Render this...

April 3, 2007

At work I run Windows XP on my desktop and SSH into our linux systems (with -term facilitated by Exceed). Q. What are the best visualization options for viewing structure and charge density from VASP/Wien2K? I have previously used Materials Studio by Accelrys. It’s slow, expensive, eats memory and can’t export vector images. It’s almost worthy of being designed by Microsoft. Here are some freeware options I have been playing with recently:

(a) XCrySden

Originally aimed at Wien2K, but there are plenty of scripts around to convert VASP files. A very powerful program. It can automatically visualize the Brillouin zone for your cell, display good structures and produces some of the best isosurfaces I’ve seen. The EPS export didn’t work for me when I tried, but it warned me it would be tough with lighting turned on. The main problem is that it is a bit sluggish when run over SSH.

(b) VEND

I had never heard of this until a few weeks ago when I started using the excellent VICS-II for viewing my structure files. VICS-II arose from the VENUS suite which aims to satisfy all your modeling needs. It runs natively on windows and links in with an array of electronic structure programs. Fast and impressive. Exporting isosurfaces as vectors is quick and with very good results. Puts expensive packages to shame. (My winner!) Update: VEND has been replaced by the amazing VESTA!

(c) Lev00

In the eighties the only computer I had was a Spectrum ZX that my cousin gave me (after they fell out of vogue). I’m sure it could have run Lev00. That’s not to say it’s not a useful suite. It can readily treat spin density files, linear combinations of density files and a has a number of nifty analysis tools. The visualization is purely contour based. You specify the plane and the contour ranges, and a second later you have a picture. Nice if you’re in a retro mood, but I like bright colours and fancy buttons.

(d) Loose ends: I remember using Vaspview at some stage and it worked somewhat, however when I downloaded the latest compilation today, it just kept freezing. I also attempted to install P4VASP. It needs a lot of extra packages to be linked. Too much effort for the moment. WXDragon is handy for checking CIF files and exporting directly as program input, but its visualization isn’t up to scratch. Poor quality pictures and limited functionality. There’s a new version due out soon which might improve things though.

Read More

Physics vs Chemistry

March 28, 2007

Having a background in chemistry and working with a group of physicists is interesting. They may have their fancy models, but we have our chemist’s intuition! One of the biggest differences I have noticed so far is the use of symmetry.

In terms of crystal lattices, I am more used of thinking in terms of conventional unit cells with lattice vectors along cartesian axes. Primitive cells for centered crystals make more sense in terms of the cost of calculations (and for calculation of band and phonon dispersion), but they are harder to visualize and construct - especially for body-centered monoclinic cells that I am currently concerned with.

I happened to come across a nice free package called Lev00. It can work with VASP, Crystal and Siesta. So far I have only used the subprogram Tetr which can produce primitive cell structure files based on previously generated VASP POSCAR files, or work directly from crystallographic data. Very handy! It seems to do a lot of other nifty things, I just haven’t had a chance to use them yet. The source code is available here.

I also picked up another nice book for $16 on amazon. “Group theory and quantum mechanics”. It is a reprint of a book from the sixties (I don’t think too much has changed since then in group theory). It’s small and concise. It reads like a lecture series, and it’s definitely not pop science. It would be good to have some basic background on group theory before starting. There are some nice references to old Physical Review papers along the way, for when more detail is needed.

Read More

Lone pairs in the solid state (Part One)

March 18, 2007

I need to get a presentation of my PhD research ready for my first group meeting, so I have been updating some old pictures now that I am more fluent in the workings of Macintosh (a welcome transition from years of Windows). Below is an animation which illustrates the main orbital interactions which result in PbO (and likewise SnO) adopting a distorted crystal structure (litharge).

Covalent interactions (sharing of electrons) in metal oxides is more important than previously considered and can explain some unusual properties of these materials.

Surprisingly, considering some of their useful properties, the subject of lone pairs in the solid state hasn’t really been updated since the intra-atomic hybridization models of L. E. Orgel and co-workers back in the 1950’s. Their model is based on direct mixing between the cation s and p states. However, our work has shown that interaction with anion p states of appropriate energy is required to bridge the energy gap between the cation s and p which are two far apart in energy to couple directly. This can go on to explain the unusual structural trends observed in these materials that could not be explained through previous hypothesis.

The strongest covalent interaction in PbO is between the anion p and cation s states. This results in a bonding interaction at the bottom of the valence band (majority Pb 6s states). There is a corresponding filled anti-bonding combination at the top of the valence band (majority O 2p states). Pb 6p can further interact and help stabilize these antibonding states only if the coordination environment of the Pb atoms is non-centrosymmetric. This is the driving force behind PbO (and likewise SnO) adopting a distorted structure. It is the coupling of Pb 6p with the antibonding O 2p/Pb 6s states that results in the observed ‘lone pair’ in the electronic distribution of the metal cation.

Read More

Water splitting

February 21, 2007

Thus use of hydrogen as a dependable fuel is still a long way off. Before it happens we need a viable storage solution and more importantly a viable production technique. Efficient photoelectrolysis of water to produce hydrogen and oxygen would be one good solution. It cuts out any lossy intermediates, simply: sun → water → hydrogen. The problem is that current catalysts either have very short activity times (low stability in water) or have band gaps too large to absorb a significant amount of visible light. Simulating new catalytically active materials is one of my first new research projects. One short and fulfilling step from basic to applied research.

Read More

Optical properties in VASP

February 10, 2007

‘NAN’ is my least favorite high frequency dielectric constant. If the optical property subroutine in VASP worked better for magnetic systems, my life would be much easier. Time to move over to WIEN2K? I think so, well at least until VASP 5 comes out (this summer hopefully). Attending a WIEN workshop in June should ease the transition for the time being.

Read More