Research Interests – Adam J. Jackson – a.j.jackson@physics.org

View Github Repository Open presentation in a new window

ajjackson

See all presentation from ajjackson

Research Interests – Adam J. Jackson – a.j.jackson@physics.org

0 0

research-interests

Presentation of research interests

On Github ajjackson / research-interests

Research Interests

Adam J. Jackson

a.j.jackson@physics.org

Info

This presentation is available online at http://ajjackson.github.io/research-interests

Timeline

2006–2011 Undergraduate MEng Chemical Engineering

  • University of Bath
  • Technical review (membranes for biomedical tech)
  • Research project (oxygen scavenging membranes)
  • 2009–2010 Industrial placement (surfactants)
  • Design projects

2011–2016 Doctoral Training Centre in Sustainable Chemical Technologies

  • University of Bath (Depts. of Chemistry and Chem. Eng.)
  • 2011–2012 MRes: two short projects
    • Oxidation of GaN
    • Dissolution of cellulose in ionic liquids
  • 2012-2016 PhD project: CZTS
    • Research group of Aron Walsh
    • Thesis accepted pending minor corrections

2016 Research assistant

  • Same research group as PhD
  • Method/code development
  • Supporting students in group

Relevant research

Oxidation of GaN

  • Ab initio study of thermodynamics of oxygen defects in GaN
  • Defect formation energy from DFT with PBEsol
  • Temperature effects from harmonic lattice dynamics
  • Chemical potentials of O2, N2 from literature

Oxidation of GaN - findings

  • Supercell size dependence: band-filling correction for dilute limit
  • ~1,000,000:1 N2 to O2 ratio needed to prevent oxidation
  • Paper published in Phys. Rev. B 10.1103/PhysRevB.88.165201

CZTS

  • Copper zinc tin sulfide
  • Thin-film absorber layer for earth-abundant PV
  • p-type "copper-poor" material typically produced
  • Poor reproducibility between experimental groups
  • Major concerns are low open-circuit voltage and presence of secondary phases

CZTS

  • Investigate relative stability of phases against \(T\), \(P\)
  • Similar approach to GaN, more materials: ab initio lattice dynamics
  • Initially looked at decomposition to binaries
  • → Sensitive to temperature, at low S pressure
  • Published in J. Mater. Chem. A 10.1039/c4ta00892h

Cu2S + ZnS + SnS + S(g) ⇌ Cu2ZnSnS4

CZTS

  • Move to include ternary phases
  • Ternary phase diagrams with "open" sulfur content
  • Vary chemical potentials with two parameters:
    • temperature and pressure

CZTS

CZTS

  • Found very narrow stability region for CZTS without CTS
  • Presented at conferences
  • Main finding of PhD thesis

Sulfur

  • While developing CZTS stability models, found inconsistent treatment of sulfur
  • Known to consist of a mixture of allotropes
  • Typically one allotrope is assumed to represent vapour phase
  • Last model of mixture developed in 1970s, made a lot of assumptions

Sulfur

  • Evolutionary algorithm used to generate candidate clusters, compared with literature
  • Used a selection to populate equilibrium model

Sulfur

  • Vibrations computed with range of DFT XC functionals
  • PBE0 frequencies with 96% scale factor provided best fit
  • Create "universal" chemical potential of sulfur atoms accounting for equilibrium as \(f(T,P)\)
  • Paper published in Chem. Sci. 10.1039/C5SC03088A

High-throughput screening

  • Work in progress
  • Supporting code is being developed openly at http://github.com/wmd-group/smact
  • "Low-fi" chemical screening of arbitrary element combinations
    • Simple tight-binding models
    • Materials availability
    • Electronegativity rules
    • Radius ratios
  • Paper and conference talks in preparation

Areas of interest

Experiment interpretation and prediction

  • Quantitative prediction of Raman and IR results
  • PV device properties more difficult - but important!
    • \(E_{g}\): Optics vs current/voltage vs VOC vs computation
    • Device modelling

Reproducible research & data sharing

  • Computational chemistry is deterministic! No excuses for difficulty reproducing work
  • Research councils require raw data to be made available
  • Go one step further by making all analysis available, e.g. http://github.com/wmd-group/sulfur_model
  • New databases of results are being developed (Materials Project, OQMD…)

Wrapping up

  • Work so far has primarily used GGA/hybrid DFT with supercells
  • Keen to branch out in both directions
    • Low-cost forcefield calculations and embedded cluster calculations
    • Beyond-DFT methods for improved optical modelling
  • Automation is good…
    • as long as it is easy to understand
  • HPC is good…
    • as long as the results are useful
1
Research Interests Adam J. Jackson a.j.jackson@physics.org