CO Oxidation Microkinetic Modeling Framework

From Periodic DFT Energetics to Regime-Dependent Catalytic Performance

Overview

This repository implements a physics-based microkinetic framework for heterogeneous CO oxidation on Pt(111).

The goal is to translate periodic DFT-derived adsorption energies and activation barriers into macroscopic catalytic behavior across temperature and gas-phase conditions.

The modeling pipeline connects microscopic energetics to system-level performance:

DFT Energetics
      ↓
Arrhenius Rate Constants
      ↓
Surface Coverage Dynamics
      ↓
Steady-State Flux
      ↓
Catalytic Performance Maps

The framework shows how surface competition, kinetic coupling, and operating conditions collectively determine catalytic activity.

Reaction Network

The following elementary steps are modeled:

1.	CO(g) + * ⇌ CO*
2.	O₂(g) + 2* ⇌ 2O*
3.	CO* + O* ⇌ CO₂*
4.	CO₂* ⇌ CO₂(g) + *

where * denotes an empty surface site.

Surface site conservation is enforced:

θ_CO + θ_O + θ_CO₂ + θ_* = 1

Rate constants follow Arrhenius form:

k = A · exp(−Ea / (kB T))

Mathematical Framework

The microkinetic model includes:

• Mean-field surface kinetics
• ODE-based surface coverage evolution
• Numerical integration to steady state
• Turnover frequency (TOF) defined as steady-state CO₂ formation rate

Temperature sweeps allow extraction of apparent activation energy from:

ln(TOF) vs 1/T

Key Results

1. Regime-Dependent Catalytic Performance

Catalytic activity varies strongly with CO partial pressure.

Three regimes emerge:

• Oxygen-activated regime (low CO)
• Balanced regime (maximum activity)
• CO-poisoned regime (high CO)

The volcano-like behavior arises from site competition and coverage redistribution, not from a single dominant barrier.

2. Surface Coverage Redistribution

Increasing CO pressure causes:

• Increase in θ_CO
• Decrease in empty sites θ_*
• Suppression of O₂ adsorption

Catalytic performance therefore depends on surface availability, not only intrinsic rate constants.

3. Emergent Apparent Activation Energy

Apparent activation energy does not correspond to a single elementary barrier.

Instead it emerges from:

• flux redistribution
• coverage shifts
• changing kinetic bottlenecks

4. Barrier Sensitivity and Regime Mapping

Barrier perturbation analysis identifies which step controls catalytic flux.

Findings:

• O₂ dissociation dominates in oxygen-rich regimes
• Surface reaction becomes controlling near optimal activity
• CO₂ desorption becomes important under CO-rich conditions

Rate control migrates across state space, demonstrating strong kinetic coupling.

Technical Implementation

The framework is implemented in Python with a modular architecture.

Features include:

• Arrhenius-based rate construction
• Stiff ODE integration (BDF)
• Temperature sweeps
• CO pressure sweeps
• Apparent activation energy extraction
• Barrier perturbation analysis
• Kinetic regime visualization

The structure is designed to integrate naturally with periodic DFT workflows.

Model Assumptions

To maintain interpretability the model assumes:

• Mean-field approximation
• Single site type
• No lateral adsorbate interactions
• No coverage-dependent barriers
• No transport limitations

These simplifications allow clear mechanistic interpretation while enabling systematic extensions.

Reproducibility

Run baseline simulation:

python scripts/run_baseline.py

Parameter sweeps:

python scripts/sweep_pco.py
python scripts/sweep_T.py
python scripts/sweep_drc.py
python scripts/sweep_heatmap.py

Core Insight

Catalytic performance is not determined by the largest intrinsic barrier.

Instead it emerges from:

• which steps control net flux
• how surface coverages redistribute
• how operating conditions reshape kinetic bottlenecks

This framework demonstrates how electronic-structure energetics can be transformed into predictive catalytic behavior across operating regimes.