Tem and Ebic Investigations of Polycrystalline Silicon Sheets Grown by the Rad Growth Process

Abstract

Oans cet article, sont prBsentBs des rBsultats d'une Btude par microscopie electronique en transmission de joints de grains Blectriquement actifs (et inactifs) dans des Bchantillons de silicium polycristallin 6laborBs par la methode RAD. Les joints de grains dans les dchantillons extraits aprhs plusieurs metres de tirage ne sont pas caractCrisBs par des relations de macles parfaites. De nombreux prdcipitbs apparaissent dans les joints proches de la surface libre.

Impurity-defect interaction in polycrystalline silicon for photovoltaic applications. The role of hydrogen

“…5) the recombining boundaries are also boundarieŝ e a segregation of carbon is observed. an electron microscopy investigation of the material, Sharko [10] has shown that carbon be precipitated as C-Si microparticles in some i grain-boundaries (mostly boundaries with a 1 deviation to coincidence). Both our and Sharko's results agree to prove that carbon seg] tion or precipitation is responsible for the min carrier recombination behaviour of these g boundaries.…”

Transmission electron microscopy studies in relation to production of solar cells from polycrystalline silicon sheets (R.A.D. process)

“…TEM investigations on type-d samples (Si layers separated by a burn-off at 1,020"C in oxygen) reveal that dislocation arrays have a very complex structure due to plastic deformation (Sharko, 1984); at the same time, precipitation occurs on dislocation cores. Under local internal stresses the pinned dislocations are bowed in and out of their glide plane (Fig.…”

On the origin of the electrical activity in silicon grain boundaries

“…parallel with the bicrystals studies, experiments were carried out on large grained Si polycrystals for solar cells, with both electrical and structural characterization performed on the same zone of the same GB. The possible means include LBIC or EBIC on the .one hand and TEM on the other hand [64, [70][71][72][73][74][75][76]. This kind of investigations permitted to establish, in silicon, that general GBs and subgrain boundaries are more active than coherent (low-1:) ones [70][71][72], and that the {111} 03A3 = 3 twin is not active, unless it is decorated [64,[71][72][73][74].…”

“…The possible means include LBIC or EBIC on the .one hand and TEM on the other hand [64, [70][71][72][73][74][75][76]. This kind of investigations permitted to establish, in silicon, that general GBs and subgrain boundaries are more active than coherent (low-1:) ones [70][71][72], and that the {111} 03A3 = 3 twin is not active, unless it is decorated [64,[71][72][73][74]. Moreover, Silvain could associate the electrical activity of sub-GBs to given types of dislocations [75], while Dianteill and Rocher could show that the electrical activity, in the 2 = 9 twin, depended strongly on the GB plane orientation, varying from 0 for the twin plane to a maximum for a random plane [76].…”

“…PRECIPITATES. -Carbon and oxygen have also been found under precipitated form : SiC precipitates were observed in GBs in TEM[73,96] ;…”