14
$\begingroup$

I have been trying to simulate the properties of Bi2Se3. But the lattice parameters given in Materials Project (which I believe is obtained computationally) is inconsistent with that of experimental data, i.e. a = 4.18976606 (Materials Project) , a=0.413(hq graphene). Even the band gap, which I believe is supposed to be 0.3 eV for bulk structure is different in Materials Project(0.54 eV).

So, When do I use the lattice parameter obtained computationally? Also, what is the reason for this difference? What to do if there is no available experimental data?

Thank You,

$\endgroup$
13
+100
$\begingroup$

To respond directly regarding the Materials Project data, I'm a staff member there so maybe I can shed some light.

Materials Project computed data is currently generated using a technique known as Density Functional Theory (DFT) with the PBE exchange-correlation functional. This results in some well-understood, systematic differences from experiment. Typically, this means that our computed lattice parameters will over-estimate experimental lattice parameters by 2-3% on average. Note that layered materials (any material where van der Waals bonding might be significant) will have larger errors in their inter-layer distance. Finally, note that these lattice parameters are nominally at 0 K, and do not take thermal expansion into account.

Note that lattice parameters on Materials Project are often given as their primitive cell, if you want the conventional lattice parameters make sure to download the CIF file in the "conventional" setting.

Band gaps will be systematically under-estimated by a large degree when using PBE (see our documentation). Spin-orbit coupling is also not included. The electronic band structures on Materials Project are most useful for seeing the shapes of the bands and the character of the gap (e.g. indirect, direct, between what symmetry points, etc.), the absolute magnitude of the band gaps are only useful for trends between different materials.

Better computational techniques can give results with smaller systematic errors, and we're constantly evaluating using some of these better techniques with the Materials Project. The trade-off here is that Materials Project tries to calculate properties for 100,000s of materials, and so using these better techniques is not always practically possible due to their computational cost.

With this context, to answer the question of "which should I use?", the question depends on what you want to use it for. If you want to know the "true" value, always defer to high-quality X-ray diffraction (bearing in mind the experimental value might be affected by grown-in strain, impurities, the temperature the measurement is taken at and other factors). However, if you want to do additional calculations with PBE, it's often easier to start from the previously-computed geometry. The computed geometry is also useful for examining differences between materials (e.g. varying composition) and also for materials where high-quality experimental data has not been acquired.

Likewise, for band gap, I would always defer to the experimental value, but of course there are also experimental issues too; experimentally, the optical gap is usually what is measured (e.g. via photoluminescence), there might be defect levels, finite temperature effects, excitonic effects, unintentional doping, Moss-Burstein shifts, etc., you might be only measuring the direct gap, whereas computationally you're predicting the fundamental gap (strictly speaking, the "Kohn Sham gap" using traditional DFT, which is another very important but subtle point). So there's no easy answer for which is better. The computational picture might give you a better picture for how a hypothetical pristine material might behave, but is typically most useful for trends and comparisons between similar materials.

Hope this helps! Happy to answer further questions.

$\endgroup$
1
  • 2
    $\begingroup$ +100. Thank you very much for your answer! $\endgroup$ – Camps Oct 9 '20 at 0:23
10
$\begingroup$

I will give a quick answer from my experience, but typically you want to use the computed lattice parameter. The reason for this is that otherwise you induce strain in your computed cell, but in general this is a red flag anyways. You really want your computed lattice parameter to match if possible which might involve using a different functional or even things such as DFT+U corrections.

I am actually not sure of any circumstance where you would want to use the experimental lattice constant over the calculated one. Maybe someone can list a reason. Also keep in mind the experimental lattice constant is taken at some experimental conditions, which are not represented in the calculation (frozen geometry, 0K, etc).

$\endgroup$
7
  • $\begingroup$ What is the problem with strain as long as you don't want to perform structural relaxations? $\endgroup$ – Gregor Michalicek Oct 3 '20 at 15:01
  • $\begingroup$ Strain will affect every property of the material almost. Band gaps, adsorption strength, etc. $\endgroup$ – Tristan Maxson Oct 3 '20 at 15:17
  • 1
    $\begingroup$ With regards to the end of @TristanMaxson's answer: The lattice constants themselves can also influence various properties. With this in mind, there can be cases where fixing the lattice constants at the experimental value can give you a more accurate property of interest, even if there is slight strain present. I've seen a few papers demonstrate this for gas binding at metal sites in MOFs. Whether this is the ideal route, fortuitous error cancellation, or luck is perhaps a topic of debate. Personally, I am with you here and rarely would suggest fixing the experimental lattice constants. $\endgroup$ – Andrew Rosen Oct 3 '20 at 15:52
  • 3
    $\begingroup$ Another scenario where one might fix the lattice constants is where it's just too computationally expensive to relax the lattice constants! Generally, you want to use a greater plane-wave kinetic energy cutoff to prevent Pulay stresses when doing volume relaxations. This can make simulations much more expensive than a simple relaxation of atomic positions. Depending on what you're looking to do, this might be the only feasible option, unfortunately. $\endgroup$ – Andrew Rosen Oct 3 '20 at 15:54
  • 2
    $\begingroup$ Be very careful of bandgaps in dft calculations. $\endgroup$ – Tristan Maxson Oct 3 '20 at 16:05
8
$\begingroup$

I have to disagree with the answer of @Tristan.

The x-ray diffraction, both powder and single crystal, has enough precision and quality (equipment and software) to give very good and real experimental values.

It is well known that very few DFT approximations (methods, functional, pseudopotential, basis set, etc.) give good lattice values. And when we said good, we mean in agreement with experimental ones (by the way, one of the path to test your simulation results is comparing with experimental ones).

So, my advice is always use data from experimental measurements. In this case, from deposited CIF files.

$\endgroup$
11
  • $\begingroup$ Can you quantify what accuracy you mean with "good lattice values"? $\endgroup$ – Gregor Michalicek Oct 4 '20 at 7:33
  • 1
    $\begingroup$ Experimental value is taken from a real material at finite temperature. If a DFT calculation can't take temperature into account, the experimental value of lattice constant will result in simulation of a stressed (stretched/compressed) material, which is not necessarily what one is after. $\endgroup$ – Ruslan Oct 4 '20 at 16:40
  • $\begingroup$ @GregorMichalicek, in agreement with experimental ones. $\endgroup$ – Camps Oct 4 '20 at 17:26
  • $\begingroup$ For many materials lattice constant predictions from DFT are within a few percent around the experimental values, even with basic XC functionals. Also the deviations are kind of systematic, which implies that you can predict ground-state structures. $\endgroup$ – Gregor Michalicek Oct 4 '20 at 21:41
  • $\begingroup$ @GregorMichalicek, the more recently CIF deposit (165226) in the ICSD database reported a value of 4.1355(5) Angs. (about the reported precision, please, take a look at this). This reported value was at the following reported work back in 1999: dx.doi.org/10.1021/ic9812858. $\endgroup$ – Camps Oct 5 '20 at 14:32

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.