# MXene-review **Repository Path**: buaaer/mxene-review ## Basic Information - **Project Name**: MXene-review - **Description**: No description available - **Primary Language**: Unknown - **License**: Not specified - **Default Branch**: main - **Homepage**: None - **GVP Project**: No ## Statistics - **Stars**: 0 - **Forks**: 0 - **Created**: 2026-07-18 - **Last Updated**: 2026-07-18 ## Categories & Tags **Categories**: Uncategorized **Tags**: None ## README # MXene first-principles optical-properties knowledge base This repository organizes practical knowledge for first-principles studies of MXene optical properties. It combines a cleaned, compilable article corpus with a structured literature index, a calculation workflow, and reusable templates for recording papers and calculations. The knowledge base is intended to answer four recurring questions: 1. Which MXene composition, surface termination, adsorption site, layer count, or stacking should be modeled? 2. Which electronic-structure and optical-response method is appropriate for the scientific question? 3. Which numerical parameters must be converged for a reproducible two-dimensional calculation? 4. Which conclusions are reported by the literature, and which are recommendations of this project? ## Start here | Goal | Entry point | | --- | --- | | Plan a calculation | [`knowledge/calculation-workflow.md`](knowledge/calculation-workflow.md) | | Compare methods and systems in the current corpus | [`knowledge/literature-map.md`](knowledge/literature-map.md) | | Read knowledge extracted from one paper | [`knowledge/papers/`](knowledge/papers/) | | Query individual scientific claims | [`knowledge/claims.csv`](knowledge/claims.csv) | | Compare conclusions across papers | [`knowledge/topics/`](knowledge/topics/) | | Query the literature metadata | [`knowledge/literature.csv`](knowledge/literature.csv) | | Record a new paper | [`templates/paper-note.md`](templates/paper-note.md) | | Record a new calculation | [`templates/calculation-record.md`](templates/calculation-record.md) | | Read or rebuild a cleaned article | [`references-tex-new/`](references-tex-new/) | The calculation guide is deliberately code-independent. Values quoted from papers are examples, not universal defaults; production settings must be converged for the selected composition, dimensionality, pseudopotential, and target observable. ## Repository layout ```text MXene/ ├── knowledge/ # paper notes, claims, and topic syntheses ├── templates/ # paper and calculation record templates ├── scripts/ # lightweight consistency checks ├── references-tex-new/ # maintained TeX article corpus └── published/doi-index.csv # DOI manifest for main articles ``` ## Article corpus The article packages live under `references-tex-new/`: | Directory | Document type | | --- | --- | | `Bai2016/` | Article | | `Bashir2023/` | Article | | `Berdiyorov2016/` | Article | | `Cardoso2026/` | Article | | `Cardoso2026SI/` | Supplemental material | | `Kandemir2022/` | Article | | `Kandemir2024/` | Article | | `Kandemir2024SI/` | Supplemental material | | `Lashgari2014/` | Article | | `Pei2021/` | Article | Each directory uses the same compact stem for its main files, for example: ```text references-tex-new/Pei2021/ ├── Pei2021.tex # cleaned LaTeX source ├── Pei2021.bib # BibLaTeX data, when the document has references ├── Pei2021.pdf # generated local PDF, ignored by Git ├── Pei2021.md # retained extraction output, when available ├── result.json # retained extraction metadata, when available └── images/ # extracted figure assets ``` The supplemental-material packages do not have separate bibliographies because they do not cite an extracted reference list. Some newly added packages may also lack retained Markdown or JSON extraction artifacts. Generated PDFs are local build outputs and are excluded from version control. Rebuild them from the maintained TeX, BibLaTeX, and image sources when needed. The DOI manifest at `published/doi-index.csv` follows the same schema as the CSP project and covers the main articles. Supplemental-material packages share their parent articles' DOI and are not indexed separately. Its `size_bytes` values record the local compiled PDFs present when the index was refreshed. ## Maintaining the knowledge index Each main article has one row in `knowledge/literature.csv`, one structured note under `knowledge/papers/`, and one or more traceable rows in `knowledge/claims.csv`. Cross-paper conclusions belong under `knowledge/topics/`. None of these layers may silently turn a paper-specific result or parameter into a project recommendation. After adding or changing an entry, run: ```bash python3 scripts/check_knowledge.py ``` The check verifies the source and claim schemas, unique IDs, per-paper notes, evidence fields, source-file paths, and coverage against `published/doi-index.csv`. Search the curated scientific content with: ```bash python3 scripts/query_knowledge.py --topic optical_response --system Ti3C2 ``` ## Building The shared `latexmkrc` selects pdfLaTeX, runs each build inside its article directory, and supports BibLaTeX/Biber where needed. From the project root, build one document with: ```bash latexmk -r references-tex-new/latexmkrc -pdf \ references-tex-new/Pei2021/Pei2021.tex ``` Build every document with: ```bash for directory in references-tex-new/*/; do stem=$(basename "$directory") latexmk -r references-tex-new/latexmkrc -pdf \ "$directory$stem.tex" done ``` Remove intermediate build files while retaining PDFs with: ```bash for directory in references-tex-new/*/; do stem=$(basename "$directory") latexmk -r references-tex-new/latexmkrc -c \ "$directory$stem.tex" done ``` ## Editing conventions - Treat each `.tex` file, its `.bib` file, and `images/` directory as the maintained source for the corresponding PDF. - Keep directory names and primary filenames synchronized using the `AuthorYear` or `AuthorYearSI` stem. - Use BibLaTeX citations rather than extracted numeric reference markers. - Use `mhchem` for chemical formulas where appropriate. - Keep figure and table captions inside their corresponding float environments. - Exclude journal headers, publication metadata, author-address markers, and publisher permission boilerplate from the cleaned article text. - Rebuild the affected PDF and inspect the log after changing a source file. See `AGENTS.md` for the detailed maintenance and validation instructions used by Codex.