CO₂ Solvent Materials Informatics Collection — Interactive Explorer
An interactive web application for browsing 3,200+ organic solvents with 3D molecular visualization
and comprehensive physical, thermodynamic, electronic, and CO₂ capture properties.
Open index.html via a local HTTP server:
cd COSMIC
python -m http.server 8080
# Then visit http://localhost:8080Note: A local server is required because the app loads CSV and XYZ data files via
fetch().
- 3D Molecular Viewer — Interactive ball-and-stick models rendered with Three.js. Rotate, zoom, and fullscreen support. Bond connectivity is computed in real-time from DFT-optimized XYZ geometries using covalent radii.
- 3,200+ Solvents — Searchable and sortable list with pagination. Filter by solvent ID or SMILES.
- Multi-Source Data Integration — Properties merged from three independent computational pipelines:
- COSMO-RS thermodynamic predictions (24 properties)
- DFT quantum-mechanical calculations (13 properties)
- CO₂ solubility predictions
- Materials Project–Inspired UI — Three-column detail view with left navigation, central content sections, and a right-hand property summary sidebar.
| File | Records | Description |
|---|---|---|
COSMO-RS/pure_solvent_properties_data.csv |
3,203 | Viscosity, density, boiling/melting/flash point, dielectric constant, vapor pressure, critical constants, enthalpies, Gibbs energy, vdW area/volume, solubility parameter, synthetic accessibility |
Quantum_mechanics_DFT/DFT_Computed_Properties.csv |
3,019 | SCF energy, dipole moment, HOMO/LUMO/gap, polarizability, partial charges, vibrational frequencies |
COSMO-RS/solubility_co2.csv |
3,015 | Predicted CO₂ solubility (mol/L) from COSMO-RS |
xyz_files/Optimized_Geometry_xyz/ |
~2,600 | DFT-optimized molecular geometries in XYZ format |
COSMIC/
├── index.html # Web application (single-file)
├── image.png # Logo
├── COSMO-RS/
│ ├── pure_solvent_properties_data.csv # Physical & thermodynamic properties
│ └── solubility_co2.csv # CO₂ solubility data
├── Quantum_mechanics_DFT/
│ └── DFT_Computed_Properties.csv # Electronic structure properties
└── xyz_files/
└── Optimized_Geometry_xyz/ # ~2,600 XYZ geometry files
| Section | Source | Properties Shown |
|---|---|---|
| Summary | All | Key properties overview (boiling point, density, viscosity, HOMO-LUMO gap, CO₂ solubility) |
| Physical Properties | COSMO-RS | Viscosity, density, boiling/melting/flash point, dielectric constant, vapor pressure, molar volume, solubility parameter |
| Thermodynamics | COSMO-RS | Ideal gas entropy, enthalpies of formation/combustion/fusion/sublimation, Gibbs energy |
| Critical Constants | COSMO-RS | Critical temperature, pressure, volume; triple point temperature |
| Molecular Descriptors | COSMO-RS | van der Waals area & volume, parachor, synthetic accessibility |
| Electronic Structure | DFT | SCF energy, HOMO, LUMO, gap, dipole moment, polarizability, partial charges, vibrational frequencies |
| Bonding | XYZ | Bond types, average distances, and counts (computed from geometry) |
| CO₂ Solubility | COSMO-RS | Solubility value and capture rating (High / Moderate / Low) |
- Three.js (v0.160.0) — WebGL-based 3D molecular rendering
- OrbitControls — Camera interaction (rotate, zoom, pan)
- Vanilla JavaScript — No build tools or frameworks required
- Single HTML file — Zero dependencies to install
XYZ files contain only atomic coordinates. Bonds are determined at runtime using covalent radii:
For each atom pair (i, j):
d = Euclidean distance between atoms i and j
threshold = (covalent_radius_i + covalent_radius_j) × 1.3
minimum = (covalent_radius_i + covalent_radius_j) × 0.4
if minimum ≤ d ≤ threshold → bond exists
This is a standard approach in computational chemistry visualization tools.
This repository is released under the MIT License.
The COSMIC database was developed as part of research on organic solvents for electrochemical CO₂ reduction (CO₂ER). Data spans quantum chemical (DFT), thermodynamic (COSMO-RS), and classical molecular dynamics (MD) calculations covering solvents from 11 chemical classes.
Ruiqi Luo GitHub: @RuiqiLuo
