Simulation and optimization of InGaN Schottky solar cells to enhance the interface quality
Introduction
Over the past few decades, III-N semiconductors have been widely used in optoelectronics [[1], [2], [3], [4], [5]]. In particular, the Indium Gallium Nitride (InGaN) has been considered one of the pivotal materials for photovoltaic applications [6].
The InxGa1-xN has emerged owing to its bandgap energy that can be adjusted by the Indium content. In essence, the bandgap energy is between 0.68 eV (1800 nm) for xIn = 1 and 3.44 eV (360 nm) for xIn = 0 [[7], [8], [9], [10]], which makes the InGaN a promising material for solar cells [11,12]. Besides, its electrical and optical properties such as high tolerance to radiation [13], high mobility [14] and high absorption coefficient [15], make this material a relevant choice for photovoltaic applications [16]. So, to enhance solar efficiency, the knowledge of InGaN properties is required. The major challenge is to obtain a good crystal quality of InGaN [6]. However, the scientific researchers have reported in Ref. [[17], [18], [19]], the failure to grow a p-InGaN layer because of the high residual density of donors [15]. so that, the application of Ohmic contact on p-InGaN material seems difficult [16]. Moreover, it is notable that the p-n and the p-i-n InGaN is still unable to exceed the photovoltaic efficiencies of both III-Nitride and silicon materials [20].
Accordingly, we propose to replace the p-layer by a Schottky contact (Schottky solar cell) to solve these critical problems. This innovation is widely used for III-Nitride based power devices and photodetectors [14]. Whereas it is exceptionally new for the based-InGaN photovoltaic technology. There have been a few authoritative investigations that reported the realization of functional Schottky solar cells [[21], [22], [23], [24]]. Recently, Sid Ould Saad et al. [25] and Abdoulwahab Adaine et al. [26] have realized the first complete numerical simulations toward that path. Yet, they did not take into account the large band offset issue [19,[27], [28], [29]] and the large lattice mismatch between GaN and InGaN effect [30].
Experimentally, the InGaN layer should be thicker than 400 nm aiming to both ensure an important absorption of photons and high efficiency for solar applications [29]. Nevertheless, thick layers of InGaN involve a high indium concentration that presents a limitation in the experimental process [29]. Indeed, the Indium tends to accumulate in the active and a large valence band offset appears which degrades the performance of the cells [19,29,31]. The Indium content is selected fundamentally by the selection of band offset among GaN and InN (from 0.5eV [32] to 1.1eV [33]. This broad scope is affirmed experimentally and theoretically, which indicates a hard determination of the content where a collection of the carriers and J-V characteristics are influenced [8]. A numerical advanced technique and different physical models are investigated to approach our structure to reality. This study is focused on the effects of (i) piezoelectric field, (ii) the lattice mismatch and (iii) the recombination of photo-generated carriers in the hetero-interface on the Schottky solar cells performances. To overcome these problems, several experimenters have used the grading procedure in the purpose of reducing the dislocations, the lattice mismatch and the surface recombination losses in the investigated solar cells [[34], [35], [36], [37], [38], [39]].
In this study, we have exclusively studied the Schottky n-InGaN/n-GaN solar cell with adding a compositionally graded layer between GaN and InGaN, to enhance the photovoltaic performances and to ameliorate the interfaces. All simulations in this study were performed using Silvaco-Atlas software.
Section snippets
Electrical properties
The electrical properties of InGaN are deduced from a linear interpolation (Vegard's law) between the values of InN and GaN. Except for the gap energy and the electronic affinity, we must add a curvature correction defined by the bowing parameter 1.43 and 0.8, respectively [19]. Vegard's law is expressed as:where A is the parameters cited in Table 1, b is the bowing parameter and x is the Indium concentration.
Table 1 shows bandgap Eg, the effective density of
Simulation part
This work was accomplished through simulation using the Silvaco-Atlas simulator. This software is based on a finite element method and can solve Poisson and Continuity equations using the standard drift-diffusion model. It includes the decoupled Gummel's and the coupled Newton-Raphson methods. Accordingly, the dense mesh should be used in critical areas (the interfaces between the different materials) to get good resolution and convergence. To get the best solar cell performance, our strategy
The first part
First, we have optimized two parameters: The Indium content in InxGa1-xNx (0 < x < 0.75) and the work function of the Schottky contact (5.1 eV < Wf < 6.3 eV), whose curves are normalized to 1. The other parameters such as doping concentration and thickness are fixed.
Fig. 2 illustrates the metal/n-InGaN/n-GaN Schottky solar cell efficiency versus Indium content for different work functions. At first glance, the increase of Indium content initially leads to an increase in solar cell efficiency.
Conclusion
The new Schottky solar cell with a compositionally graded layer is studied using numerical advanced technique and realistic physical models. Our investigation has focused on the band offset effect between n-GaN/n-InGaN and the enhancement of the interface quality. Using parameters extracted from recent published works, we have obtained an important efficiency of 21.69% for the optimal values of Wf = 6.3 eV, xIn = 54%, TInGaN = 0.18 μm and Nd = 2 × 1017 cm−3. Moreover, it is notable that the
CRediT authorship contribution statement
Abderrahim Khettou: Writing - original draft, Conceptualization, Methodology, Software, Data curation. Imen Zeydi: Writing - review & editing. Mohammed Chellali: Supervision, Project administration, Methodology, Conceptualization. Marwa Ben Arbia: Writing - review & editing, Data curation, Conceptualization. Sedik Mansouri: Supervision, Project administration, Methodology, Conceptualization. Hicham Helal: Software, Validation, Formal analysis. Hassen Maaref: Data curation, Investigation,
Declaration of competing interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Acknowledgment
The authors wish to acknowledge the significant contributions of Ms. Mazari Halima, Ms. Kheira Ameur and Mr. Taibi Brahim (University of Sidi Bel-Abbes, Algeria), Mr. Smiri Badreddine (University of Monastir, Tunisia), Mr. Khirouni Kamel (University of Gabes, Tunisia) and the General Directorate of Scientific Research and Technological Development (GDSRTD).
References (85)
- et al.
High-efficiency of AlInGaN/Al (In) GaN-delta AlGaN quantum wells for deep-ultraviolet emission
Superlattice. Microst.
(2016) - et al.
Comparative study of ionic bombardment and heat treatment on the electrical behavior of Au/GaN/n-GaAs Schottky diodes
Superlattice. Microst.
(2019) - et al.
Finite element simulations of compositionally graded InGaN solar cells
Sol. Energy Mater. Sol. Cell.
(2010) - et al.
GaInN/GaN solar cells made without p-type material using oxidized Ni/Au Schottky electrodes
Mater. Sci. Semicond. Process.
(2016) - et al.
Simulation study of a new InGaN p-layer free Schottky based solar cell
Superlattice. Microst.
(2016) - et al.
Optical polarization characteristics of semipolar (303‾1) and (303‾1‾) InGaN/GaN light-emitting diodes
Optic Express
(2013) - et al.
The effect of dislocations on the efficiency of InGaN/GaN solar cells
Sol. Energy Mater. Sol. Cell.
(2013) - et al.
Band gaps of InN and group III nitride alloys
Superlattice. Microst.
(2003) - et al.
Guidelines and limitations for the design of high-efficiency InGaN single-junction solar cells
Sol. Energy Mater. Sol. Cell.
(2014) An electron mobility model for wurtzite GaN
Solid State Electron.
(2005)