#research, paper,

Designing Workflows for Materials Characterization

(2023) - Sergei V. Kalinin, Maxim Ziatdinov, Mahshid Ahmadi, Ayana Ghosh, Kevin Roccapriore, Yongtao Liu, Rama K. Vasudevan

Abstract

Experimental science is enabled by the combination of synthesis, imaging, and functional characterization. Synthesis of a new material is typically followed by a set of characterization methods aiming to provide feedback for optimization or discover fundamental mechanisms. However, the sequence of synthesis and characterization methods and their interpretation, or research workflow, has traditionally been driven by human intuition and is highly domain specific. Here we explore concepts of scientific workflows that emerge at the interface between theory, characterization, and imaging. We discuss the criteria by which these workflows can be constructed for special cases of multi-resolution structural imaging and structural and functional characterization. Some considerations for theory-experiment workflows are provided. We further pose that the emergence of user facilities and cloud labs disrupt the classical progression from ideation, orchestration, and execution stages of workflow development and necessitate development of universal frameworks for workflow design, including universal hyper-languages describing laboratory operation, reward functions and their integration between domains, and policy development for workflow optimization. These tools will enable knowledge-based workflow optimization, enable lateral instrumental networks, sequential and parallel orchestration of characterization between dissimilar facilities, and empower distributed research.

Figures

Pasted image 20230310172959.png

Notes

Annotations (3/10/2023, 5:28:28 PM)

“ne of the materials systems where these limitations are particularly important are the hybrid perovskites for solar cells and other optoelectronic applications.13-16” (Kalinin et al., 2023, p. 4) #nodes_research references for high-throughput perovskite

“Despite some early demonstrations, this concept became mainstream only in the last 5 years, as a result of the large-scale efforts by B. Maruyama et al.26-29, A. Aspuru-Guzik et al.27, 30, 31, A. Abdolhasani et al.32-34,” (Kalinin et al., 2023, p. 5) #nodes_research
effort regarding high-throughput experiments

“high throughput and combinatorial studies of hybrid perovskite materials by M. Ahmadi et al.35-37, C. Brabec et al.38, 39” (Kalinin et al., 2023, p. 5) #nodes_research
high-throughput experiment for perovskite

“Particularly over the last five years, a number of machine learning approaches based on variational autoencoders,79 generative adversarial networks, and diffusion models have been suggested to bridge length and time scales in simulations, establish statistically significant descriptors such as order parameters, and determine their constitutive relations.” (Kalinin et al., 2023, p. 13)

“probabilistic machine learning (PML) framework based, for example, on Gaussian process, Bayesian neural network, or deep kernel learning,” (Kalinin et al., 2023, p. 15) gaussian process includes in probabilistic machine learning (PML) framework

“Recently, a deep residual learning framework with holistically-nested edge detection (ResHedNet) was ensembled to minimize the out-of-distribution drift effects in real-time SPM measurement.86 The ensembled ResHedNet was implemented on an operating SPM, where it converted the real-time SPM data stream to segmented objects of interest, e.g. ferroelastic domain wall or polycrystal grain boundary images.” (Kalinin et al., 2023, p. 18) ML for edge detection in SPM

“behavior of grain boundary junction points in metal halide perovskites was discovered.88” (Kalinin et al., 2023, p. 19) using ML in SPM the behavior of grain boundary of perovskite was discovered

“The deep kernel learning (DKL) algorithm learns what elements of the materials structure maximize this reward and guides the exploration of materials surface accordingly.” (Kalinin et al., 2023, p. 19)

Comment: 33 pages; 8 figures

need to be zotero-ed