Sodium Boride and Crystal Structures

AZoM talked with Professsor Artem Oganov, to find out more about sodium boride and his recent research.

What is sodium boride and what are its properties?

Boron is arguably the most complicated chemical element in the Periodic Table, for many reasons. Boron is an element that has extremely complicated crystal structures, which are very sensitive to impurities.

For a long time people couldn't figure out which allotrope of the element is the most stable and the phase diagram of boron was not established until very recently. Boron forms compounds with most elements in the Periodic Table and many of these have exceptionally complicated crystal structures, as well as counter-intuitive compositions. For example, YB66 or PuB100.

Why has it been difficult to determine structures of borides previously?

Boron chemistry is a challange to understand. It's also hard for researchers to figure out the actual chemical composition and the crystal structures. Part of the reason is that borides are often formed as powders, containing mixture of phases rather than a single phase.

Additionally, boron is a light element which weakly scatters x-rays. So establishing positions of boron atoms with traditional x-ray diffraction is not an easy thing, especially when they're present together with heavy elements.

So for these reasons, crystal structures and even chemical compositions of borides are rather complicated and not fully established for many of the known compounds. But it's important to understand them because borides can have a lot of interesting properties, including superhardness and thermoelectricity. Some are superconductors, but many are semiconductors and potentially useful photovoltaics.

How did your research get around these difficulties?

This is where my research will help - I have invented the methodology for predicting crystal structures, called USPEX. It actually does more than just crystal structure prediction: if you give my program a set of chemical elements (for example sodium and boron), in one single calculation it can give you all the chemical formulae and crystal structures of all the stable compounds formed by these elements. In one calculation you would find the structures of the pure elements (sodium and boron in this case) and also the structures and compositions of all of the stable borides.

Sodium boride is one of those problematic cases where experiments in the past could not agree on the chemical formula and crystal structure of one of the compounds. Two compounds had been known - Na3B20 and Na2B30/Na2B29. Researchers could not agree whether the composition of the latter was Na2B30 or Na2B29 - two atoms of sodium and either 29 or 30 atoms of boron.

Decafes ago, researchers suggested that there were 30 atoms of boron (Na2B30), but then, around the turn of the century German scientists suggested that the chemical composition and crystal structure determined previously were wrong. They thought that one boron atom needed to be removed, yielding the formula Na2B29.

This peculiar case has attracted attention of my colleagues from Nankai University (China), led by Prof. Xiang-Feng Zhou, and we jointly embarked on this problem. With our method, we discovered something new. We found no problem with the Na3B20 compound, which is an undisputed compound - the chemical formula and crystal structure are published and have never been debated. The other stable boride we found was Na2B30, and we did not see anything like Na2B29. We compared the energetics of the structure with the compound we found using the new methodology, with the previously proposed structures. It turns out, that from the point of view quantum mechanical calculations, our structure is more stable than any of the previously published ones.

We established the right chemical composition, so initial proposition of Na2B30 was correct. Unfortunately, both original groups turned out to publish incorrect crystal structures. The crystal structure that we predicted in this study was different from all of the previously published theories but matches experimental data. This is an interesting turn in the story because it shows that determination of the crystal structure is dangerous. We always knew it that there could be more than one crystal structure explaining your data, but now we see a growing number of examples, and we see that theory can provide invaluable help in discriminating between these different structural models.

There is another lesson that we have learned. If you look at the three models that have been put forward (the originally proposed experimental Na2B30, Na2B29 and our Na2B30), according to theoretical calculations the Na2B29 structure should have been metallic. The previously proposed Na2B30 structure would have been a semimetal, so a poor conductor of electricity. Our structure would be a semiconductor. Additionally, in that order, the phases also become more and more stable.

In the same sequence we also see changes in hardness - the hardness increases greatly, by a whopping 50 percent. This is all due to slightly changing the chemical composition and the crystal structure, so it really shows that every single atom matters when it comes to physical properties. You subtract an atom from a semimetal, it may become a metal. Rearrange the atoms somewhat and you get an increase in hardness. One gap opens and you get a proper semiconductor from a semimetal. It shows the sensitivity of the properties to the crystal structure.

Regarding the hardness, our calculations show that the stable phase of sodium boride Na2B30 that we have established is very close to being a superhard material.

What does it mean for a material to be superhard and why is this important?

Superhard materials are defined as those that have a Vickers hardness greater than 40 gigapascals (GPa). Our material had a hardness upward of 37 GPa, which is very close to the borderline. Given the uncertainty of theoretical predictions, the hardness could be even over that. So it is potentially a superhard material, unlike the other compounds and crystal structures that have been proposed.

So to put the long story short, this shows the power of the USPEX, which was able to predict all the stable components of this complicated system in one calculation, resolving a long-standing controversy, and establishing interesting properties for sodium boride.

About Professor Artem Oganov

Professor Artem Oganov is the current Head of Computational Materials Discovery Laboratory at the Skolkovo Institute of Physics and Technology and an Elected Member of Academia Europaea (M.A.E.). He is also a Professor of the Russian Academy of Sciences and has published over 200 research papers.