Crystalline Defects in Silicon

Crystalline Defects in Silicon

Material Name: Silicon
Record No.: 8
Primary Chemical Element in Material: Si
Sample Type: Wafer
Uses: Etching
Etchant Name: None
Etching Method: Wet etching
Etchant (Electrolyte) Composition: We have used P-type Boron doped mc-Si wafers of 10 x 10 cm2 in dimension and of about 1 Ohm x cm in electrical resistivity. They were sawed from ingots grown by the Heat Exchanger Method (HEM). In order to remove the sawing process damage, we have begun by thinning and polishing these wafers. During this step, we have used an acidic polishing solution (known as “CP4 Etch”) made by mixing nitric acid (HNO3), acetic acid (CH3COOH) and hydrofluoric acid (HF) with respectively 50%, 30% and 20% concentrations. After 6 min of etching, we rinsed thoroughly the mc-Si wafers with deionized water and dried them under a nitrogen gun. In order to test the Secco and Yang solutions, samples were cut from a polished mc- Si wafer. Mainly, the etching time and the agitation mode were varied. Before each delineation trial, the samples were immersed in diluted HF (10%) solution for 30 seconds in order to remove the native silicon dioxide (SiO2) and then rinsed in deionized water. The Secco formulation is HF/potassium bichromate (K2Cr2O7)/H2O, obtained by mixing 2 parts of HF with 1 part of K2Cr2O7/H2O at (0.15Moles) or (44 grams of K2Cr2O7 in 1litre of H2O). The Yang formulation is HF/chromic acid (CrO3)/H2O, obtained by mixing 1 part of HF with 1 part of CrO3/H2O at (1.5 Moles) or (150 grams of CrO3 in 1 litre of H2O). After the Secco or Yang etching process the samples were immediately rinsed in deionized water and nitrogen dried. Subsequently, SEM observation and other analysis have been performed.
Procedure (Condition): No data
Note: The results of Yang and Secco delineation studies confirm the revelation of crystalline defects for an immersion time from 1 to 2 minutes, by clearly delineating dislocations, twins, grain boundaries and dislocation lines. We observed that dislocation pits etched with the Yang solution have mainly triangular or quadratic forms, whereas the dislocation pits are circular when using the Secco solution. Figure 2 is an illustration of Secco defect delineation process under the optimized conditions.

In order to compare the action of dislocation localization between Secco and Yang solutions, we carried out a Secco revelation on a sample previously revealed with Yang Etch and vice-versa. The aim was to enable us to make a choice between Secco and Yang Etches for the calculation of the maximum density of defects. These tests showed that the action of Secco is higher than Yang’s and lead us to choose Secco for the calculation and mapping of dislocation density. The profile of etched dislocation pits revealed by this solution is shown on Figure 3. The dipping time of the Secco delineation process was fixed at 5 minutes for the following defect mapping study. Such dipping time removes 5 ìm of the dislocation zone as shown by the SEM micrograph of Figure 3. This fine knowledge and control of crystalline defect decoration by “Secco Etch” on mc-Si wafers were directly applied to the development of defect detection by the sheet resistance variation technique.

Once the sheet resistance mapping of a polished mc-Si wafer (10 x 10 cm2) was completed, we submitted the wafer to a Secco etch during 5 minutes. We chose this revelation time in order to strongly mark the defected zones and thus obtain an appreciable variation of sheet resistance. The mc-Si wafer revealed in this way was then precisely placed under the 4 probes tester, and measurement of the sheet resistance was carried out at the initially selected positions. We plotted the mapping of sheet resistance variation (delta R) on the 25 selected points of the mc-Si wafer. The Figure 4 shows the layout obtained for delta R. The next step was the superposition of the physical image of decorated mc-Si wafer with that of the delta R mapping. In order to accomplish this, we have taken a digitalized photo of the whole Secco etched wafer surface, scaled it, and finally successfully superposed it to delta R mapping. The Figure 4 illustrates this original result. The SEM analysis of these results allowed us to make the first classification for delta R bands according to the revealed crystalline defect type and average dislocation density. This classification is summarized on Table 1.

Reference: Mohamed FATHI, Ahmed CHIKOUCHE, New Method for Quality Evaluation of Mc-Si Wafers Implied in the Fabrication of Photovoltaic Cells, Jordan Journal of Mechanical and Industrial Engineering, Volume 4, Number 1, Jan. 2010, pp. 151 - 154.

Figure 2: Secco defects delineation on mc-Si.

Figure 3: Cross section of dislocation delineated by Secco Etch for 5 minutes.

Figure 4: Superposition of the physical image of the defects area with delta R mapping.

Table 1: Classification of defects types according to sheet resistance variation bands.