Electrically Active Defects in Plated Crystalline Silicon n+p Solar Cells: A DLTS Perspective

Abstract

Cost reduction is one of the major driving forces in the technology development of silicon solar cells. This explains the current interest in the implementation of a copper-based metallization, replacing more expensive Ag contacts [1-3]. In order to prevent the in-diffusion of copper in the underlying silicon substrate, a diffusion barrier, like, e.g., a thin nickel layer is usually deposited. Opening of the passivation layer for contact formation by laser ablation (LA), enables a better controll for narrower line features, compared with standard wet etching. In this work, potential issues with defect fomation associated with laser ablation and metal plating will be studied relying on Deep-Level Transient Spectroscopy (DLTS) [4]. P-doped emitters have been fabricated by POCl3 diffusion into 1-3 Wcm p-type Czochralski silicon substrates. Passivation of the solar cells has been achieved by PECVD of a SiNx:H layer. The dielectric layer at the front was opened by ps UV-laser ablation, using a high fluence (1.08 J/cm2) (hard LA). This was compared with wet etching (WE). Approximately 1 mm of nickel was deposited by light-induced plating, followed by 8 to 10 mm of electroplated copper. More details of the processing have been previously reported [5,6]. Diodes with area of 2 mmx2 mm were laser-cut from the finished solar cells for DLTS, using a Fourier-Transform (FT) system. Temperature (T-) scans have been performed at different bias pulses from a reverse bias VR to a pulse bias VP, for different durations tp. The pulse period was tw. Before cooling the sample, Current-Voltage (I-V) and Capacitance-Voltage (C-V) measurements have been performed at room temperature, the latter at a fixed frequency of 1 MHz. Normal I-V and C-Vs have been obtained (Fig. 1), while rather uniform doping profiles were derived from a 1/C2 versus VR plot (Fig. 2), yielding a doping density in the range of 2×1016 cm-3. The spectra for the wet etched sample in Fig. 3 exhibit two main hole traps in the p-type base layer: a peak at about 80 K and another one at ~150 K. The activation energy of the first trap amounts to 0.17 eV and a hole capture cross section sp of ~7×10-15 cm2. Applying a minority carrier injection pulse to forward VP, reveals a clear electron trap at about the same position in Fig. 4, followed by a second one at higher T. It has been shown elsewhere that the 0.17 eV peak most likely corresponds with the substitutional nickel donor level [5]. The shallow electron trap could well be the double acceptor level of Nis at EC-0.07 eV, while the deeper one may correspond with the first acceptor level. The LA sample exhibits the same 80 K hole trap as after wet etching (Fig. 5), indicating that the nickel-silicidation process introduces nickel in the underlying p-type base, irrespective of the etching. A deeper hole trap at about 200 K is observed as well and could be related to dislocation-related states [5,6]. Further proof of that is the observation of slow capture by the nickel donor level in Fig. 6, which follows an ln(tp) dependence. The capture is much slower than can be expected from the hole capture cross section and indicates that Nis may be trapped in the strain field around the dislocations, penetrating the p-type base. Similar observations have been found in the past for Fe decorating dislocations in p-type silicon [8]. It will, finally, be shown that in both cases, no active copper levels have been detected in the p-type silicon base [5,6].