Quick Start Guide

This guide will help you get started with EMOS quickly through practical examples.

Basic Workflow

Step 1: Choose Your Research Goal

First, determine what you want to accomplish:

  • Discover new materials: Use Materials Exploration features

  • Optimize electronic devices: Use Electronics Application features

  • Analyze existing materials: Use characterization and analysis features

Step 2: Select a Feature

Navigate to the EMOS application and choose from available features:

Materials Exploration Examples:

  • Material Search (Feature ID: 1)

  • Material Generation (Feature ID: 2)

  • DFT Calculation (Feature ID: 5)

Electronics Application Examples:

  • Interface Calculation (Feature ID: 10)

  • Band Structure (Feature ID: 12)

  • Thermal Management (Feature ID: 13)

Step 3: Configure Information Units

Select which databases, generators, and predictors to use:

Common Configurations:

  • Database-heavy: ICSD + Materials Project + JARVIS

  • Generation-focused: MatterGen + iMatGen + M3GNet predictor

  • Prediction-intensive: M3GNet + MatterSim + DeepMD

Step 4: Provide Inputs

Enter the required parameters. Each feature has specific input requirements.

Example Workflows

Example 1: Finding High-Conductivity Materials

Objective: Find materials with high electrical conductivity

Steps:

  1. Select “Material Search” feature (ID: 1)

  2. Configure Information Units:

    • Databases: Materials Project, JARVIS

    • Predictors: M3GNet, MatterSim

  3. Input Parameters:

    • Search criteria: “high conductivity”

    • Property range: conductivity > 10^6 S/m

    • Crystal systems: cubic, hexagonal

  4. Execute and review results

Expected Output:

  • List of candidate materials

  • Predicted conductivity values

  • Crystal structure information

  • Database references

Example 2: Interface Analysis

Objective: Analyze Al/Si interface properties

Steps:

  1. Select “Interface Calculation” feature (ID: 10)

  2. Configure Information Units:

    • Databases: ICSD (for structure data)

    • Predictors: M3GNet, DeepMD

  3. Input Parameters:

    • Interface Type: metal-semiconductor

    • Material A: Al (Aluminum)

    • Material B: Si (Silicon)

    • Calculation Method: DFT

  4. Execute and analyze interface properties

Expected Output:

  • Interface energy: ~1.247 J/m²

  • Band offset: ~1.85 eV

  • Lattice mismatch: ~2.3%

  • Interface states density

  • Charge transfer information

Example 3: New Material Generation

Objective: Generate novel semiconductor materials

Steps:

  1. Select “Material Generation” feature (ID: 2)

  2. Configure Information Units:

    • Databases: Materials Project (for training data)

    • Generators: MatterGen, GNoME

    • Predictors: M3GNet (for property validation)

  3. Input Parameters:

    • Target properties: semiconductor, band gap 1-3 eV

    • Elements: Si, Ge, C, N

    • Crystal system: any

    • Number of candidates: 50

  4. Execute and evaluate generated materials

Expected Output:

  • 50 novel material candidates

  • Predicted band gaps

  • Stability assessments

  • Synthesizability scores

Common Input Parameters

Universal Parameters

All features accept these common parameters:

  • active_databases: List of database configurations

  • active_generators: List of generator configurations

  • active_predictors: List of predictor configurations

Feature-Specific Parameters

Material Search:

  • searchTerm: What to search for

  • propertyRange: Target property values

  • elementList: Allowed elements

Interface Calculation:

  • interfaceType: Type of interface (metal-semiconductor, etc.)

  • materialA: First material

  • materialB: Second material

  • calculationMethod: Computational method

Material Generation:

  • targetProperties: Desired properties

  • elementConstraints: Allowed elements

  • numCandidates: Number of materials to generate

Understanding Results

Result Structure

All features return structured results with:

  • Primary Results: Main computational outputs

  • Metadata: Processing information and parameters

  • Information Unit Results: Individual outputs from databases, generators, predictors

  • Status Information: Success indicators and warnings

Interpreting Outputs

Database Results:

  • Material identifiers and references

  • Crystal structure data

  • Experimental properties

Generator Results:

  • New material structures

  • Generation confidence scores

  • Novelty assessments

Predictor Results:

  • Property predictions with uncertainty

  • Confidence intervals

  • Model performance metrics

Best Practices

Information Unit Selection

  1. Start Simple: Begin with 1-2 information units per type

  2. Match Purpose: Choose units that align with your research goals

  3. Consider Speed: More units = longer processing time

  4. Validate Results: Use multiple predictors for cross-validation

Input Guidelines

  1. Be Specific: Provide detailed, specific inputs for better results

  2. Reasonable Ranges: Use realistic property ranges and constraints

  3. Check Units: Ensure all units are consistent (eV, Å, etc.)

  4. Start Small: Begin with small parameter sets and expand

Result Analysis

  1. Check Convergence: Ensure calculations converged properly

  2. Cross-Validate: Compare results from different information units

  3. Consider Uncertainty: Pay attention to prediction confidence

  4. Document Process: Keep track of parameters and configurations used

Troubleshooting

Common Issues

No Results Found:

  • Broaden search criteria

  • Try different databases

  • Check input parameter validity

Slow Processing:

  • Reduce number of information units

  • Limit search scope

  • Use faster prediction models

Inconsistent Results:

  • Check information unit compatibility

  • Verify input parameter units

  • Consider model limitations

Error Messages:

  • Review input parameters for completeness

  • Check information unit availability

  • Consult the troubleshooting reference