GaInP single junction and GaInP/GaAs two junction thin-film solar cell structures by epitaxial lift-off

The epitaxial lift-off (ELO) technique was used in forming a thin-film GaInP/GaAs two-junction monolithic tandem solar cell structure. First, the GaInP single junction solar cell to be used in the tandem cell structure as a top cell was thinned by the ELO process. Although the ELO process and the transfer to the quartz substrate caused a strain in the thin-film cell after separation from the GaAs substrate, the photoluminescence peak intensity was not decreased. This shows that defects, such as those causing carrier loss, were not introduced on the thin-film cell during the thinning process. The key issue for thin-film cell fabrication is to avoid damaging the AlInP window layer during the selective etching (HF etchant), by which the thin-film cell is released from the GaAs substrate. A GaInP/GaAs monolithic tandem structure was also thinned by the same process with a GaInP single junction cell. Characteristics of the single-junction GaInP cell and individual cells in the GaInP/GaAs tandem structure were examined. It was found that the spectral response remains almost the same as that for cells with a GaAs substrate, thus confirming the feasibility of using the ELO process to fabricate thin-film GaInP/GaAs cells.     

Process damage free thin-film GaAs solar cells by epitaxial liftoff with GaInP window layer

We report on process damage free thin-film GaAs cells detached from the GaAs substrates. GaAs cells grown by gas-source MBE were thinned by the epitaxial liftoff (ELO) technique. Photoluminescence spectroscopy showed a peak splitting in the band emission, indicating that a strain was induced in the thin-film cell fixed on the quartz glass substrate. The strain, however, was found not to affect the quality of the thin-film cells, based on the fact that the peak intensity was almost twice that before ELO. The thin-film GaAs cells showed no evidence of degradation in diode characteristics and spectral responses. The keys to avoiding damage on the active region of the solar cell during the thinning process are the introducing a GaInP window layer and improving the thin film process including metallization on thin film cells. These results demonstrates that the thinning and transfer processes dol-not affect the quality of the active region of the cells.     

Shadow-epitaxy for flexible crystalline thin-film silicon solar cell design

Crystalline silicon thin-film solar modules require a cost-effective means for the fabrication of an integrated series connection. We introduce selective epitaxy by ion-assisted deposition through shadow masks for the integrated series connection of ultrathin monocrystalline and textured Si solar cells. The series connection is made in situ during Si-film deposition thus avoiding any trench etching. The open-circuit voltage of an experimental mini-module consisting of six cells is six times as large as the open-circuit voltage of a single cell. Epitaxy through shadow masks also permits the in situ realization of parallel multi-junction solar cells. The experimental quantum efficiency analysis of a parallel multi-junction device with three junctions reveals an enhanced carrier collection. The apparent effective diffusion length of the multi-junction cell is three times larger than that for a conventional single-junction device.     

Energy requirements of thin-film solar cell modules-a review

In this paper a number of energy analysis studies for thin-film solar cell modules are compared and reviewed. We start with a short introduction into methodological issues related to energy analysis (of PV systems) such as system boundary definition, treatment of different (secondary) energy types and the choice of functional unit.Subsequently we review results from 6 studies on a-Si modules and 3 studies on CdTe modules. The aim is to present results in a unified format, compare them and try to clarify observed differences. Although significant differences were found, many of these differences could be explained by the choice of materials for the module encapsulation. For categories with large observed differences, like indirect process energy and capital equipment energy, we performed additional analyses in order to gain a better understanding of these aspects.Finally we present best estimates of the energy requirement for present-day a-Si and CdTe thin film modules which are between 600 and 1500 MJ (primary energy) per m^fn3 module area, depending on cell and encapsulation type. This means that the energy pay-back time is below two years for a grid-connected module under 1700 kWhm²yr irradiation. In the near future an energy pay-back time below one year seems feasible.     

A review of thin-film crystalline silicon for solar cell applications. Part 1: Native substrates

Approximately half the cost of a finished crystalline silicon solar module is due to the silicon itself. Combining this fact with a high-efficiency potential makes thin-film crystalline silicon solar cells a growing research area. This paper, written in two parts, aims to outline world-wide research on this topic. The subject has been divided into techniques which use native substrates and techniques which use foreign substrates. Light trapping, vapour- and liquid-phase deposition techniques, cell fabrication and some general considerations are also discussed with reference to thin-film cells.     

A review of thin-film crystalline silicon for solar cell applications. Part 2: Foreign substrates

One of the most promising ways to reduce the cost of photovoltaics is thin-film crystalline silicon solar cells. This paper, together with part 1, reviews the current state of research in thin-film crystalline silicon solar cells. Deposition on silicon, novel techniques which use a high-quality, reusable silicon substrate and light trapping have been described in part 1 of this paper. This paper describes deposition on glass and ceramics and discusses cell designs for thin-film crystalline silicon solar cells.     

Epitaxial lift-off process for GaAs solar cell on Si substrate

This research has established the process to transplant GaAs solar cells from GaAs substrate to Si substrate without degrading the conversion efficiency. The conversion efficiency of GaAs solar cell bonded to Si substrate using epitaxial lift-off process is almost the same as that grown on GaAs substrate and is superior to that grown on Si substrate by heteroepitaxy.     

Photocarrier generation in organic thin-film solar cells with an organic heterojunction

Photocurrent–voltage characteristics for organic solar cells with a heterojunction formed between copper phthalocyanine and a perylene derivative (or C₆₀) were studied. The photocurrent was observed under both reverse and forward biases. From the analysis of the photocurrent action spectra, the origin of the reverse photocurrent was attributed to the excitons formed in both the organic layers, whereas that of the forward photocurrent was attributed to the excitons formed in the perylene derivative (or C₆₀) layer. The photocurrent density under reverse bias increased at higher temperatures, suggesting that the charge recombination possibility was lowered at higher temperatures. On the basis of the time responses of the photocurrents observed after pulsed photoirradiation, the charge separation and transport processes are discussed.     

Highly efficient large area (10.5%, 1376 cm) thin-film CdS/CdTe solar cell

Highly efficient large area (10.5%, 1376 cm²) thin-film CdS/CdTe solar cell sub-module has been fabricated. Very recently we also have fabricated a very large area sub-module of aperture area 5413 cm² exhibiting a conversion efficiency of 8.4%. Such a high efficiency has been achieved by depositing all the constituent films such as SnO₂: F, CdS and CdTe having greater uniformity and better quality under atmospheric pressure conditions. A post deposition treatment of CdTe surface with CdCl₂ has been optimized to improve the overall solar cell output performance significantly.     

Porous silicon as an intermediate layer for thin-film solar cell

The potential of porous silicon (PS) with dual porosity structure as an intermediate layer for ultra-thin film solar cells is described. It is shown that a double-layered PS with a porosity of % allows to grow epitaxial Si film at medium temperature (725°–800°C) and at the same time serves as a gettering/diffusion barrier for impurities from potentially contaminated low-cost substrate. A 3.5 μm thin-film cell with reasonable efficiency is realized using such a PS intermediate layer.     

Extreme radiation hardness and light-weighted thin-film indium phosphide solar cell and its computer simulation

A very light-weighted and extreme radiation hardness high-doping n⁺-i-p⁺ InP solar cell is developed. The total thickness of its epitaxial layer is only 0.22 μm. It is more radiation hardened than many other structures so that it can provide the required output power for spacecraft even after a decade at 3200 km polar orbits. Its end of life efficiency is about 10% (AM0, 1Sun); its highest power/weight ratio is about 130 W/g (only the weight of epitaxial layers is considered). The surprising fact is that all of these are obtained from a very simple structure.     

Functional requirements for component films in a solar thin-film photovoltaic/thermal panel

The functional requirements of the component films of a solar thin-film photovoltaic/thermal panel were considered. Particular emphasis was placed on the new functions, that each layer is required to perform, in addition to their pre-existing functions. The cut-off wavelength of the window layer, required for solar selectivity, can be achieved with charge carrier concentrations typical of photovoltaic devices, and thus does not compromise electrical efficiency. The upper (semiconductor) absorber layer has a sufficiently high thermal conductivity that there is negligible temperature difference across the film, and thus negligible loss in thermal performance. The lower (cermet) absorber layer can be fabricated with a high ceramic content, to maintain high solar selectivity, without significant increase in electrical resistance. A thin layer of molybdenum-based cermet at the top of this layer can provide an Ohmic contact to the upper absorber layer. A layer of aluminium nitride between the metal substrate and the back metal contact can provide electrical isolation to avoid short-circuiting of series-connected cells, while maintaining a thermal path to the metal substrate and heat extraction systems. Potential problems of differential contraction of heated films and substrates were identified, with a recommendation that fabrication processes, which avoid heating, are preferable.     

Impact of sheet resistance on 2-D modeling of thin-film solar cells

A rigorous mathematical approach was used to find a relation between the transparent-conductive-oxide (TCO) sheet resistance ρ_S (Ω/□) of a thin-film solar cell and the parameter R (Ω) that describes the TCO resistance in a two-dimensional circuit model. Additionally, the mathematical relationship that connects experimentally derived series resistance R_S (Ω cm²) of the solar cell to the TCO sheet resistance ρ_S (Ω/□) and the bulk semiconductor resistivity ρ (Ω cm) was derived. It was found that the fill factor of the solar cell is governed by a reduced dimensionless TCO sheet resistance that depends only weakly on the type and quality of the solar cell. Parameters corresponding to thin-film Cu(In,Ga)Se₂, two-junction a-Si, and an ideal solar cell were used as concrete examples.     

Cu(In,Ga)Se2 thin-film solar cells with an efficiency of 18%

An efficiency of over 18% have been achieved in Cu(In,Ga)Se₂ (CIGS) thin-film solar cells. Solar cell parameters were estimated for the cells with efficiencies of more and less than 18%. A diode quality factor n and forward current (saturated current) J₀ of the cell with over 18% efficiency are lower than those with below 18% efficiency. This would be attributed to sufficient coverage of the CdS film with excellent uniformity as a buffer and/or window layer over the CIGS film because the process of CdS film formation was improved.     

Characterization of crystalline silicon grown by plasma-enhanced CVD for thin-film solar cells

High growth-rate Si epitaxy by plasma-enhanced chemical vapor deposition (PECVD) has been investigated for a thin-film solar cell application. A high growth rate of 50 μm/h was obtained at 1050°C with plasma which is 50% larger than that by the conventional CVD without plasma. The electrical properties are almost the same for epitaxial layers with and without plasma. For undoped n-type layers, the Hall mobility and carrier density were about 600 cm²/V s and low 10¹⁵ cm⁻³, respectively. The electron diffusion length in doped p-type layers was about 20 μm. These electrical properties for the layer with plasma, in spite of higher growth rate, are comparable or better than those without plasma.     

Thin film GaAs solar cells with increased quantum efficiency due to light reflection

Thin film GaAs solar cells, separated from their substrate using the weight-induced epitaxial lift-off technique, were compared with conventional cells on a substrate. The thin film cells can be illuminated from both sides using a mirror. The thickness of the p-type GaAs layer, which is the base layer for front illumination and the emitter layer for rear illumination, was varied between 0.25 and 2.5 μm. For both front and rear illumination, the cell efficiency shows a maximum at a thickness of 1.5 μm. The rear illuminated cell current is only 10% lower than for front illumination. Light reflection in the thin film cell enhances the external quantum efficiency and the collection efficiency in the higher wavelength region from 0.84 to 0.90 and from 0.82 to 0.95, respectively.     

c-Si wafer-equivalent epitaxial thin-film solar cells on isolating substrates

A new cell concept has been developed that enables epitaxial c-Si thin-film solar cells to be made using isolating substrates. The Recrystallised Wafer-Equivalent on an Isolating Substrate (RexWISe) cell concept relies on an array of mini-silicon rods through the substrate to enable standard contacting. Processing techniques have been developed to produce the rods by drilling holes through the substrate, coating the holes with an intermediate layer, filling the holes with a seeding layer deposition and then recrystallising the seeding layer. Subsequently, the active layers of the cell are epitaxially grown onto the recrystallised layer and then this “Wafer-Equivalent” structure is metallised like a standard wafer solar cell. The first solar cells have been produced to test the RexWISe process and a “proof-of-concept” efficiency of almost 8% was achieved.     

High-efficiency drift-field thin-film silicon solar cells grown on electronically inactive substrates

A drift-field in the base region of a solar cell can enhance the effective minority-carrier diffusion length, thus increasing the long-wavelength spectral response and energy-conversion efficiency. Silicon thin-films of 20–32 μm thickness as a cell base layer were grown by liquid-phase epitaxy (LPE) on electronically inactive heavily doped p⁺⁺-type CZ silicon substrates. Growth was performed from In/Ga solutions, and in a purified Ar/4%H₂ forming gas ambient, rather than pure H₂. The Ga dopant concentration was tailored throughout the p-type film to create a drift-field in the base layer of the solar cell. An independently confirmed efficiency of 16.4% was achieved on such an LPE drift-field thin-film silicon solar cell with a total cell area of 4.11 cm². Substrate thinning, combined with light trapping which is encouraged by the textured front surface and a highly reflective aluminium rear surface, is demonstrated to improve the long-wavelength response and thus, increase cell efficiency by a factor of up to 23.7% when thinned to a total cell thickness of 30 μm.     

Silicon thin-film solar cells on insulating intermediate layers

An interdigitated front grid structure for both the emitter and base was simulated and realized. This contact design is suitable for thin-film solar cells on insulating substrates or insulating intermediate layers. Confirmed efficiencies of up to 18.2% were achieved on a 46 μm thick epitaxial silicon layer which was grown on a SIMOX wafer with an implanted compact SiO₂ intermediate layer.Samples with and without a highly doped back surface field were prepared to study the influence of the back-side recombination velocity. L_eff values of 250 and 52 μm, corresponding to S_back values of 800 and 10⁵ cm/s, respectively, were measured, thus, underlining the importance of a low back-side recombination velocity. The optical confinement properties of the SiO₂ intermediate layers were calculated depending on the angle of the incident rays. An angle from the plane normal which is larger than 23° is necessary in order to achieve the condition of total internal reflection.Future work will focus on recrystallized Si layers on foreign substrates [1]. Since the surface of the silicon layer is fairly rough after the recrystallisation process, another set of masks was designed which is more tolerant to aligning accuracy. This is mainly relevant for the area where the base contacts are located between the locally diffused emitter. The technology for CVD Si-layer deposition, zone melting recrystallization (ZMR), as well as for a simplified solar cell process is under investigation.     

Effects of grain boundaries in polycrystalline silicon thin-film solar cells based on the two-dimensional model

The two-dimensional calculation for polycrystalline Si thin-film solar cells was performed. Two models, “stripe structure” and “columnar structure”, were applied for the solar cells composed of grains. For the stripe structure of 20 μm active layer, to keep the efficiency distribution within 5% for individual unit cells, the stripe width requires more than 500 μm for a minority-carrier lifetime of 1×10⁻⁵ s and recombination velocity at the grain boundary of 1×10⁴ cm/s. For the columnar structure of 10 μm active layer, to keep the efficiency independent of grain size, the recombination velocity should be kept less than 1×10³ cm/s. If imperfect passivation of a grain boundary is given, the way of decreasing carrier concentration to 10¹⁴ cm⁻³ in an active layer may realize insusceptible output. An appropriate device modeling is needed in the two-dimensional calculation for polycrystalline Si thin films with an electron diffusion length close to or more than grain size and with a poorly passivated grain boundary. The calculated efficiency using bad model will include an error of about 1% as overestimation.