Motivation and objectives
With this context favourable to photovoltaic technologies, the motivations presented below led to the establishment of this work.
From 2015 to 2017, some proofs of concept of PK/SHJ tandem solar cells were presented in the literature (–, see part II.3.2 for more details). These devices, involving the series connection of a Perovskite solar cell and a SHJ solar cell, have thus demonstrated PCE exceeding 20 %, on small surfaces, rivalling the silicon single-junction efficiencies. Moreover, several types of architecture (NIP and PIN), texturing (polished/polished, polished/textured and textured/textured) and recombination junctions (TCO and highly doped PN junction) have been proposed in experiments and simulations. Furthermore, those early studies also showed that it is essential to reduce losses to increase the efficiencies, whether optical (due to parasitic absorption of the different layers) or related to the charges transport, in particular in the recombination junction (RJ).
However, even if this RJ between the two subcells seems a key factor, the published studies did not dwell on this subject. In fact, they demonstrated the viability of an ITO layer and a highly doped silicon PN junction in the proofs of concept but without proposing any prospects for characterisation and improvement. This justifies this work, which consists in characterising, understanding and optimising the recombination junction between the subcells, by developing adapted methods.
Therefore, the objectives of this work are as follow:
Development of a highly doped silicon tunnel junction suitable for 2T PK/SHJ tandem solar cells,
Implementation of characterisation methods to optimise the recombination junctions,
Understanding the transport mechanisms in 2T PK/SHJ tandem solar cells and the influence of their RJ.
In this last part of the general introduction, we describe the approach carried out to achieve the objectives of this work.
In chapter II, we present the basics of tandem solar cells and recombination junctions as well as the state-of-the-art of PK/SHJ tandem solar cells. Then, in chapter III, we detail the fabrication processes and the characterisation methods used along this work.
In order to study the two types (previously presented) of RJ during this work, we developed a highly doped silicon PN junction, based on microcrystalline silicon. In chapter IV, we present the development of this RJ for PK/SHJ tandem solar cells in the PIN architecture. Moreover, we detail the microcrystalline silicon layers characterisation and optimisation. In conclusion, we discuss the integration of these layers into SHJ solar cells and their viability for tandem solar cells.
Then, towards the understanding of the transport mechanisms and the optimisation of the RJ, we investigated, set up and discussed ways to characterise electrically the RJ in chapter V. Furthermore, we present advantages and limitations of these characterisation methods, depending of the RJ type and the experimental potential limitations.
Finally, to investigate the influence of the RJ in PK/SHJ tandem solar cells and allow its pairing with the Perovskite top-cell, we perform optical simulations in chapter VI. Moreover, we develop and present an innovative “tandem-like” device that could be very useful in the understanding, development and optimisation of the PK/SHJ tandem solar cells. To conclude, we present the integration of the developed microcrystalline silicon junction in PK/SHJ tandem solar cells and discuss further improvements.
This chapter presents the state-of-the-art of Perovskite/SHJ tandem solar cells and is articulated in four parts. The first part presents a brief history and the basics of tandem solar cells. The second part is an overview of c-Si solar cells and a presentation of SHJ solar cells. Then, the third part describes the recent arrival of Perovskite and Perovskite/SHJ tandem solar cell as a major field of interest in the photovoltaic community. To conclude, the fourth part focuses on the recombination junction, a key part of tandem solar cells that is deeply studied in this work. This chapter not only shows a review of the state-of-the-art technology at the beginning of this work, but also a follow-up of the major advances, in this quickly evolving field of research, during these three years.
Tandem Solar Cells
In this part, we shortly introduce the history of tandem solar cells. Then, we discuss the tandem main configurations, and to conclude we explain the operation of a monolithic tandem solar cell. After the observation of the photovoltaic effect by E. Becquerel in 1839 , the first solar cells elaborated were single-junction solar cells . That is to say that it comes down to an absorber material of N-type or P-type (or even intrinsic) in contact with a layer of the opposite type, forming a PN junction (or a PIN device). In the 1970s, in order to improve the efficiency limit of existing solar cells, the idea of tandem and multi-junction solar cells was proposed . During this period, two possibilities were considered for tandem solar cells: “stacked cells” or a splitting approach , represented respectively in Figure II.1-1 b) and c). Figure II.1-1 shows the three main tandem configurations and Table II.1-1 presents their advantages and disadvantages. In this work, we only studied monolithic tandem solar cell (Figure II.1-1 b)), mainly for its ease of implementation in module and its lower cost. The first monolithic (or two-terminal) tandem solar cell, was made of AlGaAs/GaAs, in 1978, by S. M. Bedair et al.  using a heavily doped tunnel junction between the two subcells.
A monolithic (2T) tandem solar cell is a superposition of two subcells separated by a recombination junction (II.4) or designed such that their superposition forms a recombination junction. The two subcells are connected in series, thus the current density of the tandem is the minimum current density between those of the subcells, equation (II. 1-1). Thus, a current matching between the two subcells leads to the maximum potential current in the 2T tandem solar cell and therefore to a better efficiency. Moreover, in the case of an ideal 2T tandem solar cell, the open-circuit voltage of the tandem is the addition of the subcells open-circuit voltages, equation (II. 1-2). In reality, the recombination junction often induces an additional resistance in the system, R Junction in Figure II.1-2, which leads to losses in the tandem Voc .
As for a single-junction solar cell, the Voc is dependent of the quasi-Fermi levels shown in Figure II.1-3 ( = 1 . ( − ), where q is the elementary charge and Vext is the potential between the two terminals of the solar cell; at Voc: Vext = Voc). If the recombination junction is not optimised, the tandem solar cell will not operate at its full potential. The recombination junction is therefore one of the most important part of the development of tandem solar cells, a focus on it will be done in part II.4.
In this work, we did not study multi-junction solar cells with more than two subcells but the operation theory is the same by adding other subcells and there is always a recombination junction between each subcell. In fact, the more subcells there are, the higher the theoretical efficiency limit is , –. In 2020, J. F. Geist et al. presented a six-junctions solar cell with an efficiency of 39.2 % under 1 sun (47.1 % under 143 suns concentration) , which is very promising but using very expensive III-V materials. Consequently, the vast majority of these technologies are only used for space applications, where the ratio performance/cost is not a key issue. This is not the case for terrestrial applications for massive deployment of the photovoltaic, where tandem or multi-junction solar cells need to demonstrate the lowest LCOE possible to be competitive with silicon-based technologies.
Silicon Heterojunction Solar Cells
After presenting the working principle of tandem solar cells in the previous part, in this section, we relate the history of crystalline silicon solar cells to SHJ cells, as we know them today. Then, we describe in more details the different components of SHJ solar cells.
Figure II.2-1 shows the evolution of the structure of high temperature crystalline silicon solar cells. In the 1940s, Bell Labs discovered that crystalline silicon presents very good properties for photovoltaic applications , . In 1957, they presented a solar cell on N-type crystalline silicon base with a thin P-type layer , Figure II.2-1 a), which exceeded 10 % efficiency . In order to produce solar cells for space applications, P-type c-Si (less sensitive to radiation) quickly replaced the N-type as base for the solar cells and an anti-reflection coating was added to improve optical properties, as well as a front-side metallisation grid, Figure II.2-1 b). In the 1970s, Comsat corporation improved the efficiencies with lowly resistive wafer, highly doped back-side, Figure II.2-1 c) and front-side texturing, Figure II.2-1 d) . Then, the University of New South Wales (UNSW), in Sydney, added a passivation layer (SiO2) and reached 18 % efficiency with their Metal Insulator NP junction (MINP), Figure II.2-1 e). In 1988, they exceeded 20 % efficiency with their Passivated Emitter Solar Cell (PESC), Figure II.2-1 f), by using inverse pyramid structure and partially direct Metal/emitter contact. Then, with a local (n+) doping in the emitter and a partially opened passivation layer at the back-side, their PERC structure, Figure II.2-1 g), reached 22.6 % efficiency in 1989 .
Continuing their momentum, they reached the 25 % efficiency 10 years later with the PERL structure, thanks to a local (p+) doping at the back-side, Figure II.2-1 h) . Fifteen years later, the Fraunhofer ISE developed the TOPCon structure on N-type wafers, Figure II.2-1 i), using a passivating tunnel oxide and a (n+) poly-Si at the back-side . Since then, more and more solar cells with passivated contacts, made with high temperature process, have appeared , and became a hot topic in research. In parallel, low temperature surface passivation using hydrogenated amorphous silicon and forming a heterojunction with the crystalline silicon were developed . In 1990, Sanyo Electric Co. presented the first SHJ solar cell using an a-Si:H (i) passivation layer, named HIT (Heterojunction with Intrinsic Thin-layer). In 1992, they exceeded 18 % efficiency with a heterostructure contact at the rear side, Figure II.2-2 a) . With further optimisations, they reached 24.7 % efficiency in 2013 with a very thin wafer (98 µm) . Shortly afterwards, KANEKA Corporation exceeded certified 25 % efficiency on large area, Figure II.2-2 b) , . This efficiency record was the best for front-rear contacted SHJ solar cells until the Hanergy Group achieved 25.11 % efficiency in total area (M2 wafer) for a bifacial (with a metallisation grid at the back-side instead of a full surface metallisation) solar cell, in 2019 . Figure II.2-2 c) shows the bifacial SHJ solar cells produced in the pilot-line at the CEA-INES and the best efficiencies reached. The best efficiencies for crystalline silicon solar cells are not obtained with front-rear contacted configuration but with IBC configuration (Figure II.2-1 a) for example), with passivated contacts at low temperature  or at high temperature . The IBC configuration allows higher current thanks to a larger active area (no metallisation on the front-side). Tandem solar cells with an IBC bottom-cell are 3T (3-Terminal) structures. This work is focused on the front-rear contacted configuration to make 2T tandem solar cells, thus we do not detail more IBC solar cells here.
Table of contents :
I. GENERAL INTRODUCTION
I.2. MOTIVATION AND OBJECTIVES
II.1. TANDEM SOLAR CELLS
II.2. SILICON HETEROJUNCTION SOLAR CELLS
II.3. PEROVSKITE ON SILICON HETEROJUNCTION TANDEM SOLAR CELLS
II.3.2. Review Before 2018
II.3.3. 2018-2020 Timeline
II.4. RECOMBINATION JUNCTIONS
II.4.1. Highly Doped Tunnel Junction
II.4.2. Transparent Conductive Oxide as Recombination Layer
II.4.3. Recombination Junctions in Perovskite/SHJ Tandem
II.5. CHAPTER OUTLOOK
III. MATERIALS AND METHODS
III.1. FABRICATION PROCESSES
III.1.1. Substrate cleaning
III.1.2. Plasma-enhanced CVD
III.1.2.a. Industrial Tool
III.1.2.b. Development Tool
III.1.3. Transparent Conductive Oxide Deposition
III.1.4.a. Electron-beam Evaporation
III.1.4.b. Thermal Evaporation
III.1.5. Laser Cutting
III.1.6. Perovskite Solar Cell
III.1.7. Example of a Tandem Solar cell Fabrication Process
III.2. LAYER CHARACTERISATION METHODS
III.2.1.a. Ellipsometry Measurements
III.2.1.b. Fitting Models
III.2.3. Raman spectroscopy
III.2.4. Four-terminal Sensing
III.2.5. Hall Effect Measurements
III.2.6. Thin-film’s Activation Energy
III.2.7. Contact Resistances Measurements
III.2.8. Scanning and Transmission Electron Microscopy
III.2.8.a. SEM Measurements
III.2.8.b. STEM-EDX Measurements
III.2.9. Confocal Microscopy
III.2.10. X-ray Diffraction
III.3. SOLAR CELLS CHARACTERISATION
III.3.1. Lifetime Measurements
III.3.2. Suns-Voc Measurements
III.3.3. Quantum Efficiency Measurements
III.3.4. Current-Voltage Measurements
III.3.4.a. Solar Cells J-V curves
III.3.4.b. Temperature-Dependent Dark I-V
III.4. OPTICAL SIMULATION
III.4.1. Software Presentation: CROWM
III.4.2. Optical Indexes Extraction Method
III.5. CHAPTER CONCLUSION
IV. MICROCRYSTALLINE SILICON TUNNEL JUNCTION FOR PIN TANDEM SOLAR CELLS
IV.2. MICROCRYSTALLINE SILICON LAYERS DEVELOPMENT
IV.2.1. PECVD Parameters
IV.2.2. Growth and Microcrystalline Phase
IV.2.3. Microcrystalline Silicon layers properties
IV.3. INTEGRATION OF MICROCRYSTALLINE SILICON ON SHJ SOLAR CELLS
IV.3.1. Bulk Passivation Improvement
IV.3.2. Rear Emitter SHJ Solar Cells with Front-side N-Type Microcrystalline Silicon
IV.3.3. Standard Emitter SHJ Solar Cells with Back-side Microcrystalline Silicon Tunnel Junction
IV.4. CHAPTER CONCLUSION
V. WAYS TO CHARACTERISE RECOMBINATION JUNCTIONS FOR PEROVSKITE ON SILICON TANDEM SOLAR CELLS
V.2. TUNNEL RECOMBINATION JUNCTION AS TUNNEL DIODE
V.2.1. Test-Structures for Tunnel Junctions Measurements
V.2.2. Tunnel Diode Behaviour
V.3. TRANSPORT MECHANISMS INSIGHT IN NPN JUNCTIONS
V.4. CHARACTERISATION OF RECOMBINATION JUNCTIONS
V.5. CHAPTER CONCLUSION
VI. TOWARDS PEROVSKITE ON SILICON HETEROJUNCTION TANDEM SOLAR CELLS INTEGRATION
VI.2. OPTICAL ANALYSIS OF RECOMBINATION JUNCTIONS IN PK/SHJ TANDEM SOLAR CELLS
VI.2.1. Transparent Conductive Oxides
VI.2.2. Microcrystalline Silicon Tunnel Junction
VI.3. INVESTIGATION OF THE RECOMBINATION JUNCTIONS INFLUENCE ON THE PEROVSKITE TOP-CELL
VI.4. TANDEM SOLAR CELLS INTEGRATION
VI.5. CHAPTER CONCLUSION
VII. CONCLUSION AND PERSPECTIVES