Y. Lykhach, T. Skála, N. Tsud, V. Johánek, M. Vorokhta, K. C. Prince, V. Matolín, J. Libuda, Hydrogen activation on Pt-Sn nanoalloys supported on mixed Sn-Ce oxide films, Physical Chemistry Chemical Physics, 16, 2014, 13209-13219, 10.1039/C4CP01632G
[18] Paul C. Jennings, Hristiyan A. Aleksandrov, Konstantin M. Neyman, Roy L. Johnston, DFT studies of oxygen dissociation on the 116-atom platinum truncated octahedron particle, Physical Chemistry Chemical Physics, 16 (48), 2014, 26539-26545, 10.1039/c4cp02147a
[19] Armin Neitzel , Yaroslava Lykhach , Tomáš Skála , Nataliya Tsud , Mykhailo Vorokhta , Daniel Mazur , Kevin C. Prince , Vladimír Matolín , Jörg Libuda, Surface sites on Pt–CeO2 mixed oxide catalysts probed by CO adsorption: a synchrotron radiation photoelectron spectroscopy study, Physical Chemistry Chemical Physics, 16 (45), 2014, 24747-24754, 10.1039/c4cp03346a
[20] Mahasin Alam Sk, Sergey M. Kozlov, Kok Hwa Lim, Annapaola Migani, Konstantin M. Neyman, Oxygen vacancies in self-assemblies of ceria nanoparticles, J. of Materials Chemistry A, 2 (43), 2014, 18329-18338, 10.1039/C4TA02200A
[21] Stanislav Haviar, Martin Dubau, Jaroslava Lavková, Ivan Khalakhan, Valérie Potin, Vladimír Matolín, Iva Matolínová, Investigation of Growth Mechanism of Thin Sputtered Cerium Oxide Films on Carbon Substrates, Science of Advanced Materials, 6 (6), 2014, 1278-1285, 10.1166/sam.2014.1905
[22] Firas Faisal, Arafat Toghan, Ivan Khalakhan, Mykhailo Vorokhta, Vladimír Matolín, Jörg Libuda, Characterization of thin CeO2 films electrochemically deposited on HOPG, Applied Surface Science, 350, 2015, 142-148, 10.1016/j.apsusc.2015.01.198
[23] Fiala R., M. Václavů, A. Rednyk, I. Khalakhan, J. Lavková, V. Potin, I. Matolínová, V. Matolín, Pt–CeOx thin film catalysts for PEMFC, Catalysis Today, 240 Part B, 2015, 236-241, 10.1016/j.cattod.2014.03.069
[24] Beran, J; Matolin, V; Masek, K, RHEED structural study of the novel tin-cerium oxide catalyst, Ceramics Intl, 41 (3) Part B, 2015, 4946-4652, 10.1016/j.ceramint.2014.12.057
[25] Fiala R., M. Václavů, M. Vorokhta, I. Khalakhan, J. Lavková, V. Potin, I. Matolínová, V. Matolín, Proton exchange membrane fuel cell made of magnetron sputtered Pt–CeOx and Pt–Co thin film catalysts, Journal of Power Sources, 273, 2015, 105-109, 10.1016/j.jpowsour.2014.08.093
[26] Potin V., J. Lavkova, S. Bourgeois, M. Dubau, I. Matolinova and V. Matolin, Structural and chemical characterization of cerium oxide thin layers grown on silicon substrate, Materials Today: Proceedings, 2 (1), 2015, 101-107, 10.1016/j.matpr.2015.04.014
[27] Avril L., N. Zanfoni, P. Simon, L. Imhoff, S. Bourgeois, B. Domenichini, MOCVD growth of porous cerium oxide thin films on silicon substrate, Surface and Coatings Technology, 280, 2015, 148-153, 10.1016/j.surfcoat.2015.07.055
[28] Zanfoni N., L. Avril, L. Imhoff, B. Domenichini, S. Bourgeois, Direct liquid injection chemical vapor deposition of platinum doped cerium oxide thin films, Thin Solid Films, 589, 2015, 246-251, 10.1016/j.tsf.2015.05.037
[29] Avril L., S. Bourgeois, P. Simon, B. Domenichini, N. Zanfoni, F. Herbst, L. Imhoff, Nanostructured Pt–TiO2 composite thin films obtained by direct liquid injection metal organic chemical vapor deposition: Control of chemical state by X-ray photoelectron spectroscopy, Thin Solid Films, 591B, 2015, 237-244, 10.1016/j.tsf.2015.06.007
[30] Nazon J., M. Herbst, M.C. Marco de Lucas, S. Bourgeois, B. Domenichini, WC-based thin films obtained by reactive radio-frequency magnetron sputtering using W target and methane gas, Thin Solid Films, 591B, 2015, 119-125, 10.1016/j.tsf.2015.08.035
[31] Armin Neitzel, Yaroslava Lykhach, Viktor Johánek, Nataliya Tsud, Tomáš Skála, Kevin C. Prince, Vladimír Matolín, Jörg Libuda, Decomposition of Acetic Acid on Model Pt/CeO2 Catalysts: The Effect of Surface Crowding, J. Phys. Chem. C, 119 (24), 2015, 13721-13734, 10.1021/acs.jpcc.5b03079
[32] Negreiros F.R. M. Farnesi Camellone, S. Fabris, Effects of thermal fluctuations on the hydroxylation and reduction of ceria surfaces by molecular H2, J. Phys. Chem. C, 119 (37), 2015, 21567-21573, 10.1021/acs.jpcc.5b07030
[33] Lucie Szabová, Yoshitaka Tateyama, Vladimír Matolín, Stefano Fabris, Water Adsorption and Dissociation at Metal-Supported Ceria Thin Films: Thickness and Interface-Proximity Effects Studied with DFT+U Calculations, J. Phys. Chem. C, 119 (5), 2015, 2537-2544, 10.1021/jp5109152
[34] Paul C. Jennings, Hristiyan A. Aleksandrov, Konstantin M. Neyman, Roy L. Johnston, O2 Dissociation on M@Pt Core−Shell Particles for 3d, 4d, and 5d Transition Metals, J. Phys. Chem. C, 119 (20), 2015, 11031-11041, 10.1021/jp511598e
[35] Sergey M. Kozlov, Hristiyan A. Aleksandrov, Konstantin M. Neyman, Energetic Stability of Absorbed H in Pd and Pt Nanoparticles in a More Realistic Environment, J. Phys. Chem. C, 119 (9), 2015, 5180-5186, 10.1021/jp513022m
[36] Jaroslava Lavková, Ivan Khalakhan, Mykhailo Chundak, Mykhailo Vorokhta, Valerie Potin, Vladimír Matolín, Iva Matolínová, Growth and composition of nanostructured and nanoporous cerium oxide thin films on a graphite foil, Nanoscale, 7, 2015, 4038-4047, 10.1039/C4NR06550F
[37] Sergey M. Kozlov, Ilker Demiroglu, Konstantin M. Neyman, Stefan T. Bromley, Reduced ceria nanofilms from structure prediction, Nanoscale, 7 (10), 2015, 4361-4366, 10.1039/C4NR07458K
[38] Sergey M. Kozlov, Gábor Kovács, Riccardo Ferrando, Konstantin M. Neyman, How to determine accurate chemical ordering in several nanometer large bimetallic crystallites from electronic structure calculations, Chemical Science, 6, 2015, 3868-3880, 10.1039/C4SC03321C
[39] Gábor Kovács, Sergey M. Kozlov, Iva Matolínová, Mykhailo Vorokhta, Vladimír Matolín, Konstantin M. Neyman, Revealing chemical ordering in Pt–Co nanoparticles using electronic structure calculations and X-ray photoelectron spectroscopy, Physical Chemistry Chemical Physics, 17, 2015, 28298-28310, 10.1039/C5CP01070E
[40] Hristiyan A. Aleksandrov, Konstantin M. Neyman, Georgi N. Vayssilov, The structure and stability of reduced and oxidized mononuclear platinum species on nanostructured ceria from density functional modelling, Physical Chemistry Chemical Physics, 17 (22), 2015, 14551-14560, 10.1039/C5CP01685A
[41] Nguyen M.T. M. Farnesi Camellone, R. Gebauer, On the electronic, structural, and thermodynamic properties of Au supported on α-Fe2O3 surfaces and their interaction with CO, J. Chem. Phys., 143, 2015, 034704, 10.1063/1.4926835
[42] Piccinin S., S. Fabris, Water oxidation by Ru-polyoxometalate catalysts: Overpotential dependency on the number and charge of the metal centers, Inorganics, 3, 2015, 374-387, 10.3390/inorganics3030374
[43] Myslivecek, J; Matolin, V; Matolinova, I, Heteroepitaxy of Cerium Oxide Thin Films on Cu(111), Materials, 8 (9), 2015, 6346-6359, 10.3390/ma8095307
[44] Khalakhan, I; Fiala, R; Lavková, J; Kúš, P; Ostroverkh, A; Václavů, M; Vorokhta, M; Matolínová, I; Matolín, V, Candle Soot as Efficient Support for Proton Exchange Membrane Fuel Cell Catalyst, Fuel Cells, 16 (5), 2016, 652-655, 10.1002/fuce.201600016
[45] Neitzel A., G. Kovács, Y. Lykhach, S.M. Kozlov, N. Tsud, T. Skála, M. Vorokhta, V. Matolín, K.M. Neyman, J. Libuda, Atomic ordering and Sn segregation in Pt-Sn nanoalloys supported on CeO2 thin films, Top. Catal., online only, 2016, , 10.1007/s11244-016-0709-5
[46] Monai M.; Montini, T; Melchionna, M; Duchoň, T; Kúš, P; Tsud, N; Prince, KC; Matolin, V; Gorte, RJ; Fornasiero, P, Phosphorus poisoning during wet oxidation of methane over Pd@CeO2/graphite model catalysts, Appl. Catal. B-Environ., 197, 2016, 271-279, 10.1016/j.apcatb.2015.10.001
[47] Fiala R., Figueroba, A; Bruix, A; Vaclavu, M; Rednyk, A; Khalakhan, I; Vorokhta, M; Lavkova, J; Illas, F; Potin, V; Matolinova, I; Neyman, KM; Matolin, V, High efficiency of Pt2+-CeO2 novel thin film catalyst as anode for proton exchange membrane fuel cells, Appl. Catal. B-Environ., 197, 2016, 262-270, 10.1016/j.apcatb.2016.02.036
[48] Vorokhta M., I. Khalakhan, M. Václavu, G. Kovács, S.M. Kozlov, P. Kúš, T. Skála, N. Tsud, J. Lavková, V. Potin, I. Matolínová, K.M. Neyman, V. Matolín, Surface composition of magnetron sputtered Pt-Co thin film catalyst for proton exchange membrane fuel cells, Appl. Surf. Sci., 365, 2016, 245-251, 10.1016/j.apsusc.2016.01.004
[49] Neitzel A., V. Johánek, Y. Lykhach, T. Skála, N. Tsud, M. Vorokhta, V. Matolín, J. Libuda, Reduction of Pt2+ species in model Pt—CeO2 fuel cell catalysts upon reaction with methanol, Appl. Surf. Sci., 387, 2016, 674-681, 10.1016/j.apsusc.2016.06.156
[50] Wolfbeisser A., G. Kovács, S.M. Kozlov, K. Föttinger, J. Bernardi, B. Klötzer, K.M. Neyman, G. Rupprechter. , Surface composition changes of CuNi-ZrO2 catalysts during methane decomposition: An operando NAP-XPS and density functional study, Catal. Today, online only, 2016, , 10.1016/j.cattod.2016.04.022
[51] Khalakhan I., M. Vorokhta, M. Václavů, B. Šmíd, J. Lavková, I. Matolínová, R. Fiala, N. Tsud, T. Skála, V. Matolín, In-situ electrochemical atomic force microscopy study of aging of magnetron sputtered Pt-Co nanoalloy thin films during accelerated degradation test, Electrochim. Acta, 211, 2016, 52-58, 10.1016/j.electacta.2016.06.035
[52] Rednyk A; Johanek, V; Khalakhan, I; Dubau, M; Vorokhta, M; Matolin, V, Methanol oxidation on sputter-coated platinum oxide catalysts, Int. J. Hydrog. Energy, 41 (1), 2016, 265-275, 10.1016/j.ijhydene.2015.09.147
[53] Simon P., N. Zanfoni, L. Avril, J. Lavkova, I. Matolinova, L. Imhoff, V. Potin, B. Domenichini, S. Bourgeois, Observation of surface reduction in porous ceria thin film grown on graphite foil substrate, Materials Today: Proceedings, 3 (8), 2016, 2772-2779, 10.1016/j.matpr.2016.06.026
[54] Ševčíková K., L. Szabová, M. Kettner, P. Homola, N. Tsud, S. Fabris, V. Matolín, V. Nehasil, Experimental and theoretical study on the electronic interaction between Rh adatoms and CeOx substrate in dependence on a degree of cerium oxide reduction, J. Phys. Chem C, 120, 2016, 5468-5476, 10.1021/acs.jpcc.5b11431
[55] Neitzel A., A. Figueroba, Y. Lykhach, T. Skála, M. Vorokhta, N. Tsud, S. Mehl, K. Ševčíková, K.C. Prince, K.M. Neyman, V. Matolín, J. Libuda, Atomically Dispersed Pd, Ni and Pt Species in Ceria-Based Catalysts: Principal Differences in Stability and Reactivity, J. Phys. Chem. C, 120, 2016, 9852-9862, 10.1021/acs.jpcc.6b02264
[56] Brummel O., F. Waidhas, F. Faisal, R. Fiala, M. Vorokhta, I. Khalakhan, M. Dubau, A. Figueroba, G. Kovács, H.A. Aleksandrov, G.N. Vayssilov, S.M. Kozlov, K.M. Neyman, V. Matolín, J. Libuda, Stabilization of Small Platinum Nanoparticles on Pt–CeO2 Thin Film Electrocatalysts During Methanol Oxidation, J. Phys. Chem. C, 120, 2016, 19723-19736, 10.1021/acs.jpcc.6b05962
[57] Ostroverkh, A; Johánek, V; Kúš, P; Šedivá, R; Matolín, V, Efficient Ceria–Platinum Inverse Catalyst for Partial Oxidation of Methanol, Langmuir, 32 (25), 2016, 6297–6309, 10.1021/acs.langmuir.6b01316
[58] Dvorak, F; Camellone, MF; Tovt, A; Tran, N-D; Negreiros, FR; Vorokhta, M; Skala, T; Matolinova, I; Myslivecek, J; Matolin, V; Fabris, S, Creating single-atom Pt-ceria catalysts by surface step decoration, Nat. Commun., 7, 2016, 10801 (8 pages), 10.1038/ncomms10801
[59] Yaroslava Lykhach, Sergey M. Kozlov, Tomáš Skála, Andriy Tovt, Vitaliy Stetsovych, Nataliya Tsud, Filip Dvořák, Viktor Johánek, Armin Neitzel, Josef Mysliveček, Stefano Fabris, Vladimír Matolín, Konstantin M. Neyman, Jörg Libuda, Counting electrons on supported nanoparticles, Nature Materials, 15, 2016, 284-288, 10.1038/NMAT4500
[60] Lykhach Y., A. Figueroba, M.F. Camellone, A. Neitzel, T. Skála, F.R. Negreiros, M. Vorokhta, N. Tsud, K.C. Prince, S. Fabris, K. Neyman, V. Matolín, J. Libuda, Reactivity of Atomically Dispersed Pt2+ Species towards H2: Model Pt−CeO2 Fuel Cell Catalyst, Phys. Chem. Chem. Phys., 18, 2016, 7672-7679, 10.1039/C6CP00627B
[61] Figueroba A., G. Kovács, A. Bruix, K.M. Neyman, Towards stable single-atom catalysts: strong binding of atomically dispersed transition metals on the surface of nanostructured ceria, Catal. Sci. Technol., 6, 2016, 6806-6813, 10.1039/C6CY00294C
[62] Neitzel A., G. Kovács, Y. Lykhach, N. Tsud, S.M. Kozlov, T. Skála, M. Vorokhta, V. Matolín, K.M. Neyman, J. Libuda, Steering the formation of supported Pt-Sn nanoalloys by reactive metal-oxide interaction, RSC Adv., 6, 2016, 85688-85697, 10.1039/C6RA18801J
[63] Viktor Johánek, Anna Ostroverkh, Roman Fiala, Andrii Rednyk, and Vladimír Matolín, Mass Spectrometry of Polymer Electrolyte Membrane Fuel Cells, Journal of Analytical Methods in Chemistry, 2016 online, 2016, 1-9, 10.1155/2016/6097285
[64] Khalakhan, I; Lavková, J; Matolínová, I; Vorokhta, M; Potin, V; Kúš, P; Václavů, M; Maraloiu, V-A; Kuncser, A-C; Matolín, V, Electrochemically shape-controlled transformation of magnetron sputtered platinum films into platinum nanostructures enclosed by high-index facets, Surf & Coat Technol, 309, 2017, 6-11, 10.1016/j.surfcoat.2016.11.017
[65] Sergey M. Kozlov, Konstantin M. Neyman, Insights from methane decomposition on nanostructured palladium, J. of Catalysis, 337, 2016, 111-121,10.1016/j.jcat.2016.02.010
[66] Camellone, MF; Ribeiro, FN; Szabova, L; Tateyama, Y; Fabris, S, Catalytic Proton Dynamics at the Water/Solid Interface of Ceria-Supported Pt Clusters, J. Am. Chem. Soc., 138 (36), 2016, 11560-11567, 10.1021/jacs.6b03446
[67] Sergey M. Kozlov, Konstantin M. Neyman, Effects of electron transfer in model catalyst composed of Pt nanoparticles on CeO2(111) surface, J. of Catalysis, 344, 2016, 207-5015, 10.1016/j.jcat.2016.10.014
[68] Aleksandrov H.A. K.M. Neyman, K.I. Hadjiivanov, G.N. Vayssilov, Can the state of platinum species be unambiguously determined by the stretching frequency of adsorbed CO probe molecule?, Phys. Chem. Chem. Phys., 18, 2016, 22108-22121, 10.1039/C6CP03988J
[69] Bruix A., Neyman K.M. Modeling ceria-based nanomaterials for catalysis and related applications, Catal. Lett., 164, 2016, 2053-2080, 10.1007/s10562-016-1799-1
[70] Ferstl, P.; Mehl, S.; Arman, M. A.; Schuler, M.; Toghan, A.; Laszlo, B.; Lykhach, Y.; Brummel, O.; Lundgren, E.; Knudsen, J; Hammer, L; Schneider, MA; Libuda, J, Adsorption and Activation of CO on Co3O4(111) Thin Films, J. Phys. Chem. C, 119 (29), 2015, 16688-16699, 10.1021/acs.jpcc.5b04145
[71] Mehl, S.; Ferstl, P.; Schuler, M.; Toghan, A.; Brummel, O.; Hammer, L.; Schneider, M. A.; Libuda, J, Thermal evolution of cobalt deposits on Co3O4(111): atomically dispersed cobalt, two-dimensional CoO islands, and metallic Co nanoparticles, Phys. Chem. Chem. Phys., 17 (36), 2015, 23538-23546, 10.1039/c5cp03922c
Potential Impact:
The chipCAT project has advanced the field of heterogeneous catalysis in several ways, all fundamental ones and most of the applied ones are published in peer-reviewed journals and will impact the field in the standard process of scientific verification, discourse and adoption. Among the fundamental results we consider most prominent the conclusive evidence of single-atom catalyst actual existence and function. This realization shifts the paradigm of catalysis and may have manifold effect in all parts of the catalytic domain.
The achievement of developing a class of low-Platinum content catalysts suitable for PEMFC anodes is likely to be appreciated, gradually, as hydrogen fuel cell cars become more in mainstream demand: Saving Platinum will be of essence, because its global supplies do not cover the potential demand that would be created by “all cars on HFCs”. Currently the pricing of HFC cars starts around €500 000, which means (a) that the global fleet of HFC cars can be relatively tiny in size, so the global amount of Platinum will not figure, (b) that the current price of Platinum per car (ca. €1000) does not figure much in the total car price.
At present, the more attractive aspects of the nanocatalyst technology are the “dry” physical deposition of the layers and the remarkable durability of the membrane assemblies. These could bring the catalysts into mass production and use, because they both translate into reduced total cost of the technology in the end-user products. The industry as a whole may not grasp the technology yet, though: Currently, only anodes can be produced by the “dry” physical deposition route, while cathodes still need “wet” steps to achieve the industry target power outputs. Hence, companies working with “fully wet-route” catalyst layers are more likely to stick to their production lines. Once the properly efficient cathodes can be manufactured by physical deposition – if it happens – we expect to observe a massive shift in the industrial approach.
The newly developed fuel cell testing stations marketed as products constitute a minor addition to the global market in itself. Considering their performance, though, they have a great potential of becoming the test instrument of choice in laboratory and development environment. The first items were sold to synchrotron beamline laboratories to control fuel cells in operando analyses. Since the testing stations were developed with operando analyses in mind and seeing how operando spectroscopies and microscopies are in vogue in current world of catalysis research, we expect the product to help accelerate the field considerably.
The on-chip micro-FC technology has passed the proof of concept stage and has been brought to TRL 3-4. It is likely to be developed further and eventually become an alternative in mobile energy generation. Presently it still needs to pass practical hurdles to reach the stage, when tangible impact can be foreseen.
List of Websites:
URL: http://www.chipcat.eu
Corresponding contact for the project:
prof. RNDr. Vladimír Matolín, DrSc., matolin@mbox.troja.mff.cuni.cz +420-95155-2323
