Equilibrist: A Browser-Based Platform for Fitting Equilibrium and Rate Constants from Optical and NMR Data

Area of Science:

• Chemistry
• Chemical Kinetics
• Chemical Equilibria

Background:

• Quantitative analysis of chemical equilibria and kinetics is crucial in chemistry.
• Existing platforms may lack comprehensive statistical diagnostics for model assessment.

Purpose of the Study:

• To introduce Equilibrist, a free, open-source platform for quantitative analysis of chemical equilibria and kinetics.
• To provide advanced statistical tools for robust model selection and parameter estimation.

Main Methods:

• Utilizes a plain-text scripting language for defining reactions and constraints.
• Incorporates five fitting modes including species concentration, NMR, and spectrophotometry.
• Employs L-BFGS-B and Nelder-Mead optimization methods.

Main Results:

• Implements comprehensive statistical diagnostics: information criteria (AIC, AICc, BIC), Akaike weights, Fisher F-test, and residual analysis (Durbin-Watson, Shapiro-Wilk).
• Offers advanced parameter identifiability mapping using RMSE profiles, eigen decomposition, and sensitivity analysis.
• Provides three uncertainty quantification families: bootstrap, jackknife, and Monte Carlo.

Conclusions:

• Equilibrist enables users to move beyond basic R-squared and Hessian errors for thorough assessment of chemical models and parameter estimates.
• The platform enhances the reliability and interpretability of quantitative chemical analyses.