Metal organic vapor phase epitaxy (MOVPE)

Metal organic vapor phase epitaxy (MOVPE) is a process for producing thin, high-quality crystal layers. Special gases, known as precursors, are fed into a reactor in which a wafer is heated. The molecules disintegrate on the surface of the wafer, releasing atoms that are deposited in an orderly manner, forming a new material layer by layer. Unwanted residues escape as gas or settle on the reactor walls.

MOVPE is particularly important for the production of III-V semiconductors such as gallium arsenide or gallium nitride. This method combines fast layer deposition with high quality and forms the basis of many modern technologies. MOVPE is used to produce LEDs, laser diodes, fast transistors, and solar cells, among other things. Although the principle seems simple—molecules decompose on a hot surface—the process allows for extremely precise control down to the arrangement of individual atoms.

Agnieszka Paszuk

Atomic layer deposition (ALD)

Atomic layer deposition (ALD) is a modified process of chemical vapor deposition. During that process self-limiting surface reactions allow to deposit extremely thin and homogeneous layers with low-defect density. Metal oxide layers, such as TiO2 has been shown to exhibit much higher stability than III-V semiconductors when in contact with the electrolyte under operation. Despite efforts to optimize thin film depositions, there is still ambiguity regarding the reaction mechanisms of the precursor with the substrate (interface formation) and the impact of the process parameters on the properties of the grown layer, such as structure, purity and uniformity, and more importantly, its impact on the chemical stability.

In particular, defects, which originate at the III-V semiconductor/metal-oxide heterointerface will affect the film stability once exposed to the electrolyte under light illumination. Choice of the precursors for the metal and oxide, growth temperature and the substrate preparation will strongly influence the chemical and electronic structure at the heterointerface, and the structure, surface morphology and defects of the grown film.

Manali Nandy

Electron Channeling Contrast Imaging (ECCI)

Electron Channeling Contrast Imaging (ECCI) is an advanced imaging technique used in scanning electron microscopy (SEM) to visualize crystal defects such as dislocations and stacking faults near the surface of bulk crystalline materials. Unlike traditional methods that often require thin samples, ECCI allows direct observation of defects in bulk samples without extensive preparation. The technique works by using a backscattered electron detector to measure changes in how electrons scatter when they enter a crystal. In a perfectly ordered crystal, electrons travel along specific paths between atomic planes, a process known as electron channeling. However, when the electron beam encounters a defect, the regular pattern is disrupted, altering the phase of the electron wave and changing the likelihood of electrons being scattered back out of the surface. These changes create visible contrast in the SEM image, revealing the location and nature of the defects. The phenomenon of channeling was first observed in 1967 by D.G. Coates, who noticed Kikuchi-like bands in SEM images of single crystals. These bands are caused by diffraction of the incident electrons inside the crystal and form the basis of ECCI. Because of its ability to reveal crystal imperfections with high spatial resolution and minimal sample preparation, ECCI is a powerful and non-destructive tool in materials science.

Juliane Koch

Photoelectrochemistry

Photoelectrochemical (PEC) water splitting involves investigating how sunlight can be converted directly into hydrogen. This requires a special experimental setup: a solar simulator serves as the light source, a potentiostat controls and measures current and voltage, a gas chromatograph (GC) analyzes the gases produced, and the actual reaction takes place in a PEC chamber.

The semiconductor device to be tested is inserted into the chamber and connected at the rear. In addition, a reference electrode and a counter electrode are located in the chamber, which is filled with a current-conducting electrolyte. When light hits the sample, gas bubbles form on its surface. These are analyzed by the GC to determine their composition, while the potentiostat continuously monitors voltage and current.

This allows not only the efficiency of water splitting to be measured, but also the long-term stability of the materials. The results form an important basis for specifically improving semiconductor structures and developing applications with high conversion efficiency.

Juliane Koch

Multi-tip scanning tunneling microscopy

An ultra-high vacuum-based multi-tip scanning tunneling microscope (MT-STM) is a high-precision measuring device that can be used to examine the electrical properties of tiny semiconductor components.

The MT-STM is equipped with an integrated scanning electron microscope and uses piezoelectric nanopositioners to position four tungsten tips independently of each other with an accuracy of approximately 100 nm. This allows individual free-standing nanowires to be specifically contacted and measured. By varying the distance between the measuring tips, current-voltage characteristics can be recorded and spatially resolved resistance and doping profiles determined. This method is particularly important for researching special contacts in semiconductors that separate charge carriers from each other – a crucial process for many optoelectronic components such as solar cells, LEDs, and photodetectors.

Juliane Koch

Scanning electron microscopy

In order to examine semiconductor structures in the micro- and nanometer range, a very high resolution is required that ordinary light microscopes cannot provide. For this reason, a scanning electron microscope (SEM) is used to analyze these structures. In this process, an electron beam strikes the sample and interacts with its material. The resulting secondary and backscattered electrons are detected and converted into high-resolution images.

Since electrons have a much shorter wavelength than light, the SEM achieves a significantly higher resolution, with magnifications of 100,000 times possible without any problems. This allows even tiny components such as nanowires to be analyzed reliably.

In addition, characteristic X-rays are generated when the sample interacts with the electron beam. These can be measured with special detectors and enable a spatially resolved chemical analysis of the sample. This method is called energy-dispersive X-ray spectroscopy (EDX).

Juliane Koch

Magnetron sputtering

Thin metal oxide layers, initially TiOx, are deposited as III-V-free window or protective layers. The layer structure and morphology as well as the optical and electronic properties must be adapted to the III-V material. In addition to ALD (see reference here), the project uses magnetron sputtering from the physical vapor deposition group. (In this process, Ar ions from a plasma are accelerated onto a target (here Ti) and knock out atoms, which then settle on the substrate (here the III-V material). By mixing reactive gases, such as O2, into the process gas, compounds such as TiOx can also be deposited). The layer growth can be influenced by factors such as process pressure, growth rate, and deposition angle. The sputtering system used in the project also offers the possibility of depositing mixed oxides by depositing several targets simultaneously, or of varying other layer properties thanks to the free programmability of many machine parameters.
Simone Gutsche

X-ray diffraction (XRD)

The crystalline structure of the III-V materials and window layers is examined using X-ray diffraction methods. Several systems are available for this purpose: High-resolution XRD is used to examine the orientation, cell parameters, and defects of the III-V materials, while measurements for characterizing the crystallinity and modifications of the thin TiOx window layers are performed using another device with a parallel beam.

Fraunhofer ISE, Dirk Mahler

Electrochemical etching of germanium substrates

Porous germanium layers can be produced by means of electrochemical etching. The aim is to etch two layers on top of each other, a low-porosity layer near the surface and a high-porosity layer underneath, in order to enable the germanium substrate to be split for later reuse. In this process, the germanium substrates form an electrode in an acidic electrolyte. The structure of the porous layer is significantly influenced by the flowing current, with low currents resulting in low-porosity layers and high currents resulting in high-porosity layers. A special feature here is that the direction of the current must be reversed repeatedly during etching, which protects the surface that has already been etched. During a subsequent high-temperature step, the highly porous layer transforms into a release layer, while the surface of the low-porosity layer closes, thus forming a good growth template for III-V semiconductor layers. After release, the substrate part can be reused.

[Image: Screenshot window by Thi Thu Huong Do. Copyrights of the software user interface in this screenshot belongs to Wolfram Research and Modelica Association]

Wolfram System Modeler v.14.3 and System Dynamics Modelling

Wolfram System Modeler, when combined with the Wolfram Language, brings together intuitive graphical modeling and advanced computational capabilities. It provides an interactive environment for building, exploring, and simulating models, drawing on both symbolic and numerical methods to represent real-world processes. The Modelica language, designed for dynamic systems, streamlines mathematical modeling and offers a Standard Library that can be extended with additional packages- supporting large-scale collaborative projects such as SustEnMat. Within this framework, the System Dynamics library implements Jay Forrester’s methodology for representing continuous-time mass and information flows. This approach, widely applied in the social sciences and economics, uses stock-and-flow and causal loop diagrams to simplify the representation of underlying differential equations. Our objective is to develop an integrated dynamic model of the global market for group III‑V elements, simulating extraction, supply, demand, pricing, usage, and recycling across a range of potential scenarios.

 
[Image: Screenshot window by Thi Thu Huong Do. Copyrights of the contents and user interface in this screenshot belongs to pgAdmin platform and PATSTAT]

PATSTAT Database and patents analysis using pgAdmin4 (a management tool for PostgreSQL)

The EPO Worldwide Patent Statistical Database (PATSTAT), provided by the European Patent Office, is a key reference for patent intelligence and statistics. It contains bibliographic and legal event data from patent offices worldwide, covering both industrialized and developing countries. We use tools such as pgAdmin4 to query the relational database, run statistical analyses, and create visualizations. By applying methods such as technological progress modeling (e.g., Farmer & Lafond, 2016), we monitor technology trends and commercialization developments in solar panel technologies.

Siddharood Biradar

Umberto LCA Software

Umberto is a professional life cycle assessment and material flow analysis software used to model, quantify, and evaluate the impacts of products, processes, and organizations throughout the entire life cycle.

The software provides a graphical modelling environment in which process nodes and flow connections represent complex systems with high transparency. Integrated databases, such as ecoinvent, supply consistent life cycle inventory data, while the calculation engine applies established LCA methodologies in accordance with ISO 14040/44 standards. In this way, energy and material flows can be accurately mapped and analyzed using a variety of life cycle impact assessment (LCIA) methods, which enable comprehensive evaluation of environmental performance.

By defining system boundaries and adjusting process parameters, scenario analyses and optimization strategies can be developed. The results can be visualized as Sankey diagrams, impact assessment charts, or comparative life cycle profiles. This approach is particularly valuable for identifying environmental hotspots, improving product sustainability, and supporting strategic decision-making in sectors ranging from manufacturing and energy to circular economy applications.

Molecular Dynamics

We use quantum-based molecular dynamics to investigate how water interacts with a range of surfaces and interfaces, including III-V semiconductors, metal oxides, and combined oxide/semiconductor systems with pinhole defects. These simulations provide atomistic insights into the structure and dynamics of solid-liquid interfaces, such as hydrogen-bond networks, adsorption and dissociation pathways, and the first steps of corrosion. The resulting trajectories also serve as high-quality reference data for training our machine-learning models.

Machine-learning force fields (MACE)

To study larger systems and longer timescales, we apply the MACE framework for machine-learning force fields. By training on ab initio data, MACE achieves near-DFT accuracy at a fraction of the computational cost. This enables simulations over hundreds of picoseconds to nanoseconds, capturing slow water dissociation events, interface restructuring, the growth of pinhole defects, and the influence of film thickness. In this way, we can assess the reactivity and stability of realistic solid-liquid interfaces relevant for sustainable energy and corrosion-resistant applications.