Trench-edge defects formed during epitaxial recrystallization of trench-bounded amorphized silicon regions are examined as a function of Si substrate crystal orientation. In Si (001), rectilinear features having edges aligned with the crystal’s in-plane ⟨110⟩ directions recrystallize leaving trench-edge defects along all trench edges, whereas the identical features in Si (011) recrystallize without trench-edge defects along trench edges parallel to the crystal’s in-plane ⟨100⟩ direction and with trench-edge defects along trench edges parallel to the crystal’s in-plane ⟨110⟩ direction. The positions and lateral extent of these trench-edge defects suggest that their source is defective epitaxy on slow-growing {111} planes formed during recrystallization. A heuristic model proposed to explain the formation of these {111} planes correctly predicts the essentially defect-free recrystallization seen for rectilinear features in Si (001) having edges aligned with the crystal’s in-plane ⟨100⟩ directions, but cannot completely account for the distinctively curved growth fronts sometimes seen at intermediate stages of recrystallization.
The shapes of the recrystallization fronts observed during solid phase epitaxy (SPE) in line-shaped amorphized Si regions in single-crystal (001) and (011) Si are shown to exhibit a complex time evolution inconsistent with simple models in which regrowth behavior is determined exclusively by the competition between fixed-growth-rate lateral and vertical SPE. We find that the main features of our data for ⟨110⟩-aligned lines may be explained by the propensity of the lateral and vertical growth fronts to form stable interfaces at their edges. These same interfaces have been previously implicated as the root cause of the trench-edge defects produced during the recrystallization of patterned regions bounded laterally by oxide-filled trenches. An extension of a nanofacet model recently developed to explain the trench-edge defects is shown to explain many of the observed recrystallization behaviors.
1 April 2020 in The Journal of the Acoustical Society of America
This paper provides an alternative formulation of a transfer matrix method (TMM) long used for modeling the passive resonances of wind instruments. In the conventional, impedance-based TMMs, the quantities being evaluated are the acoustic pressure P and the acoustic flow U; in the present pressure-based TMM (a variation of the Heavens matrix method developed for thin film optics), the quantities being evaluated are the amplitudes of the forward-going and backward-going pressure waves, and . Power spectra ( vs frequency) for several types of branched and non-branched open-input resonators are computed and then compared to experimental spectra obtained from a Fourier analysis of the acoustic pressure at a position inside the resonator recorded during excitation by a localized sound source placed just outside the resonator. Agreement is good for both one-end-closed polyvinyl chloride chimney pipes and a modern flute set up with a variety of lip, keywork, and open/closed tonehole configurations, although fits for very short straight tubes appeared to require unphysically high wall losses. Input impedance derived from the computed values of and at the flute's embouchure hole appears consistent with directly measured input impedance data in the literature.
The musical notes produced by recorders and other flue instruments consist primarily of spectral components with frequencies given approximately by nf1, where n is an integer and f1 is the fundamental frequency. However, the real tones of these instruments contain other spectral components that have been observed and discussed by a number of authors. We report a study of spectral components in the tones produced by recorders that are odd half-integer multiples of f1, i.e., spectral components with frequencies n±12f1. Our results, obtained through a combination of experimental and simulation studies of soprano recorders, suggest that these components are associated with the air flow in the vicinity of the window region and the labium edge. We also show that these half-harmonics can be suppressed by modifications of the instrument that alter the pattern of air flow in those regions. Speculations concerning the importance of the half-harmonics and the degree to which they are perceptible by a listener are briefly discussed.