BEGIN:VCALENDAR
VERSION:2.0
PRODID:icalendar-ruby
CALSCALE:GREGORIAN
X-WR-CALNAME:PhD Final Defense - Priyanka Bholanath Shukla
X-WR-TIMEZONE:Eastern Time (US & Canada)
BEGIN:VEVENT
DTSTAMP:20261011T062546Z
UID:tag:localist.com\,2008:EventInstance_50213716669272
DTSTART:20250730T140000Z
DTEND:20250730T160000Z
DESCRIPTION:Title: Assessing and Improving the Accuracy of Chemical Propert
 y Predictions Using Electronic Structure Theory and Machine Learning Appro
 aches\n\n \n\n Abstract: Accurate prediction of chemical properties is cru
 cial for advancing catalysis\, materials design\, and reaction discovery. 
 While density functional theory (DFT) is widely used\, it has two major li
 mitations: (1) Self-interaction errors (SIE)\, where an electron erroneous
 ly interacts with itself\, leading to inaccurate predictions of properties
  such as band gaps\, reaction barriers\, electron affinity\, and others\, 
 and (2) high computational cost\, which restricts atomistic simulations to
  relatively small systems. My thesis addresses both issues through improve
 d first-principles methods and machine learning integrations.\n\nThe first
  part of my work focuses on overcoming SIE in DFT\, particularly in system
 s with stretched bonds\, which often occur in transition states. My work a
 ddresses these limitations by applying self-interaction corrections (SIC) 
 using the Perdew-Zunger (PZSIC) and locally scaled (LSIC) methods\, implem
 ented within the Fermi-Löwdin orbital SIC (FLOSIC) scheme that uses orbit
 al-by-orbital corrections. We found that both PZSIC and LSIC significantly
  improved chemical reaction barrier predictions for the BH76 benchmark dat
 aset\, with SIE most pronounced in stretched bond orbitals. The BH76 bench
 mark dataset is comprised of only gas phase reactions. However\, since mos
 t catalytic reactions occur on transition-metal surfaces\, the impact of S
 IC on transition metal energetics remains unexplored.\n\n \n\nTo address t
 his\, we evaluated the performance of DFT\, PZSIC\, and LSIC in describing
  the energetics of 3d transition metals. We introduced a new measure calle
 d “sd energy imbalance” to assess the errors in s- and d- electron sec
 ond ionization energies of the 3d atoms\, avoiding the use of excited stat
 e energies. While LDA\, PBE\, and r2SCAN provided a reasonable balance bet
 ween s and d states\, PZSIC introduced significant errors of up to 2 eV.  
 This was attributed to an energy penalty associated with noded 3d orbitals
 . This penalty was most pronounced when an electron was removed from half-
 filled (3d5) or fully filled (3d10) subshells. LSIC mitigated these errors
 . Understanding and reducing the error associated with describing 3d orbit
 als is critical for accurately modeling redox reactions and oxidation-stat
 e changes in transition metal systems.\n\n \n\nBuilding on these insights\
 , we investigated SIC performance for NOx reduction on a Cu-SSZ-13 zeolite
  cluster model. We benchmarked the performance of DFT (LDA\, PBE\, and r2S
 CAN)\, PZSIC\, and LSIC methods against CCSD(T) reference energies. Among 
 LDA-based functionals\, LDA overestimated adsorption energies and underest
 imated reaction barriers due to SIE. While PZSIC improved the description 
 of reaction barriers in some cases\, it introduced large errors in systems
  involving changes in the Cu oxidation state\, with adsorption energy erro
 rs exceeding 3 eV. These errors resulted from the SIC energy penalty assoc
 iated with the 3d10 configuration of Cu. The energy penalty also caused PZ
 SIC to favor 3d9-like FOD configurations in singlet systems\, giving rise 
 to errors in the PZSIC densities. Thus\, LDA@PZSIC often performed worse t
 han uncorrected LDA because of these density errors. LSIC mitigated these 
 energy penalties and provided a balanced description of both adsorption an
 d reaction energetics. Overall\, LSIC outperformed LDA and PZSIC.\n\n \n\n
 To address DFT’s computational limitations\, we developed an active lear
 ning workflow that combined DeePMD-based machine learning interatomic pote
 ntials (MLIPs) with transition-state search algorithms. This workflow enab
 led efficient\, unbiased exploration of reactive potential energy surfaces
  without requiring prior knowledge of products or intermediates. Our workf
 low identified new pathways and accurate reaction barriers with near-DFT a
 ccuracy for solution-phase methanimine hydrolysis. This work provides meth
 odology to sample reactive intermediates effectively allowing new avenues 
 for automated reaction discovery in catalysis and small-molecule chemistry
 \n\n \n\nDissertation Chair:\n\nDr. J. Karl Johnson\, Department of Chemic
 al and Petroleum Engineering\, University of Pittsburgh\n\n \n\nCommittee 
 Members:\n\nDr. John Keith\, Department of Chemical and Petroleum Engineer
 ing\, University of Pittsburgh\n\nDr. Kenneth Jordan\, Department of Chemi
 cal and Petroleum Engineering and Department of Chemistry\, University of 
 Pittsburgh\n\nDr. Koblar Alan Jackson\, Department of Physics\, Central Mi
 chigan University
GEO:40.444043;-79.958003
LOCATION:Benedum Hall\, 1145
SUMMARY:PhD Final Defense - Priyanka Bholanath Shukla
URL;VALUE=URI:https://calendar.pitt.edu/event/phd-final-defense-7955
CATEGORIES:Defenses
END:VEVENT
END:VCALENDAR