3
$\begingroup$

I want to calculate Gibbs free energy of hydrogen atom adsorbed on an adsorption site (ΔGH*), to evaluate the HER activity of that particular surface of material.

How can i identify all possible H adsorption sites(*) on the corresponding surface of my material computationally ?

$\endgroup$
8
  • $\begingroup$ Say for example all possible H adsorption sites on (011) surface of MoS2. $\endgroup$ Apr 3 at 7:35
  • 1
    $\begingroup$ If using DFT which is computationally expansive, we can scan whole sheet of minimum unit cell with NSW=0 which will be faster and we can get basic idea. $\endgroup$ Apr 3 at 14:06
  • $\begingroup$ @Pranav kumar Yes i am using DFT via Quantum-Espresso.Can you elaborate a little more on the method you mentioned. $\endgroup$ Apr 3 at 14:24
  • 1
    $\begingroup$ In short, Change hydrogen position on top of fixed MoS2 structure and calculate energy (no ionic movement of MoS2). You will get two dimension energy surface. However you can allow hydrogen to relax along normal to surface if you have sufficient resources. This 2d surface will provide an approximate position of hydrogen to sit. $\endgroup$ Apr 3 at 14:48
  • 1
    $\begingroup$ You would have a better luck looking at some Monte Carlo (MC) schemes to identify the possible adsorption sites. Performing a full 2D DFT scan is computationally very expensive. $\endgroup$ Apr 3 at 19:22

1 Answer 1

1
$\begingroup$

A quick search on google scholar returned multiple papers that have used Kinetic Monte Carlo simulations to investigate the binding of small molecules with surfaces. This paper gives the methodology for using MC simulations to investigate the binding of CO$_2$ molecules with MgO(100) surfaces. The same methodology should be applicable for the particular case that you are interested in.

Other references are

  1. https://pubs.aip.org/aip/jcp/article/77/12/6296/217092/Adsorption-of-polymer-chains-at-surfaces-Scaling
  2. https://www.sciencedirect.com/science/article/abs/pii/0039602889907310
  3. https://pubs.acs.org/doi/full/10.1021/la8020595
$\endgroup$
3
  • $\begingroup$ Thank you @Hemanth Haridas $\endgroup$ Apr 3 at 19:53
  • 1
    $\begingroup$ Kinetic Monte Carlo is different then metropolis (simple mote Carlo). For kinetic Monte Carlo you need migration barrier energy of hydrogen. The indicated paper is based on classical potential not on DFT and used grand canonical as well. Monte Carlo with DFT is highly expansive and even simple metropolis of hydrogen on MOS2 surface is computational expansive than scanning only those points which are symmetrically inequivalent. $\endgroup$ Apr 5 at 6:12
  • $\begingroup$ It would be great if people who downvoted the answer could provide (i) either an answer to the question, or (ii) at least explain the reasoning behind the downvotes. Now to answer the question by @Pranavkumar I agree that running an MC algorithm with DFT is very expensive. My logic would be to run a coarser run using MC to get possible sites, and then refine them using DFT, because combinatorial explosion of sites is a very real thing. $\endgroup$ Apr 5 at 15:03

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .