Stephan Hofmann

Reliable pH sensing in liquid environments underpins applications in energy storage, biosensing, and process monitoring, yet conventional graphene-based electrical devices introduce ohmic contacts, encapsulation layers, and electrochemical artifacts that can obscure intrinsic behavior. These approaches also fail to capture the strong spatial variability of commercial CVD graphene. Here, we present a fully contactless methodology for quantifying pH-dependent conductivity of large-area CVD graphene using through-substrate reflection terahertz time-domain spectroscopy (THz-TDS). A customized liquid cell with a THz transparent window preserves unrestricted access to the graphene–electrolyte interface while enabling centimeter-scale conductivity mapping and in-operando single-point monitoring during sequential pH modulation. By integrating spatially registered Raman mapping with automated quality-metric filtering, we extract statistically robust trends across thousands of pixels rather than isolated device regions. We show that pristine monolayer graphene exhibits the highest baseline conductivity and strongest pH sensitivity, whereas O2-plasma-introduced defects systematically suppress both. This scalable, remote, and inert probing approach establishes THz-TDS as a powerful platform for high-throughput characterization and robust liquid-phase sensing in thin films and 2D systems.

One of the most promising building blocks for the development of spin qubits, single-photon sources, and quantum sensors at room temperature, as well as 2D ultraviolet light-emitting diodes, is defect color centers in 2D hexagonal boron nitride (hBN). However, a significant requirement for the realization of such devices toward scalable technologies is the deterministic localization of hBN color centers. Here, we demonstrate a novel approach to the localization of hBN color center emission by using patterned dielectric layers grown on graphene via atomic layer deposition (ALD). While color center emission is quenched within areas where hBN is transferred directly on graphene due to charge transfer, it is maintained where hBN is transferred on micrometer-sized Al2O3 pillars, which act as barriers to charge transfer. Importantly, our approach allows for device architectures where graphene layers are used as top and bottom electrodes for the application of vertical electric fields, such as for carrier injection in electroluminescent devices, Stark shifting of color center emission, and charge state control of defect color centers. Furthermore, the use of ALD to grow dielectric layers directly on graphene allows for the control of tunnel barrier thicknesses with atomic layer precision, which is crucial for many of these device applications. Our work represents an important step toward the realization of a wide range of scalable device applications based on the deterministic localization of electrically controlled hBN color centers.

Inorganic colloidal quantum dot light-emitting diodes could be used to build next-generation electroluminescent displays due to their colour properties and electrical stability. However, to create high-resolution and large-area displays, a pixel integration method is required, which can deposit quantum dot arrays on an active-matrix backplane and maintain uniformity and precision, without colour cross-contamination. Here we report a cracking-assisted transfer printing technology that can be used to pattern high-resolution full-colour pixel arrays over large areas. The technology uses a controlled cracking process to fracture interparticle cohesive bonds between quantum dots. This facilitates subsequent pick-up and transfer to a thin-film transistor backplane with high precision. With the technology, we achieve pixels down to a size of 600 nm with electroluminescent emission and uniform pixelization over areas up to 4 inches. We create a cadmium-free full-colour active-matrix display with a resolution of 341 pixels per inch, as well as a blue active-matrix display with a flexible form factor. Furthermore, the cracking-assisted transfer printing can improve electroluminescence performance—with higher maximum luminance and operational lifetime than other quantum dot patterning techniques—through precise nano-interface control and high quantum dot packing density.

Atomic layer deposition (ALD) of an ∼0.4 nm iron oxide overlayer on a Pd/Al2O3 catalyst was investigated. TEM/EDS mapping, XRD, and XPS analyses confirmed the formation of an Fe2O3 layer over the Pd particles, forming a protective ‘core-shell’ structure. The deposited film remained thermally stable and XRD-amorphous, with crystallinity related to α-Fe2O3 emerging above 900 °C. All catalytic assessments and characterizations were conducted under conditions where the Fe2O3 phase was defined. After high temperature treatment, the Fe2O3 layer maintained good interfacial contact with Pd, and this interaction inhibited coking from deactivating Pd despite exposure to CH4-rich streams at 900 °C for 100 h of operation. In contrast, conventional catalysts, prepared by incipient wetness infiltration (IWI), deactivated over this duration. When these catalysts were tested for dry reforming of methane (DRM), the ALD catalyst maintained its activity under differential conditions while the IWI-prepared catalysts exhibited decreased reaction rates after they were pretreated under CH4-rich conditions at 900 °C. Rich CH4 conditions were selected to stress the catalysts at shorter timescales. Additionally, the morphology of the coke formed on the ALD catalyst differed from that formed on IWI catalysts. Based on TGA and XPS, the carbon formed on the ALD catalyst was more disordered and could be oxidized in air at lower temperatures (500 °C), whereas the carbon formed on uncoated Pd exhibited ordered whisker-like structures and was more resistant to oxidation (700 °C). Finally, mechanistic insights into the role of Fe2O3 in stabilizing Pd for DRM and inhibiting coke formation were supported by XPS analysis.

Tungsten diselenide is a prominent candidate for nanoscale electronics and complementary metal–oxide–semiconductor technology due to its controllable p-type and n-type doping that is achievable through selective oxidation. However, characterizing the resulting oxide quality often relies on destructive or slow metrology. In this work, we demonstrate spectroscopic imaging ellipsometry (SIE) as a noninvasive technique to characterize the oxidation state of two-dimensional materials, providing spatially resolved measurements of layer thickness and uniformity. We apply this method to compare few-layer WSe2 after thermal or plasma-based oxidation treatments. We show that plasma oxidation enables a highly controllable conversion of individual WSe2 layers into uniform sub-stoichiometric tungsten oxide (WOx), whereas thermal oxidation in atmospheric conditions yields a rough interface characterized by non-uniform oxide formation and blistering. Using SIE, we quantify a layer-by-layer conversion ratio, where a single WSe2 monolayer transforms into ∼1.6 nm of WOx during plasma oxidation. These findings, corroborated by atomic force microscopy, highlight the capability of SIE to resolve dielectric evolution and interface quality in layered semiconductors.

The commercialization of emerging materials is hindered by the empirical nature of process development, with the underpinning solid-state reaction kinetics remaining elusive due to their inherent multistep and multiscale character and the vast configurational and parameter space. We combine high-throughput operando scanning electron microscopy (OSEM) with extended classical continuum and phase-field simulations and atomistic machine-learned interatomic potential (MLIP) surrogate models to demonstrate effective foundational reaction exploration, using thermal oxidation of chemical-vapor-deposited monolayer WS₂ across a temperature range of 450–680 °C as our benchmark reaction. OSEM provides statistically relevant reaction data sets of spatiotemporal basal plane nucleation kinetics and propagation of tens of thousands of individual 1D reaction facets, revealing time-dependent rates. By extending Avrami theory and employing a calibrated phase-field model, we extract an apparent nucleation barrier of 1.1 eV and show that the complex rate behavior arises from mixed reaction–diffusion control. Atomistic reaction exploration via MLIPs guided by these experimental data reveals that oxidative reaction chains leading to W volatilization and etch pit formation are driven not by the most common sulfur or substitutional oxygen point defects but by more complex defects such as W vacancies. MLIP diffusion barrier screening identifies the role of chemisorbed hydroxyl species for this reaction scenario, while systematic screening of 1D edge configurations and their terminations uncovers the structural origins of the pronounced in-plane reaction anisotropies. We discuss the potential of our synergistic approach to effectively bring experimental and computational approaches closer together and accelerate critically required process discovery for advanced materials.

The optical excitation of a bright exciton may be followed by the formation of lower-energy dark states. In these formation and relaxation processes, nonequilibrium exciton and phonon populations play a dominant role but remain so far largely unexplored, as most states are inaccessible by regular spectroscopies. Here, on the example of homobilayer 2 H − MoS 2 , we realize direct access to the full exciton relaxation cascade from experiment and theory. We find distinct changes in the time-, energy-, and in-plane momentum-resolved photoemission spectral function that can be explained only when considering the formation and subsequent thermalization of excitonic nonequilibrium occupation distributions. In agreement with microscopic many-particle calculations, we quantify the timescales for the formation of a nonequilibrium dark-excitonic occupation and its subsequent thermalization to 85 and 150 fs, respectively. Our results provide a previously inaccessible view of the complete exciton relaxation cascade, which is of importance for the future characterization of nonequilibrium excitonic phases and the efficient design of optoelectronic devices.

The advent of 2D materials has revolutionized condensed matter physics and materials science, offering unprecedented opportunities to explore exotic physical phenomena, engineer novel functionalities, and address critical technological challenges across diverse fields. Over the past two decades, the exploration of 2D materials has expanded beyond graphene, encompassing a vast library of atomically thin crystals and their heterostructures. These materials exhibit extraordinary electronic, optical, thermal, mechanical, and chemical properties, and hold promise for breakthroughs in electronics, optoelectronics, quantum technologies, energy storage, catalysis, thermal management, filtration and separation, and beyond. Many exciting new physics and phenomena continue to emerge, while select 2D materials, such as graphene, h-BN, and the semiconducting transition metal dichalcogenides (TMDCs), are transitioning from laboratory-scale demonstrations to industrial applications. In this context, a holistic understanding of synthesis, structure-property relationships, integration, and performance optimization is essential. This roadmap reviews the multifaceted challenges and opportunities in 2D materials research, focusing on the synthesis, properties and applications of representative systems including graphene and its derivatives, TMDCs, MXenes as well as their heterostructures and moiré systems.

Atomistic simulations of matter, especially those that leverage first-principles (ab initio) electronic structure theory, provide a microscopic view of the world, underpinning much of our understanding of chemistry and materials science. Over the last decade or so, machine-learned force fields have transformed atomistic modeling by enabling simulations of ab initio quality over unprecedented time and length scales. However, early machine-learning (ML) force fields have largely been limited by (i) the substantial computational and human effort required to develop and validate potentials for each particular system of interest and (ii) a general lack of transferability from one chemical system to the next. Here, we show that it is possible to create a general-purpose atomistic ML model, trained on a public dataset of moderate size, that is capable of running stable molecular dynamics for a wide range of molecules and materials. We demonstrate the power of the MACE-MP-0 model—and its qualitative and at times quantitative accuracy—on a diverse set of problems in the physical sciences, including properties of solids, liquids, gases, chemical reactions, interfaces, and even the dynamics of a small protein. The model can be applied out of the box as a starting or “foundation” model for any atomistic system of interest and, when desired, can be fine-tuned on just a handful of application-specific data points to reach ab initio accuracy. Establishing that a stable force-field model can cover almost all materials changes atomistic modeling in a fundamental way: experienced users obtain reliable results much faster, and beginners face a lower barrier to entry. Foundation models thus represent a step toward democratizing the revolution in atomic-scale modeling that has been brought about by ML force fields.