Characterization of laser edge isolation in multicrystalline silicon solar cells

Abstract

We report on a characterization study of laser edge isolation in multicrystalline silicon (mc-Si) solar cells using microscopic electrical, structural, and morphological tools of scanning capacitance microscopy (SCM), conductive atomic force microscopy (C-AFM), electron backscattering diffraction (EBSD), and scanning electron microscopy (SEM), as well as a macroscopic electrical characterization of lock-in thermography (LIT). SCM and C-AFM measurements revealed that the emitter was not completely removed by the laser ablation, and considerable amounts of emitter dopant were driven into the material. A portion of the ablated or molten material was redeposited or recrystallized on top of the laser groove, forming either single-or polycrystalline stripes. Si particles with either polycrystalline or amorphous structures were also formed on the grooves. LIT measurement on a shunted device exhibits a high temperature region centered on the groove line, indicating inadequate isolation. SEM observations show a significant different morphological/structural surface of the groove from that of the isolated devices. These techniques provide useful characterizations for failure analysis of the laser edge isolation.

Manufacturing metrology for c-Si module reliability and durability Part II: Cell manufacturing

“…Resonance ultrasonic vibration (RUV) [37,38] Scanning acoustic microscopy (SAM) [37,38] Transmission electron microscopy [23,41] Angle polishing followed by defect etching [42] Insufficient saw damage removal Atomic force microscopy (AFM) [23,43] Scanning electron microscopy (SEM) [23,41,44,45] Confocal microscopy [46][47][48] Optical profilometry [18] Compromised mechanical strength or excessive stress…”

“…Pre-breakdown [50] Trench structure or etch pit formation [43] AFM [23,43] SEM [23,41,44,45] Confocal microscopy [46][47][48] Color vision system [43] Dark lock-in thermography (DLIT) [41,50] Electroluminescence (EL) imaging [41,48,50] Light or dark current-voltage (I-V) [49] Electron beam-induced current (EBIC) [ [44,51] Electron backscatter diffraction (EBSD) [44] Scanning conductive microscopy [44] Conductive AFM (C-AFM) [44] Table 1. Failure modes, mechanisms and relevant metrology methods related to wet chemical processes.…”

“…This melting causes an undesired effect known as emitter drive-in, which is a result of laser ablation stimulating P atoms to go further underneath the isolation groove. This is why, although the P-doped n + emitter layer is only about 0.3 μm deep, the grooves made by nanosecond laser ablation have to be about 30 μm deep to overcome emitter drive-in effect [44,51]. Jiang et al [44] also observed that a portion of the ablated or molten material was redeposited or recrystallized, forming a2-4 μm-wide stripe on top of the n + region in the groove.…”

“…This is why, although the P-doped n + emitter layer is only about 0.3 μm deep, the grooves made by nanosecond laser ablation have to be about 30 μm deep to overcome emitter drive-in effect [44,51]. Jiang et al [44] also observed that a portion of the ablated or molten material was redeposited or recrystallized, forming a2-4 μm-wide stripe on top of the n + region in the groove. The stripe formed a raised fringe on the edge of the groove and was doped with P, but featuring different doping concentrations than the original and remaining emitters.…”

“…The stripe formed a raised fringe on the edge of the groove and was doped with P, but featuring different doping concentrations than the original and remaining emitters. There were also some n-type doped particles on top of the recrystallized stripes [44]. Sedao et al [51] used new generation ultrashort picosecond lasers with high pulse repetition rates, which can ablate material while minimizing melting, thereby minimizing the thermal impact of the process.…”