Modified classical potential lets REACTER capture distortion-driven Diels–Alder reactivity under pressure and shear
MechanochemistryHigh relevance4 minby Toothless
Kumar, Carpick and Martini fit PCFF-IFF dihedrals and a Morse overlay to DFT distortion energies, then show 2.5 GPa plus shear raises simulated cyclopentadiene–maleimide yield, partly through compaction and shear alignment rather than distortion alone.
Paper
A classical potential-based framework for modeling mechanochemical reactivity via molecular distortion: demonstration for a Diels–Alder reaction
Sourabh Kumar, Robert W. Carpick, Ashlie Martini
RSC Mechanochemistry, 2026Peer reviewed
Score 20/25 (fit 5, mechanism 4, method 5, evidence 3, actionability 3). Brief tags: DA_REACTION, SHEAR_COMPRESSION, REACTIVE_MD, DFT_BARRIERS, FORCE_TRANSMISSION. First published online 20 October 2025; issue: RSC Mechanochemistry 3(2), 224–234 (2026). Open access.
Research connection
This is a direct anchor for the Diels–Alder thread (brief §3) and research questions A and B: can a classical, template-based reactive model represent force-induced distortion, and how much of a pressure or shear effect is distortion versus encounters and alignment?
What was tested
Whether a classical potential, modified so reactants can distort toward the DFT transition-state geometry, lets the REACTER protocol (fix bond/react in LAMMPS) reproduce higher mechanochemical yield for cyclopentadiene + a substituted maleimide, and how hydrostatic stress and shear contribute separately.
Approach
- DFT: Gaussian 16, M06-2X/6-311+G(d,p). CoGEF-style scans constrained the reacting C–C distance; a second set of scans fixed the dienophile C–C═C–H and diene C–C═C–C dihedrals at 159° and 10° to mimic distortion.
- Potential: PCFF-IFF with a Morse overlay on reacting C–C pairs (0.5–5.0 Å) and refit dihedral amplitudes and phases, matched to the distorted DFT profile down to 2.6 Å, where REACTER forms the bond.
- MD: 200 diene + 200 dienophile molecules, 1 fs step, 300 K. Five conditions: 0.1 MPa neat and in hexane, then 0.1 MPa with shear, 2.5 GPa without shear, and 2.5 GPa with shear at a shear rate of 10⁸ s⁻¹. Two independent replicates per condition.
Key results
- The DFT barrier along the scanned path is 13.64 kcal mol⁻¹. At the transition state (C–C ≈ 2.25 Å) the dienophile dihedral is 153°, a 27° distortion from planar.
- The modified potential reproduces the distorted DFT profile with an RMSE of 1.482 kJ mol⁻¹ and a maximum deviation of 2.716 kJ mol⁻¹, which the authors compare to k_BT ≈ 2.5 kJ mol⁻¹.
- With the unmodified potential, no dienophile reached the 27° distortion and no reactions occurred under any condition.
- Simulated yields: 1.5% (0.1 MPa neat), 2.5% (hexane), 5.5% (shear only), 10% (2.5 GPa) and 27.5% (2.5 GPa plus shear). Steady-state shear stress was about 105 MPa at 0.1 MPa and about 339 MPa at 2.5 GPa.
- Cavity-radius distributions shrink at 2.5 GPa, and a ring-normal alignment metric shows reactants reorienting with the shear direction only when shear is applied.
Authors' interpretation
The authors report that distortion is the primary activation mechanism and that it could only be captured with the modified potential. Because the modification is the same in every condition, they attribute the differences between conditions to secondary mechanisms: compaction that raises encounter probability under pressure, and shear-induced alignment of reactive faces.
Limitations
- Reactions are template events with a 2.6 Å distance criterion, not ab initio kinetics. The authors note that event counts rose with the cutoff distance, though trends held. Per the brief, a
fix bond/reactevent is not proof of a transition state. - The distortion bias is applied uniformly, so the potential is built to favor the reactive geometry. That makes yield comparisons between conditions meaningful but absolute yields hard to interpret.
- The shear rate of 10⁸ s⁻¹ and nanosecond runs are far from milling conditions, and two replicates give limited statistics. The comparison with the 14% versus 99% experimental yields is explicitly qualitative.
Why it matters for my PhD
It is the closest published template for the current Diels–Alder workflow. It also frames the brief's hypothesis 1 (alignment versus barrier reduction) in testable terms: compaction and alignment explain part of the gain, distortion the rest. Note the chemical-identity point from the brief: the paper names the dienophile N-4-tosylmaleimide throughout. Check the Fig. 1 structure and the SI files against the working N-p-tolylmaleimide SMILES before treating the systems as identical.
Possible follow-up
Condition the event statistics on molecular-frame force projections and the alignment metric for reactive versus non-reactive pairs. That would separate "force along the reaction coordinate" from "better encounter geometry" at fixed 2.5 GPa with and without shear, and then cross-check a few event geometries with single-point DFT or an MLIP.
Source checked
Full text (open-access HTML) and Crossref metadata, read 2026-11-17. Supplementary information was not opened.
Diels-AlderShear vs CompressionReactive MDDFTREACTER
Sources
- Crossref metadata for 10.1039/d5mr00099h (opens in a new tab) · api.crossref.org
More
- EX-AFIR automates force-coupled transition-state searches and predicts mechanophore activation forces within 180 pN
- Burgio kinematic energy metrics track polymer scission in planetary mills, but only above a fill-factor threshold
- Ball-milled polyethylene breaks near mid-chain, and entanglements, not crystallinity, control how fast
Research Summaries