Effect of edge junction isolation on the performance of laser doped selective emitter solar cells Brett Hallam a,n , Stuart Wenham a , Haeseok Lee b , Eunjoo Lee b , Hyunwoo Lee b , Jisun Kim b , Jeoungeun Shin b a Photovoltaics Centre of Excellence, University of New South Wales, Sydney, NSW 2052, Australia b R&D Center, Solar Cell Division, Shinsung Solar Energy, Bundang, Seongnam-si, Gyeonggi-do 463-420, Republic of Korea article info Article history: Received 2 July 2011 Accepted 2 September 2011 Available online 19 September 2011 Keywords: Edge junction isolation Laser doping Light induced plating Photoluminescence imaging abstract The effect of laser and chemical edge junction isolation on electrical performance of industrially manufactured laser doped selective emitter solar cells with light induced plated n-type contacts is investigated in this work. Directly after the formation of the aluminium back surface field, photo- luminescence images indicates that laser edge junction isolation causes substantial damage around the perimeter of the cell, extending several millimeters from the laser edge isolation groove. On finished devices, regions of high series resistance are evident around the perimeter, caused by parasitic plating nucleating in the damaged laser grooved region which induce shunting and inhibits further plating taking place in the surrounding regions. The use of chemical edge junction isolation eliminates both of these issues and can result in efficiency gains of more than 2% absolute compared to that fabricated using laser edge isolation, suggesting a far superior method of edge junction isolation for the industrial manufacture of laser doped selective emitter solar cells with light induced plated contacts. & 2011 Elsevier B.V. All rights reserved. 1. Introduction 1.1. Laser doped selective emitter solar cell In recent years, several research groups have demonstrated non-confirmed efficiencies on laser doped selective emitter solar cells between 18.5% and 19% on large area devices using standard commercial grade boron-doped CZ silicon substrates [1–3], whilst Shinsung holdings have achieved independently confirmed efficien- cies as high as 19.2% on large area devices [4]. The laser doping process was first proposed by Wenham and Green [5], and overcomes many of the fundamental limitations of the conventional screen printed solar cell technology. The laser doping process simultaneously patterns the SiN dielectric layer and incorporates dopants into the underlying silicon, creating a heavily doped region to allow low contact resistance whilst minimising the metal/Si interface. This allows for a selective emitter structure to avoid the loss of current due to a poor short wavelength response of standard screen printed solar cells by increasing the sheet resistance of the emitter above 80 O=& and enable higher voltages on finished devices. Laser doping also allows a self-aligned metallisation scheme compatible with the use of light induced plating (LIP). The LIP process utilises the solar cell itself as a power source under illumination, to plate the front metal contacts of the device. This process further enhances the performance of the LDSE solar cell in comparison to their screen printed counterparts by allowing the formation of fine line widths on the front of the solar cell of less than 30 mm. As a result, fingers can be spaced more closely together whilst generating less shading losses than conventional screen printed solar cells and hence allow increased short circuit current densities. Fig. 1 shows the cell structure of the LDSE solar cell. 1.2. Edge junction isolation Typically, plasma edge junction isolation has been used for the manufacture of screen printed solar cells in commercial produc- tion. However there are several weaknesses to the plasma edge junction isolation technique. One such weakness is that the wafers are required to be stacked, hence it is not an inline process which disrupts the flow of the production line and requires more wafer handling. This increased wafer handling, along with the mechanical stresses induced on the wafers during the plasma process can adversely affect yields. The plasma process can also remove some of the SiN around the edge of the cell if the edge isolation is performed after the deposition of the antireflection coating, leading to further damage through the absence of surface passivation around the perimeter of the cells. This method of edge junction isolation also fails to remove the emitter off the rear side of the cells [7]. Contents lists available at SciVerse ScienceDirect journal homepage: www.elsevier.com/locate/solmat Solar Energy Materials & Solar Cells 0927-0248/$ - see front matter & 2011 Elsevier B.V. All rights reserved. doi:10.1016/j.solmat.2011.09.001 n Corresponding author. Tel.: þ61 2 9385 5246; fax: þ61 2 9662 4240. E-mail address: brett.hallam@unsw.edu.au (B. Hallam). Solar Energy Materials & Solar Cells 95 (2011) 3557–3563