TDDFT-ris

Official Resources

• Homepage: https://github.com/John-zzh/pyscf_TDDFT_ris
• Source Repository: https://github.com/John-zzh/pyscf_TDDFT_ris
• Documentation: README with theory and usage
• License: Open Source

Overview

TDDFT-ris is a high-performance Python implementation of a semiempirical Linear-Response TDDFT method that achieves ~300x speedup over traditional ab initio TDDFT while maintaining accuracy within 0.06 eV for excitation energies of organic molecules. The method uses Resolution-of-the-Identity (RI) approximation with minimal auxiliary basis (single s-type orbital per atom) and disables the XC kernel, providing an excellent balance of speed and accuracy for UV-Vis spectroscopy.

Scientific domain: Molecular excited states, UV-Vis absorption, large organic molecules
Target user community: Researchers needing fast, accurate TDDFT for large molecular systems

Theoretical Methods

• Linear-Response Time-Dependent DFT (LR-TDDFT)
• Resolution-of-the-Identity (RI) approximation
• Minimal auxiliary basis (ris/risp/rispd)
• Semi-empirical atomic radii parameterization
• Hybrid XC functional compatibility
• MOKIT integral interface
• PySCF backend

Capabilities

• Extremely fast TDDFT calculations (~300x speedup)
• Excitation energy calculations
• UV-Vis absorption spectra
• Excited state analysis
• Multiple auxiliary basis levels (s, sp, spd)
• Hybrid functional support
• Large molecule capability
• Command-line interface
• PySCF object interface

Key Strengths

Exceptional Performance:

• ~300x faster than ab initio TDDFT
• 0.06 eV average deviation
• 4x more accurate than sTDDFT
• Scales to large molecules

Flexible Auxiliary Basis:

• ris: single s-function per atom (fastest)
• risp: s+p functions (default, more accurate)
• rispd: s+p+d functions (most accurate)
• Atomic radii-based exponents

Production Integration:

• Built into TURBOMOLE 7.7dev
• Built into AMESP v2.1dev
• ORCA 6.0 compound script support
• PySCF native integration

Easy Workflow:

• .fch file input support
• Command-line interface
• Python API
• Automatic spectrum plotting

Inputs & Outputs

• Input formats:

  • Formatted checkpoint files (.fch)
  • PySCF mean-field objects
  • MOKIT-compatible files

• Output data types:

  • Excitation energies
  • Oscillator strengths
  • UV-Vis spectra (plotted)
  • Excited state analysis
  • Natural transition orbitals

Interfaces & Ecosystem

• Dependencies:

  • PySCF (required)
  • MOKIT (required)
  • NumPy, SciPy

• Compatible software:

  • TURBOMOLE
  • AMESP
  • ORCA 6.0
  • Gaussian (via .fch files)

Command-Line Usage

# Basic calculation
python -m TDDFT_ris input.fch --nstates 10

# Plot spectrum
python -m TDDFT_ris input.fch --nstates 20 --plot

Performance Characteristics

• Speed: ~300x faster than ab initio TDDFT
• Accuracy: 0.06 eV average deviation
• System size: Large organic molecules
• Memory: Reduced compared to full TDDFT
• Basis sets: All PySCF-supported basis sets

Limitations & Known Constraints

• Accuracy: Slightly lower than full ab initio TDDFT
• Core excitations: Not optimized for core-level
• Charge-transfer: May need careful validation
• Metals: Parameterized for organic molecules
• Dependencies: Requires MOKIT installation

Comparison with Other Codes

• vs ab initio TDDFT: 300x faster, 0.06 eV deviation
• vs sTDDFT (Grimme): 4x more accurate (0.06 vs 0.24 eV)
• vs sTDA: Similar speed, better accuracy
• Unique strength: Best speed/accuracy trade-off for organics

Application Areas

UV-Vis Spectroscopy:

• Absorption spectra of large molecules
• Chromophore design
• Photophysical properties
• Dye molecules

High-Throughput Screening:

• Virtual screening of chromophores
• Materials discovery
• Large dataset calculations

Photochemistry:

• Initial excited state characterization
• Photosensitizer evaluation
• OLED material screening

Best Practices

• Use risp (default) for balance of speed/accuracy
• Validate against full TDDFT for new molecule classes
• Use .fch files for Gaussian interoperability
• Enable spectrum plotting for visualization

Community and Support

• Open-source on GitHub
• 2 contributors
• 1 release (v2.0)
• Active development
• Python 100%

Verification & Sources

Primary sources:

1. GitHub repository: https://github.com/John-zzh/pyscf_TDDFT_ris
2. Built into TURBOMOLE 7.7dev
3. Built into AMESP v2.1dev
4. ORCA 6.0 compound script support

Confidence: VERIFIED - Production software integration confirms validity

Verification status: ✅ VERIFIED

• Source code: OPEN (GitHub)
• Documentation: Comprehensive README
• Production use: TURBOMOLE, AMESP, ORCA
• Language: Python 100%