Ab initiostudy of atomic hydrogen diffusion on the clean and hydrogen-terminated Si(001) surface

Abstract

Recent experiments have shown an unexpected diffusion behavior of hydrogen on the Si͑001͒ surface at high temperatures and high coverages. To shed some light on this behavior, we have employed densityfunctional theory to investigate H diffusion on the flat Si͑001͒ surface for different coverages with main emphasis on the high-coverage limit of Si͑001͒ monohydride. Three basic diffusion steps, intradimer, intrarow, and interrow have been studied both for isolated H atoms on the clean Si͑001͒ surface, as well as for isolated and paired H vacancies on the Si͑001͒ monohydride surface. The barrier energies depend strongly on the distance between the two Si neighbors of the diffusing H atom in the transition state. We observe that an isolated vacancy is less mobile than an isolated H atom showing that the Si͑001͒ monohydride surface is more rigid than the clean surface. Interestingly, two adjacent vacancies may transfer dangling-bond charge from one to another prior to a transition of one of them, which significantly lowers the transition barrier. We visualize the reaction mechanisms using maximally localized Wannier functions and we discuss hopping rates within the harmonic approximation to transition state theory in comparison with experimental data.

Extent of hydrogen coverage of Si(001) under chemical vapor deposition conditions from ab initio approaches

“…9,10,11 Likewise, computational studies have been devoted to desorption mechanisms and surface dynamics. 12,13,14 Two studies using temperature programmed desorption (TPD) in UHV conditions show an onset of H2-desorption from hydrogenated Si(001) surfaces to occur between 700-780 K 15 and 750±25 K 16 , respectively. Brückner et al studied the desorption of H2 from hydrogenated Si(001) both in a flow of hydrogen and nitrogen carrier gas with reflectance anisotropy spectroscopy (RAS) under MOVPE conditions.…”

“…The authors thus concluded that the silicon surface is hydrogenated for temperatures below 1070 K in a H2-atmosphere and pristine at 1270 K. Extrapolation from the low pressures used in the study to atmospheric pressure of hydrogen yields  = 0.94 ML coverage at 1070 K and  = 0.25 ML coverage at 1270 K. It should be stressed, that we will focus solely on surface coverage in thermodynamic equilibrium -adsorption and desorption dynamics are not aimed at here and were Page 4 of 21 extensively reviewed elsewhere. [9][10][11][12][13][14] Likewise, the H2 induced formation of reconstruction domains 18 is not the aim of this study.…”

“…9,10,11 Likewise, computational studies have been devoted to desorption mechanisms and surface dynamics. 12,13,14 Two studies using temperature programmed desorption (TPD) in UHV conditions show an onset of H2-desorption from hydrogenated Si(001) surfaces to occur between 700-780 K 15 and 750±25 K 16 , respectively.…”

Energetics and stability of dangling-bond silicon wires on H passivated Si(100)

“…Discrepancies are probably due to the different H coverage of the two systems compared here. Indeed, thorough computations 32 with different H coverages yield a 0.25-eV difference between their computed high H-coverage and low coverage limits.…”

“…These results are in good agreement with previous calculations at different levels of H coverage on Si (001), where a systematic larger intrarow barrier is found compared to the intradimer one. 30,32 Diffusion between rows is about one eV larger than the intrarow diffusion barrier 29,30,32 and we can effectively neglect it as a threat to the DB-wire stability.…”