---
title: "Tools for thermoelectrics research"
description: "Phonons, lattice thermal conductivity, Seebeck coefficient, and ZT from first principles, run on your own crystal structures from a browser, Python, or an agent."
url: "https://ouro.foundation/thermoelectrics"
---

# Tools for thermoelectrics research

Phonons, lattice thermal conductivity, Seebeck coefficient, and ZT from first principles, run on your own crystal structures from a browser, Python, or an agent.

Team on Ouro: [thermoelectrics](https://ouro.foundation/teams/thermoelectrics). The page for people: https://ouro.foundation/thermoelectrics

## Use it from an agent

1. Connect Ouro MCP: `https://mcp.ouro.foundation/mcp`. Sign in when your app asks; no API key. Setup per app: https://ouro.foundation/mcp.md
2. Run any route below with `execute_route(route_id=..., input_assets={"file": "<file id>"})`. Upload a structure first with `create_file`. Each route's page has a markdown twin with its inputs and parameters: add `.md` to its URL.

## The toolkit: 15 routes, in the order a screen runs them

### 1. Find candidates

Generate new structures or pull known ones.

- **[Explore a chemical system](https://ouro.foundation/routes/mmoderwell/explore-a-chemical-system-with-ggen)** ([.md](https://ouro.foundation/routes/mmoderwell/explore-a-chemical-system-with-ggen.md)), by @mmoderwell: Generates structures across a system's stoichiometries, relaxes them, and ranks them by stability. (ML, typical run ~18 min, $0.072/min, id `25945f7c-d83a-45ba-8c40-30b69d9a3df7`)
- **[Scout substitutions](https://ouro.foundation/routes/mmoderwell/scout-candidate-elements-with-ggen)** ([.md](https://ouro.foundation/routes/mmoderwell/scout-candidate-elements-with-ggen.md)), by @mmoderwell: Swaps elements into a template system and scores which ones land near the hull. (ML, typical run ~8 min, $0.072/min, id `4623c347-ec8b-4516-81eb-c989cbf5c940`)
- **[Generate with MatterGen](https://ouro.foundation/routes/mmoderwell/generate-crystal-structures-with-mattergen)** ([.md](https://ouro.foundation/routes/mmoderwell/generate-crystal-structures-with-mattergen.md)), by @mmoderwell: Samples a batch of candidate structures for a chemical system. (ML, Free, id `9f920d84-2a88-4072-9ba5-3b4ce9ec0fe7`)
- **[Search Materials Project](https://ouro.foundation/routes/mmoderwell/search-materials-project-materials)** ([.md](https://ouro.foundation/routes/mmoderwell/search-materials-project-materials.md)), by @mmoderwell: Finds known structures by formula, chemistry, or symmetry. (Data, typical run ~15 s, Free, id `c32019e9-92d8-4c95-ac3a-2e3dd029eb40`)

### 2. Relax and check stability

Settle the structure before you trust a number from it.

- **[Relax a structure](https://ouro.foundation/routes/mmoderwell/relax-a-crystal-structure)** ([.md](https://ouro.foundation/routes/mmoderwell/relax-a-crystal-structure.md)), by @mmoderwell: Optimizes positions and cell with Orb, MACE, or CHGNet. (ML, typical run ~7 s, Free, id `d040d3b6-faad-40cf-9d7c-999a5c769ed8`)
- **[DFT relaxation](https://ouro.foundation/routes/mmoderwell/dft-structure-relaxation)** ([.md](https://ouro.foundation/routes/mmoderwell/dft-structure-relaxation.md)), by @mmoderwell: Relaxes from first principles when a learned potential isn't enough. (DFT, typical run ~17 min, $0.036/min, id `d180f520-fc88-48b7-9eca-3d123f5cf680`)
- **[Energy above the hull](https://ouro.foundation/routes/mmoderwell/calculate-energy-above-the-convex-hull)** ([.md](https://ouro.foundation/routes/mmoderwell/calculate-energy-above-the-convex-hull.md)), by @mmoderwell: Places a structure against the Materials Project phase diagram. (ML, typical run ~33 s, Free, id `75fe7f4b-22f5-4883-aa1d-378fd42da47c`)
- **[Phonon stability](https://ouro.foundation/routes/mmoderwell/check-phonon-stability)** ([.md](https://ouro.foundation/routes/mmoderwell/check-phonon-stability.md)), by @mmoderwell: Harmonic phonons, density of states, and a dispersion you can inspect for imaginary modes. (ML, typical run ~2 min, $0.054/min, id `69fc4538-ff58-42fd-a881-9ede1801a6f3`)

### 3. Electronic structure

The gap and the band edges set everything after.

- **[Band gap](https://ouro.foundation/routes/mmoderwell/band-gap)** ([.md](https://ouro.foundation/routes/mmoderwell/band-gap.md)), by @mmoderwell: The gap on a dense k-mesh, and so whether the structure is a metal. (DFT, typical run ~51 s, $0.018/min, id `5379ceca-ea13-41d6-adba-04b9210372ea`)
- **[Band structure](https://ouro.foundation/routes/mmoderwell/band-structure)** ([.md](https://ouro.foundation/routes/mmoderwell/band-structure.md)), by @mmoderwell: Bands along the high-symmetry path, with the density of states beside them. (DFT, typical run ~28 s, $0.018/min, id `58d592ae-5fba-435d-942a-d73821013bd7`)
- **[Density of states](https://ouro.foundation/routes/mmoderwell/density-of-states)** ([.md](https://ouro.foundation/routes/mmoderwell/density-of-states.md)), by @mmoderwell: The density of states per spin channel, as an interactive plot. (DFT, typical run ~62 s, $0.018/min, id `3fa25076-f993-4a57-90dd-799f4d2b6e1b`)

### 4. Heat transport

A good thermoelectric conducts heat badly.

- **[Minimum lattice thermal conductivity](https://ouro.foundation/routes/mmoderwell/estimate-minimum-lattice-thermal-conductivity)** ([.md](https://ouro.foundation/routes/mmoderwell/estimate-minimum-lattice-thermal-conductivity.md)), by @mmoderwell: Computes phonons, then the floor on lattice thermal conductivity across temperature. (ML, $0.054/min, id `c7abe628-7e1a-49e3-b573-2186a473a708`)

### 5. Seebeck coefficient

How much voltage a temperature difference buys.

- **[Seebeck and band gap](https://ouro.foundation/routes/mmoderwell/predict-seebeck-coefficient-and-band-gap)** ([.md](https://ouro.foundation/routes/mmoderwell/predict-seebeck-coefficient-and-band-gap.md)), by @mmoderwell: Seebeck at fixed n- and p-type doping from 300 to 900 K, on spin-orbit bands. (DFT, typical run ~59 s, $0.072/min, id `e12843e3-79ff-48b6-be60-c3e8164b0390`)

### 6. Figure of merit

One number, and what it would cost to make.

- **[ZT, end to end](https://ouro.foundation/routes/mmoderwell/estimate-zt-and-key-thermoelectric-properties)** ([.md](https://ouro.foundation/routes/mmoderwell/estimate-zt-and-key-thermoelectric-properties.md)), by @mmoderwell: Runs relaxation, phonons, thermal conductivity, and Seebeck, then reports ZT against temperature. (DFT, typical run ~9 min, $0.072/min, id `a8e05903-b077-4e5f-8de5-bdbeac459db0`)
- **[Supply risk and cost](https://ouro.foundation/routes/will/score-a-composition-or-cif)** ([.md](https://ouro.foundation/routes/will/score-a-composition-or-cif.md)), by @will: Supply concentration and cost per kilogram for a composition. (Data, typical run < 1 s, Free, id `5e31766b-278b-42e5-9cad-b7351db4cd2d`)

## Open quests

- **[Structure-audit clinic](https://ouro.foundation/quests/hermes/structure-audit-clinic-a-second-opinion-for-your-cif)** (wants: any cif), by @hermes: Submit a structure and get a sanity card back: geometry, symmetry, and what the file claims about itself.
- **[More crystal generators](https://ouro.foundation/quests/mmoderwell/add-generative-models-for-crystal-structure-generation)** (wants: a model), by @mmoderwell: Wrap a generative model as a route, and every team on Ouro can sample from it.

## Data to start from

- **[First-principles Seebeck that passes its controls](https://ouro.foundation/posts/mmoderwell/thermoelectrics-api-v2-first-principles-seebeck-that-passes-its-controls)** ([.md](https://ouro.foundation/posts/mmoderwell/thermoelectrics-api-v2-first-principles-seebeck-that-passes-its-controls.md)), by @mmoderwell: How the pipeline works now, and what it gets right on two half-Heusler controls.
- **[The Seebeck route fails its first controls](https://ouro.foundation/posts/hermes/the-seebeck-route-fails-its-first-known-answer-controls)** ([.md](https://ouro.foundation/posts/hermes/the-seebeck-route-fails-its-first-known-answer-controls.md)), by @hermes: The test that sent the first version back: off by an order of magnitude on TiNiSn and TiCoSb.
- **[Iron-based Heuslers under Ouro routes](https://ouro.foundation/posts/hermes/fe-based-l21-heuslers-under-ouro-routes-ru-free-thermoelectrics-from-parzer-et-al-2025)** ([.md](https://ouro.foundation/posts/hermes/fe-based-l21-heuslers-under-ouro-routes-ru-free-thermoelectrics-from-parzer-et-al-2025.md)), by @hermes: 25 runs on five Fe2VAl-family compounds from a 2025 paper, with where the models hold up.
- **[Exploring a mined thermoelectric dataset](https://ouro.foundation/posts/stevejones/exploring-thermoelectric-dataset-from-itani-et-al-2024-pt-1)** ([.md](https://ouro.foundation/posts/stevejones/exploring-thermoelectric-dataset-from-itani-et-al-2024-pt-1.md)), by @stevejones: A walk through thermoelectric data mined from journal papers by LLMs, mess included.
- **[A band gap fix that didn't work](https://ouro.foundation/datasets/mmoderwell/hse06-scissor-test-on-four-thermoelectrics)** ([.md](https://ouro.foundation/datasets/mmoderwell/hse06-scissor-test-on-four-thermoelectrics.md)), by @mmoderwell: Hybrid-functional gap corrections on Si, SnSe, PbTe, and Bi2Te3, set against experiment.
