PhD Candidate -- Efficient (Bayesian) methods to estimate ReaxFF parameters

     
Employer
Published
WorkplaceFlemish Region, Ghent, Belgium
Category
Position

Description

Job description

We are pleased to announce a fully funded PhD position, for a duration of three years, within the project AutoCheMo -- Automatic generation of Chemical Models. This is an EU-funded European Industrial Doctorate network coordinated by Software for Chemistry and Materials (SCM, SME located in Amsterdam), with the Center for Molecular Modeling (CMM) at Ghent University (T. Verstraelen) and the chair of Technical Thermodynamics (K. Leonhard) at RWTH Aachen University. Four PhD candidates (Early-Stage Researchers, ESRs) will be trained and involved in research in this project and this vacancy is specific for one of those four positions, which will be hosted for 22 months at UGent and 14 months at SCM. More information on AutoCheMo can be found here: ?url=https%3A%2F%2Fwww.scm.com%2Fcollaborations%2Feu-projects%2Fautochemo%2F&module=jobs&id=1122" target="_blank" rel="nofollow">www.scm.com/collaborations/eu-projects/ autochemo/ (See project 3.) More details can also be found in a previous announcement of the job openings for this project: ?url=https%3A%2F%2Fwww.scm.com%2Fnews%2Fjob-openings-4-phd-fellowships-theoretical-chemistry-method-software-development%2F&module=jobs&id=1122" target="_blank" rel="nofollow">https://www.scm.com/news/job-openings-4-phd-fellowships-theoretical-chemistry-method-software-development/

Your goal within this AutoCheMo is to create an enhanced ReaxFF parameter calibration algorithm. (Project 3) In addition to the general utility of this algorithm in the ReaxFF community, it would also be very beneficial for the automated force field parameterization protocols that are studied by one of the other ESRs. (Project 2) As a first step, you will develop and benchmark a new optimization algorithm, building on our preliminary work, which showed how one can drastically improve the rate of convergence compared to existing algorithms. Later you will generalize this methodology to also account for the statistical uncertainties on ReaxFF parameters and their impact on simulation outcomes, within a Bayesian framework.

Profile of the candidate

Candidates must have obtained a M.Sc. degree in physics, applied physics, chemistry or chemical engineering in the past four years. We are seeking exceptionally qualified and motivated applicants for this position. Students with high grades and research experience in their B.Sc. and M.Sc. studies from renowned universities are strongly preferred. Other selection criteria include an outstanding knowledge of statistics, programming, chemical reactivity, numerical analysis and molecular simulation. Besides the scientific and technical background, we also value good communication skills (written and oral) and sufficient social aptitude to integrate in a diverse and competitive research group.

Due to EU mobility rules, you should not have spent more than 12 months in Belgium over the last three years prior to the starting date of the contract.

How to apply

Applications should be sent to Toon.Verstraelen [at] UGent[.]be before December 15, 2018, with the subject: "Application for Ph.D. position - Efficient (Bayesian) methods to estimate ReaxFF parameters".

The application should include:

  • a detailed Curriculum Vitae with a complete biography,
  • a cover letter describing why you should be considered for this position,
  • transcripts of B.Sc. and M.Sc. courses and grades,
  • copies of educational certificates and
  • a list of two professors who have supervised you in your B.Sc. and/or M.Sc. research project and who are willing to provide a letter of recommendation.


All documents should be in PDF format and written in English. From all applications, a short-list of candidates will be made, who will be contacted for an interview over Skype.

Web

www.ugent.be , www.ugent.be/en...
In your application, please refer to myScience.be
and reference  JobID 1122.