ReaxFF MD Simulations of Peptide-Grafted Gold Nanoparticles
Summary¶
ReaxFF molecular dynamics is used to study ~3 nm gold nanoparticles grafted with short cysteine/glycine-containing peptides in water, varying surface grafting density. The work emphasizes facet-dependent adsorption, charge polarization on Au, electrostatic potential of the functionalized particle, and thiol (de)protonation / proton transfer to water and carboxyl groups. The Langmuir abstract motivates peptide-grafted AuNP systems for biodetection and drug delivery, noting surface-enhanced Raman uses and thiol–gold functionalization as a standard bioconjugation handle; DFT literature on small Au clusters is cited for electronic trends, while this study targets explicit-solvent ReaxFF MD at nanoparticle sizes relevant to colloidal experiments.
Methods¶
Models / engine. ReaxFF molecular dynamics in LAMMPS (as standard for the cited reactive-FF workflow; see Langmuir + SI for the exact version and neighbors). Au cores ~3 nm diameter; peptide ligands CGCG and CGGG at low (~1.46 peptides/nm²) and high (~2.24 peptides/nm²) grafting densities; solvation in cubic ~8.7–8.9 nm-side cells with 3D PBC (N-k-class atom counts as in the paper).
ReaxFF MD. Initial relaxation in vacuum (100 ps at 5 K), then heating to 289.15 K; Berendsen thermostat (damping 100 fs); NVT; time step 0.1 fs; QEq charge updates every 10 MD steps; bond-order and nonbonded cutoffs 0.3 Å and 10 Å (hydrogen-bond and neighbor criteria as stated in the article). Coarse-grained MARTINI simulations are reported for comparison on the same systems.
Analysis. Gyration radius, normalized facet coverage, electrostatic potential from Au charges, and reaction chemistry from trajectories.
Engine / other MD. ReaxFF block above follows the Langmuir protocol; MARTINI coarse-grained runs for comparison use the companion setup in the article. Barostat / NPT: N/A for the NVT stated protocol; uniaxial stress / constant-p segments are N/A unless the SI adds them. External E-field, shear, shock, umbrella sampling: N/A for the stated peptide-Au adsorption study.
Findings¶
Peptide adsorption shows slight facet dependence, with Au(111) favored versus Au(100) and Au(110). Thiol–Au binding plus O/N–Au interactions polarize the surface (net positive on exterior Au, negative core in charge histograms); the outer peptide layer yields overall negative electrostatic potential. Simulations report thiol deprotonation and proton transfer to water and carboxyl groups. MARTINI comparisons show similar gyration radii, supporting that short peptides are relaxed within the ReaxFF runs. Charge histograms in the article illustrate facet-dependent Au oxidation-state-like patterns as peptide coverage changes, helping explain colloidal stability trends tied to zeta-potential sign.
Limitations¶
Peptide lengths and simulation times are finite; the authors note ReaxPQ+ as a future improvement for aqueous polarization accuracy.
Ion strength, pH, and competitive adsorbates in serum-like media are not exhaustively sampled in the Langmuir protocol excerpted here; extend cautiously to in vivo delivery claims without orthogonal experiments.
Peptide sequence isomers beyond CGCG/CGGG may shift facet preferences and charge patterns; repeat sampling when ligand libraries expand.
Relevance to group¶
Application of ReaxFF to biomolecular–Au interfaces in explicit water, relevant to reactive FF validation for nanoparticle–ligand chemistry.
Citations and evidence anchors¶
DOI: 10.1021/acs.langmuir.8b03951