Surfaces and nucleation

An initially singular and atomically flat surface advances only if ledges, possessing kinks, are nucleated and formed on the surface. This can occur if adatoms cluster together on the surface and nucleate new ledges at small pillbox-shaped clusters of adatoms as indicated in Fig. 12.1. The free energy to form such a cluster... 

In-situ growth monitoring, in particular by reflective high energy electron diffraction (RHEED), has provided some fundamental information on the surface and nucleation properties of nitrides. Early RHEED studies by Hughes et al [34] and Hacke et al [49] were completed by Smith et al [50], The observed surface reconstructions for (0001) and (0001) GaN surfaces have allowed the modelling of the quasi-equilibrium surface, which has been calculated to be preferentially Ga-terminated [51], Feuillet et al [52] have followed the evolution of surface lattice constants in RHEED for the nucleation of GaN on AIN or InN on GaN (and vice versa) and extracted a wide range of information on the character of nucleation and misfit relaxation. 

The surface diffusion mechanism, in which only the less noble element (zinc in copper-zinc alloys) dissolves and the remaining more noble metal is rearranged by diffusion on the surface and nucleation of islands of almost pure metal [23]. 

No other processes are discussed here, although they are important also. Among them are adsorption of reactants or products at the electrode surface, and nucleation, i.e. the formation of crystal nuclei, where a new phase is generated by deposition of a solid or by gas evolution. However, only the three specially emphasized processes are really essential. 

This section represents a continuation of Section VII-5, which dealt primarily with the direct estimation of surface quantities at a solid-gas interface. Although in principle some of the methods described there could be applied at a solid-liquid interface, very little has been done apart from the study of the following Kelvin effect and nucleation studies, discussed in Chapter IX. 

Miich effort in recent years has been aimed at modelling nucleation at surfaces and several excellent reviews exist [20, 21 and 22]. Mean-field nucleation theory is one of these models and has a simple picture at its core. 

Scaling is not always related to temperature. Calcium carbonate and calcium sulfate scaling occur on unheated surfaces when their solubiUties are exceeded in the bulk water. Metallic surfaces are ideal sites for crystal nucleation because of their rough surfaces and the low velocities adjacent to the surface. Corrosion cells on the metal surface produce areas of high pH, which promote the precipitation of many cooling water salts. Once formed, scale deposits initiate additional nucleation, and crystal growth proceeds at an accelerated rate. 

Models used to describe the growth of crystals by layers call for a two-step process (/) formation of a two-dimensional nucleus on the surface and (2) spreading of the solute from the two-dimensional nucleus across the surface. The relative rates at which these two steps occur give rise to the mononuclear two-dimensional nucleation theory and the polynuclear two-dimensional nucleation theory. In the mononuclear two-dimensional nucleation theory, the surface nucleation step occurs at a finite rate, whereas the spreading across the surface is assumed to occur at an infinite rate. The reverse is tme for the polynuclear two-dimensional nucleation theory. Erom the mononuclear two-dimensional nucleation theory, growth is related to supersaturation by the equation. 

Satisfactory performance is obtained with tubes having helical ribs on the inside surface, which generate a swirling flow. The resulting centrifugal action forces the water droplets toward the inner tube surface and prevents the formation of a steam film. The internally rifled tube maintains nucleate boiling at much higher steam temperature and pressure and with much lower mass velocities than those needed in smooth tubes. In modern practice, the most important criterion in drum boilers is the prevention of conditions that lead to DNB. 

Introduction of the surface-nucleation mechanism in numerical computation of elastic-plastic wave evolution leads to enhanced precursor attenuation in thin specimens, but not in thicker ones. Inclusion of dislocation nucleation at subgrain boundaries indicates that a relatively low concentration of subgrain boundaries ( 2/mm) and nucleation density (10"-10 m ) is sufficient to obtain predicted precursor decay rates which are comparable to those obtained from the experiments. These experiments are only slightly above the threshold necessary to produce enhanced elastic-precursor decay. 

Fig. 7.5. Nucleation in solids. Heterogeneous nucleotion con take place at defects like dislocations, grain boundaries, interphase interfaces and free surfaces. Homogeneous nucleation, in defect-free regions, is rare.

Fig. 13. Flypothetical growth process of SW tubes from a metal/carbon alloy particle (a) segregation of carbon toward the surface, (b) nucleation of SW tubes on the particle surface and, (c) growth of the SW tubes.

J. W. Evans, T. R. Ray. Interface propagation and nucleation phenomena for discontinuous poisoning transitions in surface reaction models. Phys Rev E 50 4302 314, 1994. 

Figure LI Steam generation from a heated surface, showing nucleate boiling, leading to bubbly, intermediate, annular and mist flow forms of convective boiling. Steam bubbles in water (a) leading to water droplets in steam (b).

Some limitations of optical microscopy were apparent in applying [247—249] the technique to supplement kinetic investigations of the low temperature decomposition of ammonium perchlorate (AP), a particularly extensively studied solid phase rate process [59]. The porous residue is opaque. Scanning electron microscopy showed that decomposition was initiated at active sites scattered across surfaces and reaction resulted in the formation of square holes on m-faces and rhombic holes on c-faces. These sites of nucleation were identified [211] as points of intersection of line dislocations with an external boundary face and the kinetic implications of the observed mode of nucleation and growth have been discussed [211]. 

X= 2) or (P = 0, X = 3) and the distinction between these possibilities is most satisfactorily based upon independent evidence, such as microscopic observations. The growth of compact nuclei inevitably results in the consumption of surfaces and when these outer faces, the sites of nucleation, have been eliminated, j3 necessarily is zero this may result in a diminution of n. The continued inward advance of the reaction interface at high a results in a situation comparable with the contracting volume reaction (discussed below) reference to this similarity was also made in consideration of the Mampel approach discussed above. Shapes of the deceleratory region of a time curves for nucleation and growth reactions and the contracting volume rate process are closely similar [409]. 

---