arXiv:2611.03460
Surface charges and quantum reflection of hydrogen-isotope beams at grazing incidence
A. Yu. Voronin, V. V. Nesvizhevsky
Abstract
Experiments with cryogenic beams of hydrogen isotopes (hydrogen (H), deuterium (D), and tritium (T)) can be used to study quantum reflection from solid dielectric mirrors. Most dielectric surfaces carry trapped electric charges whose field polarizes the incident atom. We study theoretically how such charges, distributed randomly, in clusters, with compensation over a finite length, or in order, change specular reflection and diffuse scattering of H, D, and T at grazing incidence on typical dielectric mirrors (SiO$_2$/Si and MgF$_2$), for longitudinal velocities of 100--300~m/s and normal velocities of 3--100~cm/s which might become available soon in grazing incidence experiments. Starting from the exact coupled-channel equations, we show that the specular wave feels only the laterally averaged polarization potential, which is fixed by the structure factor of the charges and, for uncorrelated charges, is an inverse-square tail $-C_2/z^2$. The exact solution replaces the linear threshold law $1-\Rr\propto\vperp$ by the fractional power $\vperp^{2ν}$, $ν=(1/4-ζ)^{1/2}$, with the coupling $ζ\propto C_2$; for weak coupling, this appears as a logarithmic velocity dependence, and for $ζ>1/4$, the reflection coefficient oscillates log-periodically. At the surface charge density of $10^{12}$ cm$^{-2}$, absorption increases by up to 26\% for SiO$_2$/Si and by a factor of two to three for MgF$_2$; the effect increases with atomic mass, saturates for compensated charges, and is accompanied by diffuse scattering below $10^{-3}$ of charge-induced loss. We also formulate the condition of complete absorption at the solid surface.
AI Summary
This theoretical study asks how trapped surface charges alter quantum reflection of cryogenic H, D and T beams from dielectric mirrors at grazing incidence. It considers SiO$_2$(2 nm)/Si and MgF$_2$, longitudinal velocities of 100–300 m/s, normal velocities of 3–100 cm/s, and charge densities of $10^{10}$–$10^{12}\,\mathrm{cm}^{-2}$. Starting from coupled-channel equations, it argues that rapid longitudinal motion suppresses coupling out of the specular channel, allowing reflection to be calculated using the laterally averaged polarization potential. This potential is expressed through the charge structure factor; uncorrelated charges produce $-C_2/z^2$. Bessel-function matching gives the threshold law $1-R\propto v_\perp^{\sqrt{1-4\zeta}}$, where $\zeta=2mC_2/\hbar^2$, replacing the neutral mirror’s linear law. Weak coupling yields a logarithmic enhancement. Numerical integration with Lifshitz atom–surface potentials predicts that $10^{12}\,\mathrm{cm}^{-2}$ charges at 3 cm/s increase absorption by 17–26% on silicon and roughly two- to threefold on MgF$_2$. Finite-length charge compensation cuts off the inverse-square tail and restores linear threshold behavior at sufficiently low velocity. Estimated diffuse scattering is approximately $10^{-3}$ or less of charge-induced specular loss. The reduction requires weak individual-charge scattering and sufficient spatial averaging; locally screened charges on conducting substrates do not produce the stated tail. Complete surface absorption is justified using a microscopic-disorder model, but can fail on atomically ordered surfaces. The predicted log-periodic reflection for $\zeta>1/4$ lies outside the stated beam regime. The experiments are prospective, and quantitative predictions depend on the unknown charge distribution.
Error found: The abstract presents the laterally averaged potential as the sole influence on specular reflection, although the derivation neglects a nonzero channel-coupling self-energy under explicit conditions that restrict its applicability, particularly for clustered charges. The claimed strict diffuse-loss bound below 10^-3 of charge-induced absorption is contradicted by the table's value of 1.1×10^-3 for deuterium at 10 cm/s.
Written by gpt-6.1-sol from the paper's full text. Automated: it can be wrong.