---
title: "Material screening pipeline"
url: "https://ouro.foundation/posts/mmoderwell/material-screening-pipeline"
id: "019409c2-e7f2-774f-86d3-fb457f9bbdf6"
type: "post"
---

# Material screening pipeline

Post by @mmoderwell in #superconductors (all) · 2024-12-27

- Asset id: `019409c2-e7f2-774f-86d3-fb457f9bbdf6`
- Page: https://ouro.foundation/posts/mmoderwell/material-screening-pipeline

# Material screening pipeline

Superconductivity typically emerges from strong interactions between electrons and vibrations in the crystal lattice (phonons). These interactions can lead to electron pairing, enabling resistance-free current flow. Our goal is to look for materials that show promising signs of these interactions at room temperature.

## The Pipeline Steps

###  1. Ground State Calculation

**What**: Calculate the most stable configuration of the material at 0K
**Why**: This gives us our starting point - like finding the "resting state" of the material
**Details**:

- Use Quantum ESPRESSO to run DFT calculations
- Need high accuracy because later steps build on this
- Takes ~100-1000 CPU hours per material
- Must ensure forces and energies are well converged

###  2. Heating to Room Temperature

 **What**: Carefully bring the material up to 300K
**Why**: Abruptly heating could shock the system and give unrealistic results
**How**: 

- Start from ground state
- Gradually increase temperature over 5 picoseconds
- Use velocity rescaling to control heating
- Monitor system to ensure stable heating

### 3. Equilibration Phase

**What**: Let the system stabilize at 300K
**Why**: Need to ensure the material is behaving normally at room temperature before measuring properties
**Details**:

- Run for 45 picoseconds
- Use Nosé-Hoover thermostat to maintain temperature
- Check energy conservation
- Watch for any structural instabilities
- This is like letting a pot of water settle after bringing it to a boil

###  4. Production Run & Analysis

**What**: Collect data about how atoms move and electrons behave
**Why**: This is where we look for signatures of potential superconductivity
**Technical Details**:

- Run for 30 picoseconds
- Save data every 5 femtoseconds
- Calculate Dynamic Structure Factor (DSF)
- DSF tells us how atoms move collectively
- Look for specific patterns in the DSF that suggest strong electron-phonon coupling

### 5. Pattern Recognition

**What**: Analyze the DSF patterns
**Why**: Certain patterns suggest conditions favorable for superconductivity
**Looking For**:

- Soft phonon modes (vibrations that become very easy to excite)
- Strong peaks at specific wavelengths
- Patterns similar to known superconductors

###  6. Screening Process

**What**: Rank materials based on how promising they look
**Why**: Need to prioritize which materials deserve deeper study
**Method**: 

- Compare DSF patterns to known superconductors
- Look for strong electron-phonon coupling signatures
- Consider chemical similarity to known superconductors
- Assess practical feasibility of synthesis 

## Resource Planning

### Computing Needs:

- Ground State: ~1000 CPU hours
- AIMD Runs: ~10000 CPU hours
- Analysis: ~50 CPU hours
- Storage: ~10GB per material