Beside dust and metallic impurities, dopants and carbon affect strongly the efficiency of silicon solar cells (Ceccaroli and Lohne, 2006). The substitutional carbon contamination can be determined in polysilicon samples (Hwang et al., 1991) and the resistivity of multicrystalline silicon can also be evaluated (Gosh et al., 1982; Seager, 1985; Tyagi and Sen, 1983).[10] However, most of the specified dopants and substitutional carbon are measured on monocrystalline silicon grown from polysilicon samples after FZ crystal growing (ASTM F574, n. d.; Bourbina et al., 1994; SEMI MF1708, n. d.; SEMI MF1723, n. d.; SEMI PV17-0611, n. d.).[11] Segregation must be considered because it affects the distribution of impurities along the crystal specimen and the optimal position ofits sampling (Freiheit et al., 2007).
Dopant concentration can be determined using resistivity measurement (Lowry et al., 2009; SEMI MF397, n. d.; SEMI MF84, n. d.) combined with low temperature Fourier-transform infrared spectroscopy (LT-FTIR; SEMI MF1528, n. d.) or photoluminescence (LT-PL) on the monocrystalline slugs (SEMI MF1389, n. d.).[12] Dopants can be also analyzed on-line during CVD of monocrystalline silicon without growing an FZ ingot (Yoji et al., 1991).
Alt reported on the precise analysis of substitutional carbon (Alt et al., 200 ) and nitrogen (Alt, 2007). After a measurement of the carbon level at room
temperature, the reference sample must be less than 2 x 1015 atoms/cm3 (0.04 ppma) at about 610 cm – 1(SEMI MF1391, n. d.). At this wavenumber, interfering Si-Cl peaks can also be assumed (Chevallier et al., 1983; Guirgis et al., 2009; Hiroshi et al., 1998).
