Metal flux methods are excellent for synthesizing high-quality hexagonal boron nitride (hBN) crystals, but the atomic mechanisms of hBN nucleation and growth in these systems are poorly understood and difficult to probe experimentally. Here, we harness classical reactive molecular dynamics (ReaxFF) to unravel the mechanisms of hBN synthesis from liquid nickel solvent over time scales up to 30 ns. These simulations mimic experimental conditions by including relatively large liquid nickel slabs containing dissolved boron and a molecular nitrogen gas phase. Overall, the reaction takes place almost exclusively on the surface of the liquid nickel, owing to the low solubility of nitrogen in bulk nickel and the intermediate species' preference for the metal-gas interface. The formation of hBN invariably begins by reaction of dinitrogen with nickel-solvated boron atoms at the surface, forming intermediate N-N-B species, which typically evolve into B-N-B units through a short-lived intermediate where a single nitrogen atom is coordinated by one nitrogen and two boron atoms. The resulting B-N-B units, in turn, coalesce with growing hBN nuclei and carry nitrogen between hBN nanocrystals in an Ostwald ripening process. The amount of hBN produced on the tens of nanosecond time scale depends critically on the boron concentration, while having a much weaker dependence on the N pressure for the regime considered (N pressures of 2.5-10 MPa, Ni-B solutions with 6-12% boron by atom fraction). The highest rate of hBN formation occurs at the lowest temperature considered (1750 K, just above the melting point of nickel), while no hBN sheets are formed at 2000 K or above. An analysis of the transition pathways for nitrogen atoms shows that the final step, incorporation of small B-N motifs into larger hBN sheets, is the rate-limiting step in the regimes considered. While raising the temperature from 1750 to 2000 K has little effect on the formation of intermediates (N-N-B, B-N-B, etc.), the lack of large hBN sheets at temperatures >1900 K is explained by decreased probability of the final step and increased probability of breakup of hBN into B-N motifs.
Download full-text PDF |
Source |
---|---|
http://dx.doi.org/10.1021/acsami.4c16991 | DOI Listing |
Enter search terms and have AI summaries delivered each week - change queries or unsubscribe any time!