Modified Random Sequential Adsorption Model for Understanding Kinetics of Proteins Adsorption at a Liquid–Solid Interface

Abstract

Microfabricated 83 MHz quartz crystal microbalance tracks HSA adsorption on hydrophobic hexadecanethiol–Au; an interface-depletion RSA model fits kinetics, supported by ReaxFF MD indicating oriented adsorption and slowed interfacial diffusion.

Summary¶

Experimentally, human serum albumin (HSA) adsorption on a hydrophobic hexadecanethiolated gold surface is measured in real time with 83 MHz micromachined quartz crystal resonators, focusing on asymptotic response and jamming limits under single- and multi-injection protocols. A new interface-depletion modified RSA model captures exponentially depleting interfacial concentration. Complementary ReaxFF reactive MD uses a merged Au/S/O/C/H parameter set (Jarvi et al., Keith et al., Joshi et al., Monti et al., as cited in section 5.2) on a three-layer Au(111)–hexadecanethiolate–aspartame/water surrogate for the interfacial environment: 576 Au, 36 hexadecanethiolates, 18 aspartame molecules, 60 inner-layer waters, plus either 750 waters or 400 waters with 30 dipeptide molecules in the top “bulk” region, in a 34.62 × 29.97 × 140.00 Å³ cell. ADF drives NVT trajectories at 298.15 K with a Berendsen thermostat (100 fs damping) and Δt = 0.25 fs, staging 100 ps surface-only and solution-only equilibrations before a 400 ps combined run (last 200 ps analyzed). Atomic density profiles and MSD analyses support hydrophobic insertion into the thiol layer and slowed interfacial transport, rationalizing the depletion RSA picture relative to bulk-limited kinetics.

Methods¶

Force-field training / fitting. The study does not report a new ReaxFF refit. Reactive simulations use a merged ReaxFF for Au/S/O/C/H assembled from published Au–S–H, Au–metal / Au–O–H, and protein (Monti et al.) parameter sources as summarized in section 5.2 of the Langmuir article.

MD application (atomistic dynamics). Engine / code: ADF (as stated in section 5.2) carries the ReaxFF MD (N/A — LAMMPS not claimed in the main-text protocol excerpt used here). System size & composition: Four-layer Au(111) with 36 C₁₆H₃₃S– thiolates (1 thiol per 4 Au, 576 Au total), 18 aspartame (C₁₂H₁₆N₂O₃) molecules plus 60 waters in the intermediate region, and a top region of either 750 waters (“pure water” case) or 400 waters + 30 dipeptide molecules (“dipeptide solution” case), in a 34.62 × 29.97 × 140.00 Å³ orthorhombic supercell. Boundaries / periodicity: Three-dimensional periodic supercell for the slab–solution stack (standard for this geometry class). Stages / duration: 100 ps pre-equilibration of the Au–thiol–peptide stack, 100 ps of the top solution layer alone, then 400 ps of the combined system with analysis on the final 200 ps. Ensemble: NVT at 298.15 K. Timestep: 0.25 fs. Thermostat: Berendsen with 100 fs damping constant (as reported). Barostat / pressure control: N/A — NPT is not used for these production trajectories. Target temperature: 298.15 K. Hydrostatic pressure / stress control: N/A — not applied beyond the fixed-cell NVT setup described. Electric field: N/A — not used. Replica / enhanced sampling: N/A — direct MD only.

Static QM / DFT. N/A — DFT is not the engine for the ReaxFF trajectories summarized above (quantum references in the article support parameter provenance, not on-the-fly ab initio MD here).

Review / non-simulation framing. 83 MHz micromachined QCM measurements of HSA on hexadecanethiolated Au, plus the interface-depletion RSA kinetic model, anchor the experimental narrative alongside the reactive MD surrogate. Corpus note: the tracked pdf_path is a proof PDF; pagination and any publisher tweaks should be checked against the VOR sibling [[2017hwall-min-langmuir-201-modified-random]] when available.

Findings¶

Outcomes & mechanisms. QCM data show both asymptotic sensor response and jamming-limited coverage for single- vs multi-injection protocols at matched final bulk concentration. The interface-depletion RSA fit captures exponentially depleted interfacial protein availability. ReaxFF/ADF trajectories on the Au–thiol–aspartame–water stack show hydrophobic vs hydrophilic partitioning (atomic density profiles) and layer-resolved MSDs consistent with reduced effective diffusivity near the interface—supporting the idea that slow interfacial transport couples to surface adsorption rate to produce the depleted layer invoked in the kinetic model.

Comparisons. The modified RSA treatment improves agreement with measured kinetics versus classical RSA that assumes a well-mixed interfacial reservoir.

Sensitivity & design levers. Varying injection protocol at fixed final bulk concentration changes how interfacial depletion develops, which the model ties to transport–adsorption coupling rather than bulk concentration alone.

Limitations & outlook (as authored). The atomistic layer uses aspartame and dipeptide-solution surrogates rather than full HSA; QCM frequency interpretation may require viscoelastic corrections beyond a single kinetic fit (see article discussion).

Corpus / KB honesty. Numbers above are taken from the indexed proof PDF text in section 5.2; confirm against the VOR PDF/SI if any layout or typographic ambiguities remain after corpus swaps.

Limitations¶

Proof PDF; atomistic models simplify full protein complexity—interpret as mechanistic support rather than quantitative fit to every experimental concentration. QCM frequency shifts can also couple to viscoelastic film properties not fully resolved by a single RSA-like kinetic fit without rheological calibration. ReaxFF snapshots of peptide segments should be interpreted as qualitative orientation trends because full-length HSA conformations can reorganize on long experimental timescales. Langmuir SI sections may contain additional simulation cells and analysis scripts not summarized on this page.

Relevance to group¶

Adri C. T. van Duin is a co-author; ReaxFF links surface chemistry to measurable kinetics.

Citations and evidence anchors¶

DOI: 10.1021/acs.langmuir.7b00523

Corpus PDF is a proof (*_proof.pdf).