Simulations of the Biodegradation of Citrate-Based Polymers for Artificial Scaffolds Using Accelerated Reactive Molecular Dynamics
ReaxFF simulations of poly(1,6-hexanediol-co-citric acid) use CHON-2017_weak with a bond-boost / restrain-energy protocol (after Vashisth et al.) to make room-temperature hydrolysis accessible, benchmark barrier trends against literature QM, and compare chemical plus uniaxial mechanical degradation of polymer bundles in water at 300 K.
Summary¶
The authors study two citrate-based polymer classes—polyester (PE) and polyester–ether (PEE)—that carry both ester and ether linkages, using accelerated ReaxFF MD so that hydrolysis can be forced at 300 K despite nanosecond affordable windows. A restrain (bond-boost) potential adds targeted energy to tagged atom pairs once pre–transition-state geometries are met, with parameters \((F_1, F_2, R_{12})\) scanned in NVT runs to minimize ReaxFF barriers for each hydrolysis channel; those barriers are compared to DFT/MP2 literature values for model ester and ether hydrolysis. Large bundles (10 chains × 20 monomers, ~20 Å bundle diameter, 3000 water molecules) are built in-house, relaxed with NVT then NPT equilibration, and subjected to bond-boosted NVT hydrolysis in boxes listed in the paper (e.g. 33.5 × 33.5 × 150 Å for PE at 0.868 g cm\(^{-3}\)). A second set of restraint parameters illustrates selectivity (favor ester over ether by lowering boost). Longitudinal tensile tests at \(1\times 10^8\) and \(2\times 10^8\ \mathrm{s}^{-1}\) probe modulus on pristine and partially hydrolyzed bundles.
Methods¶
1 — MD application (reactive, polymer + water).
- Engine / integrator: ReaxFF with leapfrog Verlet integration; QEq charges; CHON-2017_weak parameter set for hydrocarbon–water weak interactions; short-range 5 Å cutoff, Coulomb 10 Å (as stated in Computational Details). N/A — the article text does not name a standalone MD package (see SI for implementation notes).
- System size & composition: PE and PEE bundles: 10 chains, 20 monomers each, bundle diameter ~20 Å; 3000 explicit water molecules. Final hydrolysis cells: PE box 33.50 × 33.50 × 150 Å, density 0.868 g cm\(^{-3}\); PEE 31.40 × 31.40 × 245 Å, 0.885 g cm\(^{-3}\) (Table 1).
- Boundaries / periodicity: Three-dimensional periodic simulation cells for bulk bundle models (as standard for the reported box dimensions).
- Ensemble / stages: Energy minimization; NVT 100 ps; NPT ~200 ps to reach ~0.8–1.0 g cm\(^{-3}\); production bond-boosted hydrolysis in NVT at 300 K. For restraint activation, when all distance criteria are met the extra \(E_\mathrm{res}\) is applied for 15,000 steps (3.75 ps at the paper’s step length—see timestep). N/A — NVE is not the production ensemble for the large-bundle hydrolysis stage.
- Timestep: 0.25 fs (inferred: 3.75 ps / 15,000 steps in the restraint subprotocol).
- Duration / stages: 100 ps NVT + 200 ps NPT equilibration; hydrolysis and mechanical runs continue as in Results (multi-segment, N/A for a single “production” length in one line—stated per figure/table in the PDF). N/A — metadynamics / umbrella / replica exchange are not used; enhanced chemistry uses bond-boost / restrain only.
- Thermostat: Berendsen, temperature damping 100 fs, applied in “all MD simulations” per Computational Details.
- Barostat: N/A — NPT is only the ~200 ps equilibration stage; the analysis narrative focuses on NVT hydrolysis. Hydrostatic pressure servicing is implied only in that equilibration window, not in the final NVT bond-boost runs. N/A — anisotropic stress control not detailed in the main text.
- Temperature: 300 K for biodegradation and mechanical tests; N/A for broad multi-\(T\) sweeps in this study (barrier search uses isothermal NVT as described).
- Pressure: N/A in the NVT bond-boost production windows (isochoric NVT for those segments).
- Electric field: N/A — no applied field in the MD protocol.
- Replica / enhanced sampling: N/A for umbrella or metadynamics; bond-boost / restrain energy accelerates specific hydrolysis pathways.
Mechanical testing: uniaxial strain along the longitudinal bundle axis at \(1 \times 10^8\) and \(2 \times 10^8\ \mathrm{s}^{-1}\) on prehydrolyzed and partially hydrolyzed systems.
2 — Force-field / QM validation. N/A — the paper does not re-fit ReaxFF parameters; it adopts CHON-2017_weak and benchmarks ReaxFF ester/ether barrier heights (reported as ~23 and ~49 kcal mol\(^{-1}\) in the discussion) against published DFT/MP2 literature on model hydrolyses.
Findings¶
- Chemical degradation: ReaxFF barrier estimates for model ester and ether hydrolyses are reported to agree semiquantitatively with prior ab initio studies, supporting use of the force field in this polymer + water context. In PEE, RDFs and restraint energies show faster ester scission and favorable selectivity when boost parameters are tuned; lowering parameters can nearly suppress ether hydrolysis while leaving ester reactivity, demonstrating restraint selectivity between functional groups in one polymer.
- Comparisons: QM literature barriers and mechanisms for methyl acetate-like and ether model reactions serve as the DFT/MP2 references; N/A — the paper is not a direct head-to-head new DFT PES study of the full polymer, but a ReaxFF-vs-literature consistency check.
- Sensitivity / levers: restraint \((F_1, F_2, R_{12})\) sets control which bonds react and reaction priority; lower boosts reduce ether paths relative to ester. Strain rate increases the tensile modulus in the two simulated rates, showing rate-dependent mechanical response. PEE exhibits a higher tensile modulus but yields sooner than PE under the reported tests, so PE is described as more ductile than PEE in this comparison.
- Limitations (as in the work): Accelerated MD is not a literal laboratory timescale; enzymatic pathways, pH variation, and device-scale transport are outside the atomistic model. Viscous losses and very long creep of entangled matrices are not resolved.
Limitations¶
Bond-boost trajectories are biased toward reaction; clinical in vivo degradation includes biological and continuum transport physics not in the model.
Relevance to group¶
Penn State / van Duin-group ReaxFF on biodegradable elastomers with a documented restraint / Vashisth-type ReaxFF acceleration workflow tied to QM barrier checks and mechanics.