Project Topics
Get Complete Project Material File(s) Now! »
Generated carrier dynamics under the influence of electric fields
In this section we will study one additional parameter that is believed to have an influence on the generated charge carriers. That parameter is the electric field that is created once the junction is formed within the photovoltaic material. We will also precise under which experimental conditions it is relevant to consider the effect of charge separation on the probed carrier dynamics.
R.K. Ahrenkiel[74] and A.J. Nozik [75] conducted through simulation work, the first rigorous studies of the effects of electric fields on the time-resolved photoluminescence spectra in III-V semiconductors.
The need for this analysis originated from the fact that under experimental conditions where drift is important enough to have a clear influence on the recombination channels of charge carriers. The importance of the drift depends on the width of the space-charge region WSCR and the intensity of the electric field. The experimental conditions such as the value of the applied voltage-bias and the photoexcitation level will determine the “screening” degree of the electric field.
Spectrally Resolved Photoluminescence
Any system of temperature T emits light following the Planck’s law of radiation. This body can be brought out of equilibrium, as for example a semiconductor under a laser pump. As a result of excess carrier concentration, spontaneous radiative recombination of electron hole pairs occurs. This effect strongly increases the emission of photons with energy higher than the bandgap.
The recombination of a hole at an energy E1 and an electron at an energy E2 results in an emission of a photon at an energy E3 = E2 – E1. The following recombination rate is obtained [159]: 3 0 1 2 1 2 2 3 ( ) , 1 D E n c R E M g E E f E f E refr sp (1.23).
2 M is a matrix element coupling the initial and final states, f (E) the occupation probability of the electrons, 0 c the speed of light in vacuum, refr n the material refractive index and D the photon density of states. ( , ) 1 2 g E E is a combined density of states, reflecting the density of states at 1 E & 2 E , but also the conservation of momentum [159]. The first quantities are also related to the absorption coefficient ( ) 3 E so that at equilibrium, the emission .
Spectrally-Resolved Electroluminescence & LED-PV Reciprocity Relations
If we describe the operation of a solar cell and a light emitting diode (LED), one can deduce that both systems function in a reciprocal way such as the first system’s topology of functioning depends on the inverse phenomenon of the other. In a general way, a solar cell’s physical action is transforming radiation energy into electricity while an LED transforms electrical energy into light. Theoretically, based on the principle of detailed balance[42], we consider that at thermodynamic equilibrium if one photon injected causes the creation of one electron-hole pair (photovoltaic operation), such an event must be compensated by the opposite phenomenon stating that one electron-hole pair injected should lead to the creation of one photon (electroluminescence). Based on the complementary behavior of these two systems, one might believe that a perfect solar cell would also be a perfect LED, however in reality this is not the case.
If we take for example today’s leading high-efficiency silicon solar cells that have already approached the SQ-limit, therefore, in a practical way we can consider such solar cells to be almost perfect. If we are ever to produce LEDs based on the same designs of the previous solar cells we might find that the result isn’t as “perfect” as we expected. For example, a highly sophisticated record holder silicon solar cell, once in LED mode, results in an external quantum efficiency EQELED of approximately 1% only. On the opposite side, organic semiconductor-based LEDs prove to have quite a respected EQELED of 15% [43] comparing to solar cells made from organic materials that have very low power conversion efficiencies. Therefore, it is safe to conclude that at the practical level, quality requirements for solar cells and LEDs diverge in a considerable way.
Information derived from the luminescence intensity, its spectral and spatial variations
We have seen that the generalized Planck’s law and the reciprocity relations can indeed inform us about the characteristics of the solar cells. From equation (1.26) we have learned that the photoluminescence intensity and the corresponding spectrum depend on three parameters: The quasi-Fermi levels splitting profile Δμ(r), the absorptivity A(E,r) and the temperature T. Based on these observations, numerous contributions show how characteristics of devices are determined. In particular, from equation (1.4), Δμ was determined. In PL and EL, it was shown to equal the voltage applied to the junction [45-54]. Therefore, PL is an access route to Voc. In some materials, the Δμ depth profile is found to fluctuate. These depth variations are highly interesting when we trace them back to their origin. For example, important transport properties were found to be at the origin of these fluctuations such as the carrier diffusion length and the surface recombination velocities [55]. Therefore, analyzing the PL signal leads us to the value of Δμ from which we finally extract the material transport properties.
Now, once we determine the Δμ, we can see from equation (1.4) that other key parameters can therefore be accessed such as the material absorption [56] and the charge carrier temperature [57]. Finally, to all of the above we can add the spatial dimension and consequently establish mappings of the latter key parameters. This will allow us to assess and quantify their spatial inhomogeneity. This type of investigation is essential for highly inhomogeneous photovoltaic materials that behave and perform differently throughout their depth or surface. For example, maps of the saturation currents[58,59] and the diffusion lengths [60] were achieved on silicon solar cells. Given its contactless nature, PL has also been demonstrated as an efficient tool for quality control of production lines [61-63]. Maps of Δμ have been recently demonstrated on CIGS solar cells[64].
TRPL studies on CIGS solar cells: The literature contradiction dilemma
So far, we have detailed the theoretical aspects of the spectral and temporal dependencies of the luminescence of PV devices. In the scope of the present thesis, the major experimental focus took place on the temporal aspect of the luminescence. The majority of our discussions in this thesis will focus on the time-resolved techniques. The idea is to analyze the evolution of the luminescence intensity as a function of time. The luminescence “decay” contains crucial information regarding the PV material such as the quantification of its carrier dynamics, the trap density it contains and many more key information.
Theoretical aspect: Physics of charge carrier dynamics in Cu(In,Ga)Se2 PV devices
Time-resolved measurements of the photoluminescence decay upon a pulsed excitation allow studying the recombination kinetics in photovoltaic absorber materials. Hence the PL decay time can be an indicator or a measurement tool for the electronic quality of a given semiconductor. TRPL can thus constitute an attractive technique for the qualitative or quantitative inspection of photovoltaic absorber materials.
The theory sketched in this section provides the basic concepts of TRPL measurements of photovoltaic devices and their absorber materials. The physical concepts to interpret TRPL results have been first described in detail by R.Ahrenkiel [66].
Table of contents :
Index of Symbols
Introducing the PhD project
The General Context of Photovoltaic Energy
Chapter I. Theoretical background on CIGS PV devices and their luminescence properties
1.1 Chalcogenide Cu(In,Ga)Se2 thin-film technology
1.1.1 Introduction on CIGS PV technology & Cell configuration
1.1.1.1 Back Contact
1.1.1.2 CIGS Absorber Layer – Deposition Methods
1.1.1.3 CdS Buffer Layers
1.1.1.4 Front Contact
1.1.2 The particular case of micro-CIGS solar cells used in our experiments
1.2 Physics of PV device luminescence
1.2.1 Recombination Mechanisms
1.2.2 Time-Resolved Photoluminescence (TRPL)
1.2.3 Generated carrier dynamics under the influence of electric fields
1.2.4 Spectrally Resolved Photoluminescence
1.2.5 Spectrally-Resolved Electroluminescence & LED-PV Reciprocity Relations
1.2.6 Reciprocity between Photoluminescence and Electroluminescence of solar cells
1.2.7 Information derived from the luminescence intensity, its spectral and spatial variations
1.3 TRPL studies on CIGS solar cells: The literature contradiction dilemma
1.3.1 Theoretical aspect: Physics of charge carrier dynamics in Cu(In,Ga)Se2 PV devices
1.3.2 Photo-excitation-dependent TRPL decay analysis on CIGS
1.3.3 Voltage-bias-dependent TRPL decay analysis
1.3.4 Photo-excitation energy-dependent TRPL decay analysis
Chapter II. Optoelectronic Characterization of PV Devices
2.1 Standard Solar Cells Characterization Techniques
2.1.1 J-V Characteristics
2.1.2 External Quantum Efficiency
2.1.3 Capacitance-based Techniques
2.2 Advanced Luminescence-Based Characterization Techniques
2.2.1 Hyperspectral Imager
2.2.2 Scanning Confocal Microscope (SCM)
2.2.3 Comparison between the HI and the SCM
2.2.4 Time-Resolved Fluorescence Lifetime Imaging (TRFLIM)
2.2.4.1 Overview of technologies and Operating principles of TRFLIM
2.2.4.2 Design and development of a setup dedicated to TRFLIM
2.2.4.3 Spatial and Temporal resolutions of the detection system
2.2.4.4 Characteristics of the excitation system
2.2.4.5 Concluding the construction of the complete TRFLIM setup with a proof of concept
Chapter III. Quantitative local access to optoelectronic properties of Cu(In,Ga)Se2 PV devices using Time- Resolved Luminescence
3.1 General introduction and objectives
3.2 Development of a unified understanding of charge carrier dynamics in Cu(In,Ga)Se2 PV devices
3.2.1 Introduction
3.2.2 Physical model based on carrier recombination centers and shallow defects acting as carrier traps
3.2.3 Photo-excitation intensity-dependent and electrical carrier-injection dependent time-resolved
luminescence experiments
3.2.4 Investigating the potential influence of the built-in electric field
3.2.5 Identifying the characteristics of probed carrier recombination
3.2.6 Brief summary and conclusion
3.3 Optical alternative for the study of metastabilities in Cu(In,Ga)Se2 PV devices
3.3.1 Metastabilities-induced distortions of Current-Voltage characteristics
3.3.2 Optical evidence of the effect of metastabilities activation on minority carrier dynamics
3.3.2.1 Introduction & Methodology
3.3.2.2 Basic Characterization
3.3.2.3 Qualitative optical observation of a blue photon recovery effect
3.3.2.4 Hysteresis cycle of the “DCD” parameter
3.3.2.5 Hysteresis cycle of the carrier trapping lifetime
3.3.2.6 Brief Conclusion
3.4 Contactless quantification of trapping defects density in Cu(In,Ga)Se2 PV devices
3.4.1 Development of a physical model for the reconstruction of the time-resolved luminescence decays
3.4.2 Application to the photo-excitation intensity-dependent time-resolved luminescence decays
3.4.3 Application current injection-dependent time-resolved luminescence decays
3.4.4 Discussion on the application to the photo-excitation intensity-dependent and current injectiondependent time-resolved luminescence decays
3.4.5 The effect of the trapping defects density on global photovoltaic performance
3.4.6 An attempt at a quantitative optical spectroscopy of trapping defects
3.4.7 Discussing the choice, the limitations and relevance of the physical model
Chapter IV. Quantitative luminescence-based imaging of optoelectronic properties of thin film PV devices
4.1 Introduction
4.2 Spatially-resolved, spectrally-resolved and time-resolved luminescence of Cu(In,Ga)Se2 PV devices
4.2.1 Quantitative electrical mapping of the variations of PV performance indicators following
metastabilities activation
4.2.2 Quantitative optical mapping of trapping defects-induced variations in PV performance indicators
4.2.2.1 Mapping the effect of light-soaking on the quasi-Fermi levels splitting deduced from electroluminescence imaging
4.2.2.2 Micrometric mapping of charge carrier lifetimes and trapping activity
4.2.2.3 Optical micrometric mapping of the absolute trapping defects density
4.2.2.4 Mapping the micrometric losses in quasi-Fermi levels splitting
4.2.2.5 Investigating the effect the photo-excitation energy
4.3 Time-Resolved Fluorescence Lifetime Imaging (TRFLIM) of Solar Cells
4.3.1 GaAs solar cell – An experimental study
4.3.2 GaAs solar cell – A numerical study & material properties extraction
Conclusion
Bibliography
