Matteo Amati

Graphene oxide windows for in situ environmental cell photoelectron spectroscopy

The performance of new materials and devices often depends on processes taking place at the interface between an active solid element and the environment (such as air, water or other fluids). Understanding and controlling such interfacial processes require surface-specific spectroscopic information acquired under real-world operating conditions, which can be challenging because standard approaches such as X-ray photoelectron spectroscopy generally require high-vacuum conditions. The state-of-the-art approach to this problem relies on unique and expensive apparatus including electron analysers coupled with sophisticated differentially pumped lenses. Here, we develop a simple environmental cell with graphene oxide windows that are transparent to low-energy electrons (down to 400 eV), and demonstrate the feasibility of X-ray photoelectron spectroscopy measurements on model samples such as gold nanoparticles and aqueous salt solution placed on the back side of a window. These proof-of-principle results show the potential of using graphene oxide, graphene and other emerging ultrathin membrane windows for the fabrication of low-cost, single-use environmental cells compatible with commercial X-ray and Auger microprobes as well as scanning or transmission electron microscopes.

Soft X‐ray Imaging and Spectromicroscopy: New Insights in Chemical State and Morphology of the Key Components in Operating Fuel‐Cells

Fuel cells are one of the most appealing environmentally friendly devices for the effective conversion of chemical energy into electricity and heat, but still there are key barriers to their broad commercialization. In addition to efficiency, a major challenge of fuel-cell technology is the durability of the key components (interconnects, electrodes, and electrolytes) that can be subject to corrosion or undesired morphology and chemical changes occurring under operating conditions. The complementary capabilities of synchrotron-based soft X-ray microscopes in terms of imaging, spectroscopy, spatial and time resolution, and variable probing depths are opening unique opportunities to shed light on the multiple processes occurring in these complex systems at microscopic length scales. This type of information is prerequisite for understanding and controlling the performance and durability of such devices. This paper reviews the most recent efforts in the implementation of these methods for exploring the evolving structure and chemical composition of some key fuel cell components. Recent achievements are illustrated by selected results obtained with simplified versions of proton-exchange fuel-cells (PEFC) and solid-oxide fuel-cells (SOFC), which allow in situ monitoring of the redox reactions resulting in: 1) undesired deposits at interconnects and electrodes (PEFC); 2) material interactions at the electrode-electrolyte interface (PEFC); 3) release of corrosion products to the electrolyte phase (PEFC, and 4) mass-transport processes and structural changes occurring at the high operation temperatures of SOFC and promoted by the polarization.

Dynamic High Pressure: a novel approach toward near ambient pressure photoemission spectroscopy and spectromicroscopy

A Dynamic High Pressure (DHP) system has been developed, tested and implemented in the scanning photoelectron microscope (SPEM) operated at ESCAmicroscopy beamline at Elettra synchrotron. The system consists of a compact gas injection set up that allows experiments with local pressure near the sample several orders of magnitude higher that the allowable pressure for X-ray photoelectron spectroscopy setups. The DHP setup controls the amount of gas injected toward the sample by fine tuning the time and spatial profiles using a pulsed valve and a nozzle, respectively. The DHP functionality and effectiveness has been demonstrated by in operando oxidation experiments of Ru and Si. The obtained results confirmed that using the DHP the gas exposure onto the sample is equivalent to a static pressure between 10−3 and 10−2 mbar, about 3 orders of magnitude higher than the maximum gas pressure for the XPS machines under operation.

In Situ X‐Ray Spectromicroscopy Investigation of the Material Stability of SOFC Metal Interconnects in Operating Electrochemical Cells

The present in situ study of electrochemically induced processes occurring in Cr/Ni bilayers in contact with a YSZ electrolyte aims at a molecular-level understanding of the fundamental aspects related to the durability of metallic interconnects in solid oxide fuel cells (SOFCs). The results demonstrate the potential of scanning photoelectron microspectroscopy and imaging to follow in situ the evolution of the chemical states and lateral distributions of the constituent elements (Ni, Cr, Zr, and Y) as a function of applied cathodic potential in a cell working at 650 °C in 10(-6) mbar O(2) ambient conditions. The most interesting findings are the temperature-induced and potential-dependent diffusion of Ni and Cr, and the oxidation-reduction processes resulting in specific morphology-composition changes in the Ni, Cr, and YSZ areas.

In-situ Photoelectron Microspectroscopy and Imaging of Electrochemical Processes at the Electrodes of a Self-driven Cell

The challenges in development of solid oxide fuel cells (SOFCs) are reducing their dimensions and increasing their efficiency and durability, which requires physicochemical characterization at micro-scales of the device components during operation conditions. Recently, the unique potential of scanning photoelectron microscopy (SPEM) has been demonstrated by in-situ studies of externally-driven SOFCs, which mimic real devices. Here we overcome the gap between model and real systems using a single-chamber Ni|YSZ|Mn SOFC, supporting a range of self-driven electrochemical reactions in variable gas environments and temperatures. The reported SPEM results, obtained during spontaneous electrochemical processes occurring in reactive gas ambient, demonstrate the chemical evolution of electrodic material, in particular the lateral distribution of the oxidation state and the induced local potential, clearly marking out the electrochemically most active micro-regions of the Ni anode.

Quasi-in-situ single-grain photoelectron microspectroscopy of Co/PPy nanocomposites under oxygen reduction reaction.

This paper reports an investigation into the aging of pyrolyzed cobalt/polypyrrole (Co/PPy) oxygen reduction reaction (ORR) electrocatalysts, based on quasi-in-situ photoelectron microspectroscopy. The catalyst precursor was prepared by potentiostatic reverse-pulse coelectrodeposition from an acetonitrile solution on graphite. Accelerated aging was obtained by quasi-in-situ voltammetric cycling in an acidic electrolyte. Using photoelectron imaging and microspectroscopy of single Co/PPy grains at a resolution of 100 nm, we tracked the ORR-induced changes in the morphology and chemical state of the pristine material, consisting of uniformly distributed ∼20 nm nanoparticles, initially consisting of a mixture of Co(II) and Co(III) oxidation states in almost equal amounts. The evolution of the Co 2p, O 1s, and N 1s spectra revealed that the main effects of aging are a gradual loss of the Co present at the surface and the reduction of Co(III) to Co(II), accompanied by the emergence and growth of a N 1s signal, corresponding to electrocatalytically active C-N sites.

Intermetallics as key to spiral formation in In–Co electrodeposition. A study based on photoelectron microspectroscopy, mathematical modelling and numerical approximations

This paper reports on spiral pattern formation in In–Co electrodeposition. We propose an approach to the understanding of this process based on: (i) compositional and chemical-state distribution analysis by high-resolution photoelectron microspectroscopy and (ii) a mathematical model able to capture the morphological features highlighted in the experiments. Microspectroscopy—complemented by electrochemical, structural and morphological characterisations—combined with mathematical modelling, analytical and numerical investigations, converge in pointing out the key role played by intermetallic electrodeposition in spiral formation.

Electrodeposition and pyrolysis of Mn/polypyrrole nanocomposites: a study based on soft X-ray absorption, fluorescence and photoelectron microspectroscopies

Electrodeposition of manganese/polypyrrole (Mn/PPy) nanocomposites has been recently shown to be a technologically relevant synthesis method for the fabrication of Oxygen Reduction Reaction (ORR) electrocatalysts. In this study we have grown such composites with a potentiostatic anodic/cathodic pulse-plating procedure and characterised them by a multi-technique approach, combining a suite of in situ and ex situ spectroscopic methods with electrochemical measurements. We have thus achieved a sound degree of molecular-level understanding of the hybrid co-electrodeposition process consisting of electropolymerisation of polypyrrole with incorporation of Mn. By in situ Raman spectroscopy we followed the formation of MnOx and the polymer by monitoring the build-up and development of the relevant vibrational bands. The compositional and chemical-state distribution of the as-deposited material has been investigated ex situ by soft X-ray fluorescence (XRF) mapping and micro-absorption spectroscopy (micro-XAS). XRF shows that the spatial distribution of Mn is consistent in a rather wide range of current densities (c.d.s), while micro-XAS reveals a mixture of Mn valencies, with higher oxidation states prevailing at higher c.d.s. Pyrolysis of electrodeposits, desirable for obtaining more durable and active catalysts, has been followed in situ by photoelectron microspectroscopy, allowing to assess the evolution of: (i) the electrodeposit morphology, resulting in a uniform distribution of nanoparticles; (ii) the chemical state of manganese, changing from a mixture of valences to a final state consisting of Mn(III) and Mn(IV) oxides and (iii) the bonding nature of nitrogen, from initially N-pyrrolic to a combination of pyridinic and Mn–N/graphitic.

Chemically exfoliated MoS2 layers: Spectroscopic evidence for the semiconducting nature of the dominant trigonal metastable phase

A trigonal phase existing only as small patches on chemically exfoliated few layer, thermodynamically stable 1H phase of MoS2 is believed to influence critically properties of MoS2 based devices. This phase has been most often attributed to the metallic 1T phase. We investigate the electronic structure of chemically exfoliated MoS2 few layered systems using spatially resolved (≤120 nm resolution) photoemission spectroscopy and Raman spectroscopy in conjunction with state-of-the-art electronic structure calculations. On the basis of these results, we establish that the ground state of this phase is a small gap (~90 meV) semiconductor in contrast to most claims in the literature; we also identify the specific trigonal (1Tʹ) structure it has among many suggested ones. 2D transition metal dichalcogenides have emerged as a viable alternative to graphene with extraordinary properties and potential applications. Molybdenum disulfide (MoS2) is undoubtedly the preeminent member in the family for applications in transparent and flexible electronics. While the usual crystallographic form of MoS2 is the hexagonal 1H phase, MoS2 exhibits a number of trigonal polymorphic forms, such as 1T, 1T′, 1T′′ and 1T′′′ (see Fig. 1), distinguished by small distortions. These metastable states can be kinetically formed as small patches embedded in the majority 1H phase during chemical exfoliation, which is an attractive, easily scalable route to obtain one or few layer MoS2 in substantial quantities . Even mechanically exfoliated MoS2 may have small quantities of these metastable forms, influencing its material and device properties. The stable 1H form has been extensively studied and its electronic properties are well understood as a semiconductor with a large (1.9 eV) band gap. Unfortunately, electronic structures of different polymorphic MoS2 are not known, though it may potentially limit or enhance the applicability of two dimensional MoS2 devices by its presence within 1H MoS2 samples. It has been generally assumed that the metastable phase is of 1T form and metallic in nature. This presumed metallic nature is considered to be the cause of some novel beneficial device properties as well. For example, metallic 1T phase is believed to be responsible for the very high energy and power densities in supercapacitors and also for the remarkably high hydrogen evolution reaction efficiency achieved using chemically exfoliated MoS2. Whatever little is known of electronic structures of metastable phases is primarily from theoretical calculations that present contradictory views, ranging from being metallic (1T phase) to normal insulator (for 1T′ phase) or even ferroelectric insulator (1T′′′ phase). Surprisingly, direct structural investigations, based mostly on TEM and STM, also lack any agreement between different reports with the crystal structure of the metastable patches of MoS2 being variously ascribed to the 1T form, 19,26 the distorted 1T′′ form with a 2a×2a superstructure, the 1T′ form with a zigzag chain-like clustering of Mo atoms, and also the distorted 1T′′′ with a √3a×√3a superstructure where a trimerization of Mo atoms takes place (see Fig.1). Thus, the hope to understand the true electronic structure of this important phase of MoS2 via an experimental determination of its geometric structure in conjunction with electronic structure calculations has not been realized so far. The main difficulty in experimentally probing this metastable phase of MoS2 is that it exists only in small patches in the 1H matrix of few layer MoS2. Photoemission spectroscopy is the only direct probe of the electronic structure due to its inherent extreme surface sensitivity. However, the usual practice of photoemission spectroscopy does not have the required spatial resolution to enhance the relative contribution from the microscopic patches of the metastable phase. Therefore, we have carried out spatially resolved photoemission investigation with a ~120 nm photon beam diameter to directly determine the electronic structure of this metastable phase of MoS2 and conclusively establish that this elusive phase is actually a small gapped (~90 meV) semiconductor in sharp contrast to the dominant belief of it being metallic. We use state-of-the-art electronic structure calculations to provide evidence that this phase corresponds to the 1T′ structure, supported by micro-Raman experiments. The minority phase of MoS2 was extensively stabilized on conducting indium doped tin oxide (ITO) substrates using the well-known organolithium route. The mechanically exfoliated pure sample of MoS2 and the one after chemical treatment to stabilize the metastable polymorphic form are referred to as A and B, respectively in this manuscript. Raman scattering is a powerful and sensitive tool to detect presence of different phases of any given material and has been extensively used for low dimensional materials in recent times, providing vibrational fingerprints for different phases. Therefore, we characterized samples A and B using a micro-Raman probe with a spatial resolution of ~1 μm, with only three representative spectra shown for each sample. Raman spectra of the pure 1H sample (Fig. 2a) exhibit two peaks at 383 cm and 408.5 cm due to E2g and A1g modes, respectively, consistent with earlier reports. For sample B (Fig. 2b), we observe three additional peaks at 156 cm, 227 cm and 330 cm, referred to as J1, J2 and J3 modes respectively, in the literature 21,32 and associated with the formation of the metastable trigonal phase. We could not find any spot on this sample with only signals of the metastable phase with no signature of the 1H phase, clearly indicating that the patch size of this chemically induced MoS2 phase is smaller than 1 μm. Scanning Photo Electron Microscopy (SPEM) measurements were performed to understand the electronic structure of the chemically exfoliated system. Fig. 3a shows a typical photoelectron spectrum in the Mo 3d region with two narrow peaks at binding energies of 229.3 eV (Mo 3d5/2) and 232.4 eV (Mo 3d3/2) and a broad, low intensity feature at ~226.5 eV due to S 2s contributions. We compare this with a typical spectrum obtained from the sample B. Clearly, peaks in the spectrum of the mixed phase are specifically broadened on the lower binding energy side. Since the intensity of Mo 3d5/2 peak from pure 1H appears almost entirely between 228.7 and 230.0 eV in Fig. 3a, the additional intensity from sample B between 228 eV and 228.7 eV must arise from polymorphic forms of MoS2 other than 1H. This shift in the binding energy of the metastable form allows us to map its presence in the form of an image by scanning the sample through the focused photon beam and plotting the relevant Mo 3d5/2 intensity. In the SPEM imaging mode, first we carried out a detailed mapping of Mo 3d5/2 intensity over the entire 228-230 eV binding energy window, covering contributions from both phases and thus, imaging the distribution of MoS2 without any reference to its polymorphic forms on the ITO substrate as shown in Fig. 3b. Then, we plot in Fig. 3c an image of the ratio of intensities corresponding to energy windows, I and II, shown in Fig. 3a, corresponding to Mo 3d5/2 intensities arising primarily from the metastable phase and the 1H phase, respectively. The contrast in the intensity ratio, I/II, being independent of topographic features, reveals the relative abundance of the metastable phase in the sample B with the dark blue regions corresponding to the metastable MoS2 rich region. We point out that almost all spots imaged in Fig. 3c contain signals of both the 1H and the metastable phase, explicitly checked by recording Mo 3d5/2 spectra from over 30 different spots; there were only few spots that corresponded to the pure 1H phase and no spot that had contribution only from the metastable phase. Coupled with the fact that there is a considerable intensity contrast even with the present photon spot size, this implies that the typical size of metastable patches is in the order of ≤120 nm, but not very much smaller. We identified a region with the largest contribution from the metastable phase, marked by the rectangular window in Fig. 3c; the intensity of the Mo 3d5/2 from the metastable phase was essentially uniform over this region. Thereafter, we carried out a detailed spectroscopic investigation with the photon beam focused at the center of this window to maximize contributions from the metastable phase that we are interested in. The Mo 3d spectrum obtained from this spot is shown in Fig. 3d. We have decomposed this spectrum in terms of contributions from the 1H phase and the metastable phase using a least squared-error approach with the spectral feature of the 1H component as determined from measurements of sample A and that for the metastable phase approximated by a Lorentzian line convoluted with a Gaussian function with full width at the half maximum, FWHM, representing the life-time and the resolution broadenings. The resulting components, also shown in Fig. 3d, establish the dominance of the metastable phase at this spot on the sample, with the area ratio (~2.8) of the two components providing a quantitative measure of the relative abundance of the two phases. We also note that the electronic binding energy difference (~0.7 eV) between the two components is in agreement with earlier publications. 23 There have been suggestions in the literature that the presence of Li ions on such chemically exfoliated samples and consequent charge doping may influence the formation of a specific metastable state. Therefore, we scanned carefully the binding energy region corresponding to Li 1s level with a high counting statistics and found no evidence of any presence of Li in our samples. This shows that the sample preparation method, that involves repeated washing of the chemically exfoliated samples first with hexane and then with water is effective in removing all traces of

Spatially Resolved Chemical Characterization with Scanning Photoemission Spectromicroscopy: Towards Near‐Ambient‐Pressure Experiments

The major experimental challenges in investigations of heterogeneous catalysis are the morphologically complex and dynamic micro‐ and nanosystems and the exploration of events that occur at the catalyst surface, which determine the catalyst activity and selectivity. Modern‐day investigations of catalytic reactions require a multitechnique experimental and computational approach, in which each tool provides specific and complementary information. The unique combination of surface and chemical sensitivity has ranked X‐ray photoelectron spectroscopy (XPS) as one of the most important experimental methods for the characterization of catalytic systems. After its invention, more than half a century ago, a revolutionary step in XPS development was the addition of sub‐micrometer lateral resolution intermediate between light microscopy and electron microscopy. XPS microscopes have responded to the increasing demands on nanotechnology to have access to the local chemical composition, electronic and magnetic structure, and reorganization processes at morphologically complex surfaces and interfaces. The high spatial resolution in XPS microscopes is achieved by two different approaches, magnification of the image of the irradiated surface area (X‐ray photoemission electron microscopy) or demagnification of the incident photon beam by using X‐ray focusing optics (scanning photoemission microscopy; SPEM). In the present article, using selected examples, we demonstrate the capabilities of SPEM for studies relevant to catalysis, and we will discuss the next steps in the ongoing development. In the first part, we present the successful characterization of the oxidation of (i) polycrystalline PtRh particles and (ii) Pd thin films that decorate carbon nanotubes. In the second part, we describe two new setups, developed at Elettra, to overcome the “pressure gap” for photoemission spectromicroscopy experiments, which is the major limitation in the exploration of “real world” catalytic reactions. The first measurements of core‐level photoemission spectroscopy and imaging obtained with spatial resolution of the order of 100 nm at near‐ambient pressure are presented.