Library

Notes: Fully-coherent Si0.7Ge0.3 layers were deposited on Si(001) by gas-source molecular beam epitaxy (GS-MBE) from Ge2H6/Si2H6 mixtures in order to probe the effect of steady-state hydrogen coverages θH on surface morphological evolution during the growth of compressively strained films. The layers are grown as a function of thickness t at temperatures, Ts=450–550 °C, for which strain-induced roughening is observed during solid-source MBE (SS-MBE) and deposition from hyperthermal beams. With GS-MBE, we obtain three-dimensional (3D) strain-induced growth mounds in samples deposited at Ts=550 °C for which θH is small, 0.11 monolayer (ML). However, mound formation is dramatically suppressed at 500 °C (θH=0.26 ML) and completely eliminated at 450 °C (θH=0.52 ML). We attribute these large differences in surface morphological evolution primarily to θH(Ts)-induced effects on film growth rates R, adatom diffusion rates Ds, and ascending step-crossing probabilities. GS-MBE Si0.7Ge0.3(001) growth at 450 °C remains two dimensional, with a surface width 〈w〉〈0.15 nm, at all film thicknesses t=11–80 nm, since both R and the rate of mass transport across ascending steps are low. Raising Ts to 500 °C increases R faster than Ds leading to shorter mean surface diffusion lengths and the formation of extremely shallow, rounded growth mounds for which 〈w〉 remains essentially constant at (similar, equals)0.2 nm while the in-plane coherence length 〈d〉 increases from (similar, equals)70 nm at t=14 nm to 162 nm with t=75 nm. The low ascending step crossing probability at 500 °C results in mounds that spread laterally, rather than vertically, due to preferential attachment at the mound edges. At Ts=550 °C, the ascending step crossing probability increases due to both higher thermal activation and lower hydrogen coverages. 〈w〉(t) increases by more than a factor of 10, from 0.13 nm at t=15 nm to 1.9 nm at t=105 nm, while the in-plane coherence length 〈d〉 remains constant at (similar, equals)85 nm. This leads, under the strain driving force, to the formation of self-organized 3D {105}-faceted pyramids at 550 °C which are very similar to those observed during SS-MBE. © 2002 American Institute of Physics.

Notes: Oestrogen levels in urine from 21 normotensive and 13 hypertensive pregnant women were moderately correlated (r=0.48) with levels of 3β-hydroxy-5-ene steroids (oestrogen precursors) in urine from their infants. In five infants from otherwise normal pregnancies in which oestrogen excretion was very low, levels of 3β-hydroxy-5-ene steroids were significantly lower than normal while there was no difference between hypertensives and normals. Levels of urinary cortisol metabolites in the infants were moderately correlated with 3β-hydroxy-5-ene steroids (r=0.55) and were especially low in 2 out of 5 infants in the series suffering from distress during delivery. We conclude that subnormal fetal steroidogenesis rather than reduced placental metabolism is the most common cause of low oestrogen excretion of unknown aetiology. A factor in the increased perinatal risk in this group may be an associated insufficient cortisol synthesis by the fetus.

Notes: The growth rates RGe of epitaxial Ge films deposited on Ge(001)2×1 and Si(001)2×1 substrates from Ge2H6 by gas-source molecular beam epitaxy were determined over a wide range of temperatures Ts (300–800 °C) and impingement fluxes JGe2H6(0.1–1×1016 cm−2 s−1). Steady-state RGe(Ts, JGe2H6) curves were well described at both low and high growth temperatures (Ts≤325 °C and Ts(approximately-greater-than)500 °C) using a model based upon dissociative Ge2H6 chemisorption followed by a series of surface decomposition reactions with the rate-limiting step being first-order hydrogen desorption from Ge monohydride for which the activation energy was found to be 1.56 eV. At intermediate temperatures, however, experimental RGe results exhibited a large positive deviation from model predictions due, as demonstrated by temperature programmed desorption measurements and transmission electron microscopy (TEM) observations, to kinetic surface roughening. Extensive (113) faceting resulted in both an increase in the number of active surface sites and higher reactive sticking probabilities. With increased growth temperatures, the facets became more rounded and film surfaces appeared sinusoidal in cross section. The zero-coverage Ge2H6 reactive sticking probability on Ge(001) in the high-temperature flux-limited regime was found to be 0.052, more than two orders of magnitude higher than that for GeH4. In situ reflection high-energy electron diffraction and post-deposition TEM examinations showed that Ge films deposited on Ge(001) at Ts≤325 °C grew in a layer-by-layer mode exhibiting a smooth flat surface. © 1995 American Institute of Physics.

Notes: Si1−xGex layers with x ranging from 0 to 0.30 were grown on Si(001)2×1 substrates at temperatures ranging from 450 to 950 °C by gas-source molecular-beam epitaxy (GS-MBE) from Si2H6 and Ge2H6. In the low-temperature surface-reaction-limited growth regime, the deposition rate RSiGe increases with increasing Ge concentration due to an enhancement in the hydrogen desorption rate resulting in a correspondingly higher steady-state dangling bond density. In the high-temperature impingement-flux-limited regime, where the steady-state hydrogen coverage approaches zero, RSiGe is controlled by the Si2H6 and Ge2H6 reactive sticking probabilities S which decrease with increasing Ge2H6 flux but are not strongly temperature dependent. SSi2H6 and SGe2H6 range from 0.036 and 0.28 on Si(001) to 0.012 and 0.094 during growth of Si0.82Ge0.18 at Ts=800 °C. In both growth regimes, large changes in RSiGe require only modest increases in incident Ge2H6 to Si2H6 flux ratios, JGe2H6/JSi2H6, due to Ge segregation which is strongly coupled to the steady state hydrogen coverage. The Ge to Si ratio in as-deposited films increases linearly, while SGe2H6/SSi2H6 remains constant, with increasing JGe2H6/JSi2H6. Hydrogen desorption and Ge segregation rates, together with Si2H6 and Ge2H6 reactive sticking probabilities, were quantitatively determined from D2 temperature-programmed desorption (TPD) measurements. The combined results from film growth kinetics and TPD studies, together with the assumption of linear superposition, were then used to develop a predictive model, with no fitting parameters, for RSiGe(Ts,JSi2H6,JGe2H6) during Si1−xGex GS-MBE. © 1998 American Institute of Physics.

Notes: B-doped Si(001) films, with concentrations CB up to 1.7×1022 cm−3, were grown by gas-source molecular-beam epitaxy from Si2H6 and B2H6 at Ts=500–800 °C. D2 temperature-programed desorption (TPD) spectra were then used to determine B coverages θB as a function of CB and Ts. In these measurements, as-deposited films were flash heated to desorb surface hydrogen, cooled, and exposed to atomic deuterium until saturation coverage. Strong B surface segregation was observed with surface-to-bulk B concentration ratios ranging up to 1200. TPD spectra exhibited β2 and β1 peaks associated with dideuteride and monodeuteride desorption as well as lower-temperature B-induced peaks β2* and β1*. Increasing θB increased the area under β2* and β1* at the expense of β2 and β1 and decreased the total D coverage θD. The TPD results were used to determine the B segregation enthalpy, −0.53 eV, and to explain and model the effects of high B coverages on Si(001) growth kinetics. Film deposition rates R increase by ≥50% with increasing CB〉˜1×1019 cm−3 at Ts≤550 °C, due primarily to increased H desorption rates from B-backbonded Si adatoms, and decrease by corresponding amounts at Ts≥600 °C due to decreased adsorption site densities. At Ts≥700 °C, high B coverages also induce {113} facetting. © 1997 American Institute of Physics.