• Energy Research

During the last years, we developed both an efficient High-Frequency Modulated Photoluminescence (MPL) setup, covering a frequency range of 10Hz-200MHz [1] and a corresponding analytical theory [2] allowing to explain the appearance of V-shaped phase patterns in the MPL bode diagrams of probed semiconductors layers by the presence of Shockley Read Hall (SRH) recombination. An example of V-shape in phase behavior can be seen in the bode diagram on figure 1a. We precedingly investigated a fitting strategy based on curves obtained at several illumination levels. However, experiments at several temperatures are also a promising way to explore recombination paths. This year, we performed theoretical calculations and analytical simulations in order to show that the variations of the V-shape corner frequencies w1, w2, w3 with respect to the temperature allow for recovering some trap parameters such as the energy position in the bandgap (activation energy) and the minority carrier capture cross section (see Figures 1b and 1c). The validity domain of this theory will be discussed for low and high injection as well as doped and intrinsic materials. We also extended the previous theory [2] by finding supplementary rules for the appearance and disappearance of V-shapes. We proofed that the MPL amplitude curve also contains information about the injection level during the experiment. Finally, we are currently performing MPL versus temperature experiments in order to validate our theory. We will present all these results during the conference. [1]W. Zhao et al., « Coupled time resolved and high frequency modulated photoluminescence probing surface passivation of highly doped n-type InP samples », J. Appl. Phys., vol. 129, no 21, p. 215305, juin 2021, doi: 10.1063/5.0033122.[2]N. Moron, B. Bérenguier, J. Alvarez, et J.-P. Kleider, « Analytical model of the modulated photoluminescence in semiconductor materials », J. Phys. Appl. Phys., vol. 55, no 10, p. 105103, mars 2022, doi: 10.1088/1361-6463/ac39c4.

During the last years, we developed both an efficient High-Frequency Modulated Photoluminescence (MPL) setup, covering a frequency range of 10Hz-200MHz [1] and a corresponding analytical theory [2] allowing to explain the appearance of V-shaped phase patterns in the MPL bode diagrams of probed semiconductors layers by the presence of Shockley Read Hall (SRH) recombination. An example of V-shape in phase behavior can be seen in the bode diagram on figure 1a. We precedingly investigated a fitting strategy based on curves obtained at several illumination levels. However, experiments at several temperatures are also a promising way to explore recombination paths. This year, we performed theoretical calculations and analytical simulations in order to show that the variations of the V-shape corner frequencies w1, w2, w3 with respect to the temperature allow for recovering some trap parameters such as the energy position in the bandgap (activation energy) and the minority carrier capture cross section (see Figures 1b and 1c). The validity domain of this theory will be discussed for low and high injection as well as doped and intrinsic materials. We also extended the previous theory [2] by finding supplementary rules for the appearance and disappearance of V-shapes. We proofed that the MPL amplitude curve also contains information about the injection level during the experiment. Finally, we are currently performing MPL versus temperature experiments in order to validate our theory. We will present all these results during the conference. [1]W. Zhao et al., « Coupled time resolved and high frequency modulated photoluminescence probing surface passivation of highly doped n-type InP samples », J. Appl. Phys., vol. 129, no 21, p. 215305, juin 2021, doi: 10.1063/5.0033122.[2]N. Moron, B. Bérenguier, J. Alvarez, et J.-P. Kleider, « Analytical model of the modulated photoluminescence in semiconductor materials », J. Phys. Appl. Phys., vol. 55, no 10, p. 105103, mars 2022, doi: 10.1088/1361-6463/ac39c4.

We propose a simple and non-destructive method to access the properties of the defects within the subcells of multijunction solar cells (MJSC) based on the extension of two well-known techniques, namely Admittance Spectroscopy (AS) and Deep-Level Transient Spectroscopy (DLTS). Due to the electrical coupling between the subcells of an MJSC, the AS and DLTS signatures of the defects present in the subcells are difficult to identify and to properly analyse, and other independent phenomena could be misinterpreted as such. From numerical simulations and theoretical developments, it is shown how the use of dedicated experimental conditions, in particular specific light biases, on MJSCs can reveal the signatures of the defects present in each subcell and allow one to analyse them without misinterpretation.

We propose a simple and non-destructive method to access the properties of the defects within the subcells of multijunction solar cells (MJSC) based on the extension of two well-known techniques, namely Admittance Spectroscopy (AS) and Deep-Level Transient Spectroscopy (DLTS). Due to the electrical coupling between the subcells of an MJSC, the AS and DLTS signatures of the defects present in the subcells are difficult to identify and to properly analyse, and other independent phenomena could be misinterpreted as such. From numerical simulations and theoretical developments, it is shown how the use of dedicated experimental conditions, in particular specific light biases, on MJSCs can reveal the signatures of the defects present in each subcell and allow one to analyse them without misinterpretation.

AbstractThis paper presents a simple and nondestructive method to determine doping densities and built‐in potential of subcells by adapting the well‐known capacitance‐voltage (C‐V) technique to two‐terminal (2 T) tandem solar cells. Because of the electrical coupling between the two subcells in a monolithic 2 T tandem solar cell, the standard method using a Mott‐Schottky plot (1/C2 vs V) cannot be applied. Using numerical modeling, it is demonstrated that, by under chosen illumination conditions where only one subcell can absorb the light, it is possible to explore the bias dependence of the capacitance and to extract the parameters of the other subcell if the appropriate frequency conditions are present. This method is experimentally applied to an AlGaAs/Si tandem cell, and parameters of both AlGaAs and Si cells are extracted. Finally, the validity of that method is assessed by the very good agreement obtained when comparing the values extracted from our measurements on the tandem cell to those extracted from measurements on isotype cells and to the values targeted during the fabrication process of the AlGaAs/Si tandem solar cell.

AbstractThis paper presents a simple and nondestructive method to determine doping densities and built‐in potential of subcells by adapting the well‐known capacitance‐voltage (C‐V) technique to two‐terminal (2 T) tandem solar cells. Because of the electrical coupling between the two subcells in a monolithic 2 T tandem solar cell, the standard method using a Mott‐Schottky plot (1/C2 vs V) cannot be applied. Using numerical modeling, it is demonstrated that, by under chosen illumination conditions where only one subcell can absorb the light, it is possible to explore the bias dependence of the capacitance and to extract the parameters of the other subcell if the appropriate frequency conditions are present. This method is experimentally applied to an AlGaAs/Si tandem cell, and parameters of both AlGaAs and Si cells are extracted. Finally, the validity of that method is assessed by the very good agreement obtained when comparing the values extracted from our measurements on the tandem cell to those extracted from measurements on isotype cells and to the values targeted during the fabrication process of the AlGaAs/Si tandem solar cell.

It is now well established that halide perovskite materials, such as Methyl Ammonium Lead Iodide (MAPI), contain low-mobility ions that affect the device operation and performance. Ionic motion is believed to be primarily responsible for the long transients observed in dark J-V measurements (hysteresis) of perovskite devices.In this work, we use a drift-diffusion numerical simulation to evaluate the main characteristics of the current transients produced by a mobile ion after biasing simple Metal-Semiconductor-Metal (MSM) structures. We compare the theoretical results with experimental measurements performed in monocrystals and thin-films devices of halide perovskites. We observe that most of the transient characteristics can be explained with our model and we present a discussion of the possible causes for some observed discrepancies.We relate the semiconductor parameters, such as the concentration of dopants and ionic species, with the current transient shape and time position. We deduce two analytical formulas for extracting the ionic mobility and the ratio between ionic and dopant concentrations from the current transient. Finally, we perform measurements at different temperatures for extracting the activation energy of the ionic mobility.

It is now well established that halide perovskite materials, such as Methyl Ammonium Lead Iodide (MAPI), contain low-mobility ions that affect the device operation and performance. Ionic motion is believed to be primarily responsible for the long transients observed in dark J-V measurements (hysteresis) of perovskite devices.In this work, we use a drift-diffusion numerical simulation to evaluate the main characteristics of the current transients produced by a mobile ion after biasing simple Metal-Semiconductor-Metal (MSM) structures. We compare the theoretical results with experimental measurements performed in monocrystals and thin-films devices of halide perovskites. We observe that most of the transient characteristics can be explained with our model and we present a discussion of the possible causes for some observed discrepancies.We relate the semiconductor parameters, such as the concentration of dopants and ionic species, with the current transient shape and time position. We deduce two analytical formulas for extracting the ionic mobility and the ratio between ionic and dopant concentrations from the current transient. Finally, we perform measurements at different temperatures for extracting the activation energy of the ionic mobility.

The need for non-contact techniques to improve the characterization of electronic defects in materials and components is critical for the assessment of new photovoltaic technologies. Advanced characterization tools are required to study such defects and quantify their density within the material. The Modulated Photoluminescence (MPL) technique that can operate in a wide frequency range [10Hz-200MHz] has proven to be a promising candidate, showing its utility in characterizing defects in semiconductor materials. A previous study revealed singularities in the phase dependence of MPL as a function of frequency, called V-shapes, where the phase is not monotonously varying with frequency, but exhibits a local extremum. On this basis, in our study, we carried out a theoretical analysis by solving the continuity equations, as described in a previous article. This suggests that the existence of a V-shape may be related to the presence of minority carrier traps within the material, but this relationship is not always clear and requires additional explanation. In studying the variation of the V-shape with temperature, we extended previous theoretical work by identifying rules for the appearance and disappearance of the V-shape in the MPL bode diagrams of probed materials in the presence of Shockley-Read-Hall (SRH) recombination centers. From modelling of such materials in low injection, by varying the energy level position of the defect and the capture cross-section of minority and majority carriers, we show that the V-shape curves appear when the defect is a minority carrier trap. In this case, we propose a method to extract the information of the defect properties. However, for other conditions like higher injection and intrinsic material,V-shapes can appear even if the defect is a majority carrier trap. These results will be illustrated using modelling and experimental works.