DeePMD-kit v2: A software package for deep potential models.

DeePMD-kit is a powerful open-source software package that facilitates molecular dynamics simulations using machine learning potentials known as Deep Potential (DP) models. This package, which was released in 2017, has been widely used in the fields of physics, chemistry, biology, and material science for studying atomistic systems. The current version of DeePMD-kit offers numerous advanced features, such as DeepPot-SE, attention-based and hybrid descriptors, the ability to fit tensile properties, type embedding, model deviation, DP-range correction, DP long range, graphics processing unit support for customized operators, model compression, non-von Neumann molecular dynamics, and improved usability, including documentation, compiled binary packages, graphical user interfaces, and application programming interfaces. This article presents an overview of the current major version of the DeePMD-kit package, highlighting its features and technical details. Additionally, this article presents a comprehensive procedure for conducting molecular dynamics as a representative application, benchmarks the accuracy and efficiency of different models, and discusses ongoing developments.

Toward design of magnetic nanoparticle clusters stabilized by biocompatible diblock copolymers for T2-weighted MRI contrast.

We report the fabrication of magnetic particles comprised of clusters of iron oxide nanoparticles, 7.4 nm mean diameter, stabilized by a biocompatible, amphiphilic diblock copolymer, poly(ethylene oxide-b-D,L-lactide). Particles with quantitative incorporation of up to 40 wt % iron oxide and hydrodynamic sizes in the range of 80-170 nm were prepared. The particles consist of hydrophobically modified iron oxide nanoparticles within the core-forming polylactide block with the poly(ethylene oxide) forming a corona to afford aqueous dispersibility. The transverse relaxivities (r2) increased with average particle size and exceeded 200 s(-1) mM Fe(-1) at 1.4 T and 37 °C for iron oxide loadings above 30 wt %. These experimental relaxivities typically agreed to within 15% with the values predicted using analytical models of transverse relaxivity and cluster (particle core) size distributions derived from cryo-TEM measurements. Our results show that the theoretical models can be used for the rational design of biocompatible MRI contrast agents with tailored compositions and size distributions.

Magnetic Nanoclusters with Hydrophilic Spacing for Dual Drug Delivery and Sensitive Magnetic Resonance Imaging.

Magnetic Block Ionomer Clusters (MBIClusters) with hydrophilic ionic cores and nonionic coronas have been prepared that have ultrahigh transverse NMR relaxivities together with capacities for incorporating high concentrations of polar antibiotic payloads. Magnetite-polymer nanoparticles were assembled by adsorbing the polyacrylate block of an aminofunctional poly(ethylene oxide-b-acrylate) (H2N-PEO-b-PAA) copolymer onto magnetite nanoparticles. The PEO blocks extended into aqueous media to keep the nanoparticles dispersed. Amines at the tips of the H2N-PEO corona were then linked through reaction with a PEO diacrylate oligomer to yield MBIClusters where the metal oxide in the precursor nanoparticles were distinctly separated by the hydrophilic polymer. The intensity average spacing between the magnetite nanoparticles within the clusters was estimated to be ~50 nm. These MBIClusters with hydrophilic intra-cluster space had transverse relaxivities (r2's) that increased from 190 to 604 s-1 mM Fe-1 measured at 1.4 T and 37 °C as their average sizes increased. The clusters were loaded with up to ~38 wt% of the multi-cationic drug gentamicin. MRI scans focused on the livers of mice demonstrated that these MBIClusters are sensitive contrast agents.

Aqueous dispersions of magnetite nanoparticles complexed with copolyether dispersants: experiments and theory.

Magnetite (Fe3O4) nanoparticles have been synthesized and complexed with carboxylate-functional block copolymers, and then aqueous dispersions of the complexes were investigated as functions of their chemical and morphological structures. The block copolymer dispersants had either poly(ethylene oxide), poly(ethylene oxide-co-propylene oxide), or poly(ethylene oxide-b-propylene oxide) outer blocks, and all of them had a polyurethane center block that contained pendent carboxylate groups. The complexes were formed through interactions of the carboxylates with the surfaces of the magnetite nanoparticles. The magnetite cores of the magnetite-copolymer complexes were near 10 nm in diameter, and the particles were superparamagnetic. Complexes with mass ratios of polymer to magnetite varying from 50:50 to 85:15 were studied. One of our objectives is to design complexes that form stable dispersions of discrete particles in water, yet that can be actuated (moved together) upon exposure to a uniform magnetic field. DLVO calculations that accounted for magnetic attractive interparticle forces, as well as van der Waals, steric, and electrostatic forces are presented. Compositions were identified wherein a shallow, attractive interparticle potential minimum appears once the magnetic term is applied. This suggests that it may be possible to tune the structures of superparamagnetic nanoparticle shells to allow discrete dispersions without a field, yet weak flocculation could be induced upon exposure to a field.