Impurities front surface

Second, in the fast-cooling solidification process, the m-c interface is concave to the crystal and the melt flows from the periphery to the center along the solidification front surface. Thus, the impurities in the melt and the SiC particles generated at the interface are transported to and accumulated in the... 

Multicrystalline silicon wafers, with crystal sizes in the range 1-100 mm, ° are currently the main workhorse for the photovoltaics industry. Hydrogen passivation steps used in cell manufacture in recent years, particularly those involved in the deposition of silicon nitride layers on front surfaces, have reduced the impact of electrically active impurities and defects in MC-Si cells and reduced the performance deficit relative to monocrystalline cells. 

Illustrations of velocity vectors for both kinds of these particles are illustrated in Figure 2. This concept has been useful in explaining many drop and bubble phenomena. For example, it has been found that trace amounts of surface-active materials can hinder the development of internal circulation by means of a differential surface pressure. Small bubbles rising slowly are apt to behave like particles with rigid surfaces. This phenomenon can lead to a decrease in k as the age of a bubble increases (Levich. 1956). Larger bubbles, rising more quickly, may sweep their front surface free of trace impurities and therefore escape... 

One can take advantage of the attraction of impurities to dislocation cores to reduce undesired species by intentionally introducing dislocation loops on the back surface of a wafer. This is done by mechanically damaging the back side after the polishing and etching of the front surface is complete. Most detrimental impurities in Si diffuse much faster than dopant atoms. Many are interstitial atoms and move exceptionally fast. Therefore, annealing a wafer with back surface damage collects point defects... 

Before spreading, the surface must be carefully cleaned and freed from contamination. Redistilled water should be used. The liquid surface in front of and behind the float can be swept clean by moving barriers towards the float and sucking away any surface impurities with a capillary joined to a water pump. 

The crack frequently initiates from the breakdown of a craze that formed at an internal defect, as a void or impurity particle. Then, as shown by various investigators, as crack speed increases, the crack jumps rapidly from one craze bulk interface to another and from one craze to another. This can lead to a so-called mackerel type pattern on the fracture surface or to a craze island type structure see also Chapter 1 and 3. As crack length increases and local stress rises, numerous secondary fractures, as shown in Fig. 2 b, are generated ahead of the crack front. 

Based on the experimental observations and the simulation results, we can tentatively identify the following key processes leading to H embrittlement of metals (i) Application of external stress produces local concentration of tensile stress in the vicinity of cracks, which attracts H since H prefers to stay in slightly enlarged interstitial sites. H impurities can also be adsorbed at the crack surfaces near the crack tip the relative ease of H diffusion on the crack surfaces may provide a rapid means of transporting H to the fresh crack front. 

---