Student or educator? Sign up with your school email and get 6 months of Pro free.
CoChat

The DOI matches its Crossref record; 46 of the 46 references it cites were found in a registry or an index; none of the 3 sources its claims cite is checked for support yet.

An automated computational approach to kinetic model discrimination and parameter estimation

Open access

Abstract

) and corresponding time-series concentration data to determine the kinetic information of the chemistry of interest. This is performed with minimal human interaction and several case studies were performed to show the wide scope and applicability of this process development tool. The approach described herein can be employed using experimental data from any source and the code for this methodology is also provided open-source.

Overview

CoChat Summary

## TL;DR

The paper reports an open-source automated workflow for identifying chemical reaction network (CRN) models and estimating kinetic parameters directly from species lists and time-series concentration data. Across three experimental case studies, it recovered the reported reaction models, including zero-order, first-order, non-integer-order, and catalytic behaviors, with small residuals.

## Key findings

  • The method evaluates all mass-balance-feasible reactions and rate-law variants, then ranks models using AICC after local parameter fitting by minimizing SSE.
  • In the SNAr case study, 3320 unique model evaluations were performed; the best model had first-order dependence on both reactants and estimated parameters of k25°C = 0.499 ± 0.006 M−1 min−1 and Ea = 44.19 ± 0.57 kJ mol−1 for the major product, and k25°C = 0.384 ± 0.009 M−1 min−1 and Ea = 36.57 ± 0.88 kJ mol−1 for the minor product.
  • In the amino-acid protection case, 30 unique model evaluations were performed; the best model had first-order dependence on PfBr and zero-order dependence on Al-Me, with k35°C = 1.06 × 10−2 ± 0.01 × 10−2 min−1 and Ea = 62.91 ± 0.23 kJ mol−1.
  • In the maleic acid/methanol case, 5086 unique model evaluations were performed; the best model matched the reported catalytic mechanism with 0.5-order dependence, with kmono 50°C = 3.85 × 10−3 ± 0.01 × 10−3 M−0.5 min−1, Ea = 72.61 ± 0.12 kJ mol−1, and kdi 50°C = 4.66 × 10−4 ± 0.01 × 10−4 M−0.5 min−1, Ea = 69.74 ± 0.10 kJ mol−1.
  • The paper reports fit residuals of less than 0.3%, less than 0.4%, and less than 0.2% for the three case studies, respectively, and says the final case study was automatically identified in about 17 hours of computational time.

## Methods

  • The workflow is a black-box, automated CRN identification and parameter-estimation pipeline: identify all feasible transformations by mass balance using integer linear programming, generate all possible CRNs and rate-law variants, fit each model by gradient-based local minimization of SSE, and select the best model by corrected Akaike information criterion (AICC).
  • The paper applies the method to three batch experimental case studies using time-series concentration data collected by NMR and HPLC across specified temperature ranges: SNAr chemistry (−25 °C to 50 °C), amino-acid protection (30 °C to 40 °C), and maleic-acid methylation (40 °C to 60 °C).
  • The text says the user inputs the species present and experimental data; in the examples, catalytic species and orders of 0, 0.5, and 1 were considered.

## Limitations

  • The paper states that only species orders of 0, 0.5, and 1 were considered; further non-integer orders were not included because they were rare and would add significant computational burden.
  • The text makes evident that residuals were not normally distributed in at least one case, with possible causes including NMR integration error, solvent evaporation, viscosity changes, and acid/base effects; this is an observed limitation of fit quality and data quality.
  • The paper also notes that expert chemists are still needed to critically appraise the ranked models and may need to alter the identified CRN or prescribe further experiments; this is stated in the paper and indicates the method is not fully self-sufficient.

Model-written summary. Check it against the paper before relying on it.

Integrity signals

Registry

Crossref · 10.1039/d1re00098e

Cites

46 of 46 references verified0 citation errors · 0 not checkable

Statistics

Nothing to recompute.

Figures

Fig. 1 An overview of the approach, consisting of two main stages: model generation and kinetic fitting. Every feasible reaction model based on mass balance is assembled and subsequently evaluated, optimising the kinetic parameters to maximise convergence to the experimental data. Each model is also assigned an AICC evaluation, from which the most likely CRN is obtained.
Fig. 1 An overview of the approach, consisting of two main stages: model generation and kinetic fitting. Every feasible reaction model based on mass balance is assembled and subsequently evaluated, optimising the kinetic parameters to maximise convergence to the experimental data. Each model is also assigned an AICC evaluation, from which the most likely CRN is obtained.CC BY-NC · PubMed Central
Scheme 1 The presupposed reaction model of the SNAr case study.
Scheme 1 The presupposed reaction model of the SNAr case study.CC BY-NC · PubMed Central
Fig. 2 Results from the four kinetic experiments (at −25 °C, 0 °C, 25 °C and 50 °C) in the SNAr case study in CD3OD, where: = 2,4,6-trichloropyrimidine, = major product, = minor product, = 2,4,6-trichloropyrimidine (ODE), = major product (ODE), = minor product (ODE).
Fig. 2 Results from the four kinetic experiments (at −25 °C, 0 °C, 25 °C and 50 °C) in the SNAr case study in CD3OD, where: = 2,4,6-trichloropyrimidine, = major product, = minor product, = 2,4,6-trichloropyrimidine (ODE), = major product (ODE), = minor product (ODE).CC BY-NC · PubMed Central
Scheme 2 The presupposed reaction model of the PfBr protection case study.
Scheme 2 The presupposed reaction model of the PfBr protection case study.CC BY-NC · PubMed Central
Fig. 3 Results from the kinetic experiments in the PfBr protection case study in 50:50 acetonitrile/dichloromethane, where red plots indicate PfBr concentrations and blue plots indicate Pf-Al-Me concentrations. At 30 °C: = experimental data, = ODE. At 35 °C: = experimental data, = ODE. At 40 °C: = experimental data, = ODE.
Fig. 3 Results from the kinetic experiments in the PfBr protection case study in 50:50 acetonitrile/dichloromethane, where red plots indicate PfBr concentrations and blue plots indicate Pf-Al-Me concentrations. At 30 °C: = experimental data, = ODE. At 35 °C: = experimental data, = ODE. At 40 °C: = experimental data, = ODE.CC BY-NC · PubMed Central
Scheme 3 The presupposed reaction model of the maleic acid case study.
Scheme 3 The presupposed reaction model of the maleic acid case study.CC BY-NC · PubMed Central
Fig. 4 Kinetic profiles for two kinetic experiments at 50 °C in methanol, with the initial concentration of maleic acid at 0.4 M and 0.8 M. At 0.4 M: = maleic acid, = maleic acid (ODE), = mono-product, = mono-product (ODE), = di-product, = di-product (ODE). At 0.8 M: = maleic acid, = maleic acid (ODE), = mono-product, = mono-product (ODE), = di-product, = di-product (ODE).
Fig. 4 Kinetic profiles for two kinetic experiments at 50 °C in methanol, with the initial concentration of maleic acid at 0.4 M and 0.8 M. At 0.4 M: = maleic acid, = maleic acid (ODE), = mono-product, = mono-product (ODE), = di-product, = di-product (ODE). At 0.8 M: = maleic acid, = maleic acid (ODE), = mono-product, = mono-product (ODE), = di-product, = di-product (ODE).CC BY-NC · PubMed Central

Verification

Claims the paper makes

0 Valid · 0 Not valid · 3 Not checked · 30 Own results and interpretations

13 more claims

Create a free account to see all 33 claims and every source verdict.

See all claims — free account
  • It may state two separate claims as one. 11
  • The quote only partly states this claim. 5
  • Probably not a substantial claim. 1

Create a free account to see these claims.

Citations

Cites

Research works printed but not linked: 4

  1. C. Georgakis and R.Lin , Stoichiometric Modeling of Complex Pharmaceutical Reactions, in Proceedings of the Annual AIChE meeting in Cincinnati, OH , November, 2005
  2. Searson, 2007, Chem. Prod. Process Model., 2, 10
  3. Taylor, 2020, Chem. Eng. J., 127017
  4. Lood, 2015, Eur. J. Org. Chem., 3793, 3805

Cited by

See all 37

Metadata

Authors
Taylor, C. J., Seki, H., Dannheim, F. M., Willis, M. J., Clemens, G., Taylor, B. A., Chamberlain, T. W., & Bourne, R. A.
Affiliations
  • University of Leeds (GB)
  • University of Cambridge (GB)
  • University of Newcastle Australia (AU)
  • AstraZeneca (United Kingdom) (GB)
Corresponding author
  • Chamberlain, Thomas W.
  • Bourne, Richard A.
Venue
Reaction Chemistry & Engineering
Published
2021
Volume
6
Issue
8
Pages
1404-1411
ISSN
2058-9883 (ISSN-L)
OpenAlex
W3161872738
Cited by
37
References
44
Citations by year
2026: 6 · 2025: 5 · 2024: 10 · 2023: 11 · 2022: 4 · 2021: 1
FWCI
2.89
Type
journal article
Language
en
Open access
Locations
  • Reaction Chemistry & Engineering (publishedVersion, cc-by-nc)
  • PubMed Central (publishedVersion, cc-by-nc)
  • PubMed (publishedVersion)
  • White Rose Research Online (University of Leeds, The University of Sheffield, University of York) (submittedVersion)
  • Newcastle University ePrints (Newcastle Univesity) (cc-by-nc)
Funders
  • Engineering and Physical Sciences Research Council (EP/R032807/1, EP/N509681/1)
  • University of Leeds
  • AstraZeneca (EP/N509681/1)
  • Biotechnology and Biological Sciences Research Council
  • Royal Academy of Engineering (RCSRF1920\9\38)
Topics
  • Physical Sciences › Engineering › Control and Systems Engineering › Control Systems and Identification
  • Social Sciences › Decision Sciences › Management Science and Operations Research › Optimal Experimental Design Methods
  • Physical Sciences › Materials Science › Materials Chemistry › Thermal and Kinetic Analysis
Keywords
kinetic model discrimination, parameter estimation, reaction kinetics, automated analysis, open source software

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

A language model drew the claims from the paper on 2026-10-10, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).

A second model labelled each claim on 2026-10-10 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review

Data sources

Bibliographic data: Semantic Scholar, OpenAlex, Crossref.

Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).

Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.