Study of the Radiative and Non-Radiative Recombination Processes at Dislocations in Silicon by Photoluminescence and LBI
- PDF / 350,342 Bytes
- 6 Pages / 420.48 x 639 pts Page_size
- 98 Downloads / 173 Views
contamination, the latter seeming to affect only their intensity. As the D1 emission falls in the third window of optical communications and it is still detectable at room temperature [3], the dislocation related luminescence (DRL) could be of potential technological interest, provided its quantum yield (10"6) would be suitably increased. To this scope a large number of investigation were recently addressed to the understanding of the mechanism of light emission and of the nature of the centres responsible of it. In spite of the amount of work done, there is still a controversy about the origin of the radiative and non radiative recombination at dislocations and concerns the intrinsic or extrinsic nature of DRL [18]. Furthermore, while there is in literature a substantial agreement on the origin of the D3 and D4 bands, the origin of the DI e D2 bands is still under discussion. The Dl e D2 lines usually appear in heavily deformed FZ and CZ silicon samples as well as in the case of misfit dislocations in Si-Ge samples and their intensity increases with the deformation or with the increase of the dislocation density. About the controversy on the origin of the DI e D2 bands, Suezawa and Sumino[9] attributed these bands to irregular parts of dislocations and suggested that they can originate from dislocation interaction. S.A. Shevchenko et al.[1] concluded, instead that they could come from dislocation defects, like kinks, jogs, short constrictions, without neglecting impurity states. This last conclusion is supported by the works of Higgs [4-5], which show that the D1 and D2 luminescence appears only on slightly metallic impurity decorated dislocations in FZ silicon and showed a straight dependence of DRL on transition metal contamination The detailed 117 Mat. Res. Soc. Symp. Proc. Vol. 588 ©2000 Materials Research Society
mechanisms of the effect of impurities on electrical and optical properties is however not clear up to now, also because it is very difficult to prepare a truly clean dislocation array and contradictory evidences could be found in literature about impurity effects on DRL [5-7]. In order to go more insight on the problem, we started a systematic work on intentionally dislocated material aimed at understanding the properties of the dislocation luminescence. EXPERIMENTAL The samples were cut from dislocation- free B- and P - doped CZ grown 4" diameter silicon wafers. The initial content of oxygen was 20 ppma. Dislocation sources are nucleated first by scratching the surface of the samples along the direction with a diamond tip loaded with 0.3 N weight. Then the samples are elastically bent at room temperature in the cantilever mode along the transversal axis . Finally the samples are heated under stress for lh at T = 650 0 C using a quartz and ultrapure graphite deformation
apparatus in pure argon atmosphere. The temperature of 650 TC has been selected after having shown that in this condition the segregation of oxygen takes a minimum value [10]. A getter trap was used to remove residual water and ox
Data Loading...
