Neural Network based “inversion” of XAS to Structural Descriptors

Prahlad Kumar Routh

“Machine learning based unsupervised approach for nanostructure characterization using theoretical and experimental X-ray absorption spectroscopy”

Local structural information about nanoparticles during catalysis can be measured quantitatively by X-ray absorption fine structure spectra. While there are theoretical approaches to analyze the extended part of XAFS, namely EXAFS fitting, the near edge region (XANES) is only qualitatively analyzed using forward modeling approaches. Deep learning-based XANES analysis (Xiang et. a. 2022) was successfully applied to this new materials system to characterize the atomic neighborhood of single atoms.

 

Latent Representation Learning for Structural Characterization of Catalysts

Prahlad Kumar Routh

Local structural information about nanoparticles during catalysis can be measured quantitatively by X-ray near edge absorption spectra (XANES) studies using Neural Network methods. However, the extent of information encoded in XANES spectra is not known a priori. We developed an approach to solving the information content problem: Latent Space Analysis of Spectra (LSAS).

 

LSAS approach compresses the information in the input (spectrum) to a much smaller dimensionality of the latent space, which is then used to analyze and extract structural descriptors. Our information centric approach identifies and extracts the key descriptors in the hydrogen disassociation and Pd hydride formation reaction. Such an unsupervised approach was demonstrated to be crucial for learning factors behind the real-time changes in catalytic activity.

For more details, read our manuscript: Latent Representation Learning for Structural Characterization of Catalysts, Phys. Chem. Lett. , 12, 2021, DOI: 10.1021/acs.jpclett.0c03792

0D-2D Semiconducting Hybrid Nanomaterials

Prahlad Kumar Routh

Layered metal dichalcogenides (LMDs) are emerging 2D semiconductors for energy harvesting, chemical sensing and photon detection applications due to their tunable band gap with thickness, high surface-to-volume ratio and strong light-matter interactions but exhibit poor light absorption properties. On the other hand, Colloidal Quantum dots exhibit size dependent band gap properties, and high absorption cross-section over a broad light spectrum.

A novel 0D-2D hybrid materials combining SnS2 layers with CdSe/ZnS Qunatum Dots, can overcome the drawbacks of each materials and significantly improve system’s photon absorption capability along with enhanced charge mobilities. Using single crystal time—resolved spectroscopy, we demonstrated that interactions between SnS2 layers are driven by Energy Transfer, as opposed to charge transfer which is associated with change on ON-times and OFF-times probability distributions.

Self-Assembly and Patterning

Prahlad Kumar Routh

“Optoelectronic characterization and self-assembly based patterning of conducting polymers and its nanocomposites”

Dr. Routh developed a cost-effective method to pattern polymers and its nanocomposites for its use in semi-transparent photovoltaic cells. These patterned thin film structures made out of commercially available conjugated polymers could provide a new route to create cost effective semitransparent films acting as light harvesters in photovoltaic and sensory applications, where usually a large surface to volume ratio is sought. This cost-effective approach incorporated into semi-transparent PVs can meet current market demands of low energy applications. Furthermore, these micro and nano- patterned surfaces given the 150% larger surface area, as compared to flat thin films, have applications in advanced nanostructured materials such as micro-array of lens and photonic crystals; and nanoparticle decorated surfaces for sensors.