Process and device simulations aimed at improving the emitter region performance of silicon PERC solar cells

Abstract

Achieved levels of Silicon-based passivated emitter and rear cell (PERC) solar cells' laboratory and module-level conversion efficiencies are still far from the theoretically achievable Auger limit of 29.4% for silicon solar cells, prominently due to emitter recombination and resistive losses. The emitter region in PERC devices is formed by using either ion implantation followed by a diffusion process or POCl3 diffusion. In ion-implanted emitter-based PERC, the process variables such as dose, energy, diffusion time, and temperature play a vital role in defining the characteristics of the emitter region. Detailed investigation of these parameters could provide a pathway to mitigate the recombination as well as resistive losses; however, it requires a considerable budget to optimize these parameters through a purely experimental approach. Therefore, advanced industrial standard process and device simulation are perceived in this work to carry out the comprehensive study of process variables. Investigation of ion implantation and diffusion process parameters on the PV performance of an upright pyramid textured, industrial standard stacked dielectric passivated PERC solar cell is carried out to deliver 22.8% conversion efficiency with improved PV parameters such as short circuit current density (JSC) of 40.8 mA/cm2, open-circuit voltage (VOC) of 686 mV, and fill-factor (FF) of 81.54% at optimized implantation and diffusion parameters, such as implantation dose of 5×1015 cm-2 with energy 30 keV followed 950 oC diffusion temperature and 30 min of diffusion time. The performance of the optimized PERC device is compared with already published large area screen printed contact-based device. This work may open up a window for the experimental work to understand the influence of process parameters on the emitter region to develop the highly efficient PERC solar cell in the future.

Double POLO Carrier Selective Contact Based PERC Solar Cell for 25.5% Conversion Efficiency: A Simulation Study

“…Afterward, an n + emitter has been formed on the front surface with the help of ion implantation and diffusion process. Different processes such as etching, material deposition of stacked layers, front and rear metallization have been performed (25). Then, an optimized dielectric passivation stack of SiNX (70 nm)/SiO2 (10 nm) for front Si surface passivation and Al2O3 (10 nm)/SiNX (100 nm) stacked materials for rear surface passivation are used.…”

“…Different tunneling parameters such as QTNLSC.EL, QTNLSC.HO, QTNLSC.BBT and QTNL.DERIVS are invoked in this work to perform the charge carriers (electrons/holes) and band-to-band tunneling (27). Also, different models such as the Schenk bandgap narrowing model (28), Richter model for Auger recombination (29), and Klaassen mobility model (30) as per the published work have been employed (22,25). The performance of the double-POLO PERC device is accomplished by varying the thickness of the tunnel oxide layer on the front surface.…”

Design and parametric optimization of ion-implanted PERC solar cells to achieve 22.8% efficiency: a process and device simulation study

“…The diffusion process was allowed to run for 30 minutes at a temperature of 950 C, and all the process parameters for the ion-implanted emitter were optimized in a previous work. 36 As a result, the p-n junction was formed, as depicted in Fig. 1.…”

“…[37][38][39] Further, the transient activation model was utilised to compute activated dopants that were accommodated into the substitutional lattice out of the total chemical concentration of the implanted phosphorous. 36 The excess dopants that fail to unite are considered to be deactivated. The point at which this occurs at a particular temperature for a particular semiconductor is known as the solid solubility limit.…”

“…The Newton method was chosen as the numerical calculation method for this work. To simulate the device, various models were used as per the published literature, 36,45 and the iterative process for the PERC device was completed using various models, as summarized in Table 1.…”