# Xiangyu Yin

> Personal website of Xiangyu Yin, Assistant Computational Scientist at Argonne National Laboratory, working at the intersection of X-ray science, artificial intelligence, and scientific automation. The site is static HTML — the full text of every page is served at its URL.

Author: Xiangyu Yin — Google Scholar: https://scholar.google.com/citations?user=YqRWi_kAAAAJ, GitHub: https://github.com/xyin-anl, LinkedIn: https://www.linkedin.com/in/xiangyu-yin/, X: https://x.com/shawn_xyin. Also runs https://www.nodeology.ai (interactive courses, simulators, and resource boards for AI and X-ray science).

## Publications

- [Publications](https://xiangyu-yin.com/publications.html): journal articles, preprints, and conference proceedings, each with DOI link and BibTeX

## Blog posts

- [Building an MCP Server for the Materials Project](https://xiangyu-yin.com/content/post_mp_mcp.html): how to build a Model Context Protocol (MCP) server that lets Claude and other AI assistants query the Materials Project materials database
- [Building an AI-Driven Scientific Workflow & Chatbot](https://xiangyu-yin.com/content/post_nodeology_example.html): an interactive particle trajectory simulation workflow and chatbot built with Nodeology using prompt templates and plain Python functions
- ["Deep Research" for Scientific Literature Review](https://xiangyu-yin.com/content/post_deep_research.html): can OpenAI and Gemini Deep Research produce rigorous scientific literature reviews? A head-to-head test on generative models for crystal structures
- [Identifying Local Doping Patterns Around Atomic Sites](https://xiangyu-yin.com/content/post_doping_pattern.html): a Python workflow that identifies neighbors around a target site and enumerates symmetry-distinct local doping patterns in crystal structures
- ["Wave Tracker" for Pressure Swing Adsorption](https://xiangyu-yin.com/content/post_wave_tracking.html): a wave-tracking implementation of equilibrium theory that analyzes pressure swing adsorption cycles without expensive PDE-based simulations
- [Designing Materials with Cluster Expansion and MatOpt](https://xiangyu-yin.com/content/post_cluster_expansion.html): train a cluster expansion on DFT data with icet, then use MatOpt to formulate a MILP that finds the optimal atomic arrangement for target properties

## About

- [Home](https://xiangyu-yin.com/): bio and news
- [Resume](https://xiangyu-yin.com/resume/): experience and education, as a web page
- RSS feed: https://xiangyu-yin.com/blog.xml