NCSI: Parallel DFT Calculations and Birch-Murnaghan Analysis for Fundamental PV Research

Status	Completed
Mentor Name	Stephen Harnish
Mentor's XSEDE Affiliation	I received 3 Education Allocations through the BWSIP and mentored 7 XSEDE EMPOWER students. I have also participated in various NCSI workshop opportunities since 2003.
Mentor Has Been in XSEDE Community	4-5 years
Project Title	Parallel DFT Calculations and Birch-Murnaghan Analysis for Fundamental PV Research
Summary	The intern will assist mathematics, computer science and physics faculty and graduate students in large-scale, density functional theory (DFT) calculations and molecular dynamics (MD) simulations and visualizations. The general goal is to analyze properties of Si-based and related semiconductors to aid ongoing research in photovoltaic (PV) technologies. An important role for this apprentice/intern will be to develop bash scripts to create directories for the large parameter space and associated VASP input files and implement parallel jobs to analyze VASP output files. A specific focus will be to fit VASP output for various Si-based semiconductors to the Birch-Murnaghan equation of state, yielding four physical quantities - bulk modulus, ground state energy, volume, and the first derivative of bulk modulus. These will be compared to known values, when available for that semiconductor. Scripts and parallel code will automate this analysis of the VASP output files.
Job Description	The primary work will be with the Bluffton mentor, with partial assistance from colleagues at the University of Toledo. Some secondary consultations will also occur with Dr. Jay Gupta and his colleagues at The Ohio State University Center for Emergent Materials. The intern will continue to adapt VASP input files to initiate a systematic analysis of ground state and band-gap energies of silicon lattices with a wide range and geometry of dopants. This work will primarily be performed on the University of Toledo's Antec3 server. The second phase will involve fitting VASP output files for various Si-based semiconductors to the Birch-Murnaghan equation of state, yielding four physical quantities—bulk modulus, ground state energy, volume, and the first derivative of bulk modulus. These will be compared to known values, when available for that semiconductor. Scripts and parallel code will automate the analysis of their VASP output files. The third, and final phase will focus on very specific DFT calculations and MD simulations of interest to the Khare group contributing towards 1-2 published journal articles by the end of summer 2021. During this work, the apprentice/intern will also be trained on the Materials Project (https://materialsproject.org/)--a useful source of materials properties and calculation parameters.
Computational Resources	The student researcher will use resources of Bluffton and Toledo faculty as well as some gleaned from SC'11, SC'12 and BWSIP education programs. In addition, the intern will explore resources available at the HPC University. An education allocation from the OSC will play an integral role in MD simulations and visualizations to augment the use of VASP on the Toledo Antec3 server. The proposed intern is presently using Bridges-2 at the PSC for the Berkeley class in Applications of Parallel Computers. The final project for that course will initiate preliminary work for this summer research.
Contribution to Community	Ultimately, this project will contribute to foundational photovoltaic research, and holds solid potential for applications to meet energy sourcing needs in the coming years. As with prior mentorships, it will also serve to train the next-generation of researchers and computational scientists.
Position Type	Intern
Training Plan	For a well-prepared and motivated intern, we would expect quick mastery of resources and instructions provided. Appropriate training adaptations will be made if and as needed.
Student Prerequisites/Conditions/Qualifications	The ideal student researcher must have a strong background or demonstrated ability in mathematics and/or computer programming, along with core knowledge in solid-state physics, crystallography and physical chemistry. He or she should have a working knowledge of Python and/or C and the ability to adapt code from one to the other. Candidates should have the capability and motivation for self-directed research, literature review, and code adaptation and development. In addition, they should possess the personal and intellectual maturity for self-guided as well as collaborative research. Finally, the intern must have a firm grasp of fundamentals of photovoltaics (e.g., as taught in Tonio Buonassisiâ's MIT course) and the ability to process VASP input files. The student researcher should be available for 10-12 weeks at sufficient hours per week to qualify for the available internship. Based on all of the above qualifications, the proposed student would be an ideal intern for this research project. While local, in-person mentoring is intended for most segments, interns should be capable of remote research with guidance via e-mail, phone and/or Zoom meetings.
Duration	Semester
Start Date	06/01/2021
End Date	08/31/2021