Epitaxial growth of silicon whiskers without tapering at the base

Development of Growth Theory for Vapor–Liquid–Solid Nanowires: Contact Angle, Truncated Facets, and Crystal Phase

“…Usually, the initial contact angle of the droplet resting on a substrate surface is smaller than that for a developed nanowire and hence the bottom part of nanowire tapers to transfer one geometry to the other. 7,11,15,16 Most droplets observed on tops of long nanowires have the contact angles of 90°or larger. 7,10−14 However, it is poorly known why and how this tapering goes and when it stops, partly due to the mentioned uncertainty in the steady state contact angle.…”

“…The existing models for the initial nanowire shapes are based either on fully isotropic geometries without crystal facets 15 or equilibrium thermodynamic considerations. 16 The VLS growth start is also important for understanding the incubation times that are required to nucleate anisotropic nanowires and critically influence the collective properties within the resulting nanowire ensembles (such as the length distributions). 17 Furthermore, the Glas condition for the triple phase line nucleation, 7 which is also the necessary condition for the WZ phase formation in III−V nanowires with planar growth interfaces, is identical to the Nebol'sin−Shchetinin condition for stable VLS growth of cylindrical nanowires.…”

“…The energetically preferred growth configurations of the droplet with respect to the nanowire top are highly sensitive to the contact angle. − ,− It also enters the conditions governing the crystal phase of III–V nanowires, in both planar and nonplanar geometries of the growth interface (in the latter case, this is simply related to whether the top facet is truncated or not). Quite surprisingly, almost all growth theories developed so far treat the contact angle as a free parameter − that depends, for example, on the gallium flux in Au-catalyzed VLS growth of III–V nanowires .…”

“…Another issue is related to the very early stage of the VLS process. Usually, the initial contact angle of the droplet resting on a substrate surface is smaller than that for a developed nanowire and hence the bottom part of nanowire tapers to transfer one geometry to the other. ,,, Most droplets observed on tops of long nanowires have the contact angles of 90° or larger. ,− However, it is poorly known why and how this tapering goes and when it stops, partly due to the mentioned uncertainty in the steady state contact angle. The existing models for the initial nanowire shapes are based either on fully isotropic geometries without crystal facets or equilibrium thermodynamic considerations .…”

“…Usually, the initial contact angle of the droplet resting on a substrate surface is smaller than that for a developed nanowire and hence the bottom part of nanowire tapers to transfer one geometry to the other. ,,, Most droplets observed on tops of long nanowires have the contact angles of 90° or larger. ,− However, it is poorly known why and how this tapering goes and when it stops, partly due to the mentioned uncertainty in the steady state contact angle. The existing models for the initial nanowire shapes are based either on fully isotropic geometries without crystal facets or equilibrium thermodynamic considerations . The VLS growth start is also important for understanding the incubation times that are required to nucleate anisotropic nanowires and critically influence the collective properties within the resulting nanowire ensembles (such as the length distributions) …”