Search for: All records

Nanoscale phononic materials are critical to next generation devices for energy management and information technology. A nanojunction (NJ) consisting of two gold nanoparticles (AuNPs) bridged by Ferrocenyl (Fc) molecules is one such material. We use molecular dynamics simulations to demonstrate that they can exhibit a thermal switching ratio of R > 200, allowing for directed control of heat transport. Both states—with electric field switched ‘ON’ and ‘OFF’—are represented in the models through corresponding atomistic partial charges. We report response of the NJ across a broad range of parameters by varying the electric field strength, temperature set point, AuNP size, AuNP-to-AuNP distance, and number of Fc molecules. We find a nonlinear relationship between the thermal switching ratio and the number of Fc molecules. The optimum performance of R > 300 is achieved when 2 to 4 Fc molecules bridge between the AuNPs. In a Medusa AuNP—with 140 Fc molecules on a 4 nm diameter AuNP—we can achieve R = 31, which is still larger than previously reported devices. 

Characterization and prediction of the interactions between engineered nanoparticles (ENPs), proteins, and biological membranes is critical for advancing applications to nanomedicine and nanomanufacturing while mitigating nanotoxicological risks. In this work, we employ a coarse-grained dissipative particle dynamics (DPD) simulation to investigate the interactions among cytochrome c (CytC), lipid bilayers, and citrate-coated gold nanoparticles (AuNPs). We updated the DPD potential to accurately represent binding potentials between molecules, and validated the model relative to an all-atom representation. The DPD simulations successfully replicate experimental observations: CytC facilitates the binding of citrate-coated AuNPs to lipid bilayers composed of 90% dioleoylphosphatidylcholine (DOPC) mixed with 10% stearoylphosphatidylinositol (SAPI) or 10% tetraoleoyl cardiolipin (TOCL) but not to pure 100% DOPC bilayers. In addition, the simulations reveal nuanced differences in binding preferences between CytC, the lipid bilayers, and the ENP, at a scale that is not presently directly observable in experiments. Specifically, we found that the surface coating of the nanoparticles─viz variations in the CytC surface density─affects the protein-mediated binding with the bilayers. Such a molecular-sensitive result underscores the utility of DPD simulations in simulating complex biological systems. 

Polymer-networked nanoparticles are a promising alternative to silicon semiconductors for the realization of neuromorphic computing platforms. Variations in the interaction between gold nanoparticles (AuNPs) and polyelectrolyte linkers lead to the controlled formation of engineered nanoparticle network (ENPN) structures exhibiting a broad range of topologies and dynamics. Using dissipative particle dynamics (DPD) simulations, we designed tri-block copolymers with polyelectrolyte ends that can selectively attach to each of two AuNPs, and bridged them together through a middle polymer segment (or block). We leverage our earlier finding that AuNPs have well defined valencies—that is, an optimal number of polymers that can fit (or fill) their surface—for a specific choice of the outer blocks at a given polymer length. The precise selection of the AuNP valence allows for controlled binding between polymers and AuNPs. Meanwhile, the choice of the middle block enables control over inter-nanoparticle spacing and network topology. We found that ENPNs can achieve distinct and stable states, satisfying a necessary condition for primitive neuromorphic computing. By swapping the surface coating ligand from citrate to mercaptopropionic acid (MPA), the valence on a given sized nanoparticle is also increased. Thus, we found that the selection of the surface coating consequently affects the designed ENPN structures, allowing for more flexibility in searching for optimal components. 

Artificial intelligence (AI) models remain an emerging strategy to accelerate materials design and development. We demonstrate that convolutional neural network (CNN) models can characterize DNA origami nanostructures employed in programmable self- assembling, which is important in many applications such as in biomedicine. Specifically, we benchmark the performance of 9 CNN models—viz. AlexNet, GoogLeNet, VGG16, VGG19, ResNet18, ResNet34, ResNet50, ResNet101, and ResNet152—to characterize the ligation number of DNA origami nanostructures in transmission electron microscopy (TEM) images. We first pre-train CNN models using a large image data set of 720 images from our coarse-grained (CG) molecular dynamics (MD) simulations. Then, we fine-tune the pre-trained CNN models, using a small experimental TEM data set with 146 TEM images. All CNN models were found to have similar computational time requirements, while their model sizes and performances are different. We use 20 test MD images to demonstrate that among all of the pre-trained CNN models ResNet50 and VGG16 have the highest and second highest accuracies. Among the fine-tuned models, VGG16 was found to have the highest agreement on the test TEM images. Thus, we conclude that fine-tuned VGG16 models can quickly characterize the ligation number of nanostructures in large TEM images. 

The development of devices that improve thermal energy management requires thermal regulation with efficiency comparable to the ratios R ∼ 105 in electric regu- lation. Unfortunately, current materials and devices in thermal regulators have only been reported to achieve R ∼ 10. We use atomistic simulations to demonstrate that Ferrocenyl (Fc) molecules under applied external electric fields can alter charge states and achieve high thermal switch ratios R = Gq/G0, where Gq and G0 are the high and low limiting conductances. When an electric field is applied, Fc molecules are positively charged and the SAM-Au interfacial interaction is strong, leading to high heat conductance Gq. On the other hand, with no electric field, the Fc molecules are charge neutral and the SAM-Au interfacial interaction is weak, leading to low heat conductance G0. We optimized various design parameters for the device performance, including the Au-to-Au gap distance L, the system operation temperature T, the net charge on Fc molecules q, the Au surface charge number Z, and the SAM number N. We find that Gq can be very large and increases with increasing q, Z, or N, while G0 is near 0 at L > 3.0 nm. As a result, R > 100 was achieved for selected parameter ranges reported here. 

Not AvailableNanoparticle networks have potential applications in brain-like computing yet their ability to adopt different states remains unexplored. In this work, we reveal the dy namics of the attachment of polyelectrolytes onto gold nanoparticles (AuNPs), using a bottom-up two-bead-monomer DPD (TBM-DPD) model to show the heterogeneity of polymer coverage. We found that use of one polyelectrolyte homopolymer limits the complexity of the possible engineered nanoparticle networks (ENPNs) that can be built. In addressing this challenge, we first found the commensurability rules between the numbers of AuNPs and poly(allylamine hydrochloride)s (PAHs). This gives rise to a well-defined valency of a AuNP which is the maximum number of AuNPs that it can accommodate. We further use an engineered block copolymer, which has a conductive middle block to mediate the distance between a dimer of AuNP. We argue that by controlling the length of conductive block that is connecting the AuNPs and their respective topology, we can have ENPNs potentially adopt multiple states necessary for primitive neuromorphic computing. 

We have pursued the use of polymer-networked engineered nanoparticles as a candidate material capable of retaining information or perhaps even processing information in some prescribed way. Such operations would be of use for the neuromorphic engineering of materials that can compute intrinsically—that is, that they are in no way subject to a von Neumann architecture—and they have been identified as autonomous computing materials. Using trajectories integrated to much longer time steps than previously observed, we can now confirm that the response of the polymer-networked engineered nanoparticle arrays are highly sensitive to external perturbations. That is, the specific internal connections around given nanopar- ticles can be assigned to states useful for information processing, and the variations in their physical properties can result in specific responses allowing the state to be read. Moreover, their resulting equilibrium properties also depend on such external driving, and hence are subject to control which is a minimal requirement for these materials to be candidates for autonomous computing. We also demonstrate that using long polymer chains can help regulate the networks structures by increasing the 1st nearest links and reducing other links.