The efficient identification of compositional areas of interest in thin film materials systems fabricated by combinatorial deposition methods is essential in combinatorial materials science. We use a combination of compositional screening by EDX together with high-throughput measurements of electrical and optical properties of thin film libraries to determine efficiently the areas of interest in a materials system. Areas of interest are compositions which show distinctive properties. The crystallinity of the thus determined areas is identified by X-ray diffraction. Additionally, by using automated nanoindentation across the materials library, mechanical data of the thin films can be obtained which complements the identification of areas of interest. The feasibility of this approach is demonstrated by using a Ni-Al thin film library as a reference system. The obtained results promise that this approach can be used for the case of ternary and higher order systems.

We apply variational autoencoders (VAE) to X-ray diffraction (XRD) data analysis on both simulated and experimental thin-film data. We show that crystal structure representations learned by a VAE reveal latent information, such as the structural similarity of textured diffraction patterns. While other artificial intelligence (AI) agents are effective at classifying XRD data into known phases, a similarly conditioned VAE is uniquely effective at knowing what it doesn’t know: it can rapidly identify data outside the distribution it was trained on, such as novel phases and mixtures. These capabilities demonstrate that a VAE is a valuable AI agent for aiding materials discovery and understanding XRD measurements both ‘on-the-fly’ and during post hoc analysis.

Thin films are ubiquitous in modern technology and highly useful in materials discovery and design. For achieving optimal extrinsic properties, their microstructure needs to be controlled in a multi-parameter space, which usually requires too high a number of experiments to map. Here, we propose to master thin film processing microstructure complexity, and to reduce the cost of microstructure design by joining combinatorial experimentation with generative deep learning models to extract synthesis-composition-microstructure relations. A generative machine learning approach using a conditional generative adversarial network predicts structure zone diagrams. We demonstrate that generative models provide a so far unseen level of quality of generated structure zone diagrams that can be applied for the optimization of chemical composition and processing parameters to achieve a desired microstructure.

In thin film deposition processes, the deposition temperature is one of the crucial process parameters for obtaining films with desired properties. Usually the optimum deposition temperature is found by conducting several depositions sequentially in a time consuming process. This paper demonstrates a facile and rapid route of the simultaneous thin film deposition at six different deposition temperatures ranging from 100 to 1000 8C. Cuprite (Cu 2 O) was chosen for the study as this material is of interest for energy applications. The thin films are assessed for their crystallographic, microstructural, Raman scattering, and photoelectrochemical properties. The results show that the utilization of a step heater leads to the rapid optimization of thin film microstructures of an absorber material used in photoelectrochemistry. This results in a structure zone diagram for Cu 2 O. For a substrate temperature of 600 8C, an optimum between crystallinity and morphology occurs. 1 Introduction Thin film semiconductors in photoelectrochemical applications need to fulfil numerous properties. Whereas the search for new absorber and catalytic materials by combinatorial and high-throughput methods is currently addressed, the high-throughput optimization of deposition parameters, e.g., deposition temperature, however, is less developed. Finding optimal deposition conditions for even relatively simple metal oxide semiconductors for photoelectrochemical applications [1], usually requires careful evaluation of the different properties of a thin film like morphology, crystal structure, and interface formation [2][3][4][5]. In reactive magnetron sputtering, these properties can be tuned by deposition parameters like applied power, gas flow, substrate rotation, deposition geometry, total or partial gas pressure, and substrate temperature. These parameters are, however, interdependent. Attaining the desired metal oxide phase (e.g., competing phase formation between CuO and Cu 2 O) adds to the overall complexity of the problem of finding optimal deposition parameters [6]. Evaluating all possible optimization

The complete ternary system Ni-Co-Al was fabricated as a thin film materials library by combinatorial magnetron sputtering and was annealed subsequently in several steps in Ar and under atmospheric conditions at 500 °C. Ni-Co-Al is the base system for both Ni- and Co-based superalloys. Therefore, the phases occurring in this system and their oxidation behavior is of high interest. The Ni-Co-Al materials library was investigated using high-throughput characterization methods such as optical measurements, resistance screening, automated EDX, automated XRD, and XPS. From the obtained data a thin film phase diagram for the Ni-Co-Al system in its state after annealing at 500 °C in air was established. Furthermore, a surface oxide composition map of the full Ni-Co-Al system for oxidation at 500 °C was concluded. As a result, it could be shown that at 500 °C an amount of 10 at. % Al is necessary for a Ni-Co-Al thin film to produce a protective Al-oxide scale.