Boron nitride (BN) is a versatile material with applications in a variety of technical and scientific fields. This is largely due to an interesting property of BN called “polymorphism”, characterized by the ability to crystallize in more than one type of structure. It usually happens in response to changes in temperature, pressure, or both. Moreover, the different structures, called “polymorphs”, differ remarkably in their physical properties despite having the same chemical formula. Accordingly, polymorphs play an important role in material design, and knowledge of how to selectively promote the formation of the desired polymorph is crucial in this regard.
However, BN polymorphs pose a particular problem. Despite performing several experiments to assess the relative stabilities of BN polymorphs, a consensus has not emerged on this topic. While computational methods are often the approach of choice for these problems, BN polymorphs have posed serious challenges to standard computational techniques due to the low “van der Waals interactions (vdW)” between their layers, which is not taken into account in these calculations. In addition, the four stable BN polymorphs, namely rhombohedral (rBN), hexagonal (hBN), wurtzite (wBN) and zinc-blende (cBN) , manifest in a narrow energy range, making it even more difficult to capture small energy differences with vdW interactions.
Fortunately, an international research team led by Assistant Professor Kousuke Nakano of the Japan Advanced Institute of Science and Technology (JAIST) has now provided evidence to settle the debate. In their study, they tackled the problem with a state-of-the-art first-principle computational framework, namely fixed-node diffusion Monte Carlo simulations (FNDMC). FNDMC represents a step in the popular quantum Monte Carlo simulation method, in which a parameterized quantum many-body “wave function” is first optimized to reach the ground state and then fed to FNDMC.
Additionally, the team also calculated the Gibbs energy (the useful work obtainable from a system at constant pressure and temperature) of the BN polymorphs for different temperatures and pressures using the functional theory of the density (DFT) and phonon calculations. This article was posted on March 24, 2022 published in The Journal of Physical Chemistry C.
According to the FNDMC results, hBN was the most stable structure, followed by rBN, cBN and wBN. These results were consistent at both 0 K and 300 K (room temperature). However, the DFT estimates gave conflicting results for two different approximations. Dr. Nakano explains these conflicting results: “Our results reveal that the estimation of relative stabilities is strongly influenced by the correlational exchange functional, or the approximation used in the DFT calculation. Therefore, a quantitative conclusion cannot be drawn using the DFT results. , and a more precise approach, such as FNDMC, is required.”
Notably, the results of FNDMC were in agreement with those generated by other refined computational methods, such as the “coupled cluster”, suggesting that FNDMC is an effective tool for dealing with polymorphs, especially those governed by vdW forces. The team has also shown that it can provide other important information, such as reliable reference energies, when experimental data is not available.
Dr. Nakano is excited about the future prospects of the method in the field of materials science. “Our study demonstrates the ability of FNDMC to detect tiny energy changes involving vdW forces, which will stimulate the use of this method for other van der Waals materials,” he says. “Furthermore, molecular simulations based on this accurate and reliable method could enable materials design, enabling the development of drugs and catalysts.”
Solving the riddle of relative stability is undoubtedly a huge step forward. But there is still work to be done and the pace will certainly pick up!