The Division of Materials Research and the Office of Advanced Cyberinfrastructure contribute funds to this award that supports research and education towards developing methodologies for the computational modeling of materials. Our society depends on novel materials for many aspects of life, ranging from smart phones and computers to transportation and industrial processes. Historically, such materials have been developed in experimental labs, but that approach is very expensive and time consuming. With advances in computer technology, we are now at a point where we can model the properties of materials before they are ever made in a lab, providing a potentially less expensive and faster way to develop new materials.
All materials are held together by fundamental forces between their constituent atoms and/or molecules. Although weaker than most other forces, the van der Waals force is crucial for many chemical/biological/industrial processes and phenomena of every-day life, allowing for example a gecko to run up a wall. Unfortunately, the most widely used computational technique for modeling materials has difficulty capturing van der Waals forces. Over time, two distinct approaches have been developed that represent different design philosophies. With this NSF award, the PI seeks to combine both, with the objective to derive an improved and unified theory that combines the advantages of both approximations and overcomes their shortcomings. The main scientific outcome of the award will be a freely available computer code that allows researchers to more accurately model materials where van der Waals forces play an important role.
For the educational part of this project, young scientists will be trained and mentored to become independent researchers in fields that are important to our nation's continued success. In addition, the research activities are coupled with outreach at a local science museum to increase scientific literacy and engagement with science in the community. The proposed Molecular Playground at the local science museum will serve over 200,000 visitors annually, broadening participation in science among those who live in the Piedmont-Triad area in North Carolina.
The Division of Materials Research and the Office of Advanced Cyberinfrastructure contribute funds to this award that supports research and education towards developing a unified theory of van der Waals forces within nonlocal density functional theory. Our society's increased need for novel materials has shifted their development strategy to modeling a material's properties on the computer before it is ever synthesized in the lab. The capability of modern methods to accurately model a variety of interactions in novel materials has thus come into focus. Of particular interest are van der Waals interactions, which are crucial for the structure of many materials reaching from cement to DNA, playing intricate roles in catalysis, gas sequestration/storage, and biochemical processes. Due to its favorable computational cost, density functional theory (DFT) is the ab initio modeling method of choice. However, with standard approximations it unfortunately cannot reliably capture van der Waals interactions. Two density-based approximations to the exchange-correlation functional have been developed that incorporate van der Waals interactions in DFT, both with advantages and shortcomings. The objective of the current proposal is to derive an improved and unified theory that combines the advantages of both approximations and overcomes their shortcomings. The result will be a combined, parameter-free, and generally applicable and practical method to capture van der Waals interactions in DFT with improved accuracy.
While both approximations have a common starting point in many-body theory, they differ in the construction of their nonlocal correlation density functional. In both cases, the dielectric function of the material is approximated through a simple plasmon-pole model, but the specific models adhere to different numbers of exact physical constraints and thus result in models of different complexity. Surprisingly, the simpler method with fewer exact constraints often performs better in terms of accuracy, but suffers from problems of transferability to materials in different situations. The proposed work presents a pathway to unify both methods and find an improved description of the plasmon-pole model that adheres to the more important constraints, such as charge conservation, but is also simple to evaluate. The main outcome of the award will be a computational tool to more accurately capture nonlocal van der Waals effects within DFT. The PI plans to release the developed software as open source, as part of the public-domain electronic-structure package Quantum-Espresso, and build a user community around the language by ensuring that interested researchers are able to contribute to the codebase. This will allow a wider growth of the project. This aspect is of special interest to the software cluster in the Office of Advanced Cyberinfrastructure, which has provided co-funding for this award.
For the educational part of this project, young scientists will be trained and mentored to become independent researchers in fields that are important to our nation's continued success. In addition, the research activities are coupled with outreach at a local science museum to increase scientific literacy and engagement with science in the community. The proposed Molecular Playground at the local science museum will serve over 200,000 visitors annually, broadening participation in science among those who live in the Piedmont-Triad area in North Carolina.
