Search

We study amorphous silicon-germanium (a-SiGe:H) as intrinsic absorber material for thin-film silicon-based triple and quadruple junction solar cells. First, we present the development of a-SiGe:H single junction devices, in particular the Ge-content grading in the absorber layer, the influence of the Ge-content on electrical properties and (infra)red-response, and the influence of using different types of players. We subsequently show the incorporation of optimized single-junction devices in triple junction cells and discuss the interplay between Ge-content and intermediate reflector thickness. For triple junction devices with amorphous silicon (a-Si:H) top cells, a-SiGe:H middle cells and microcrystalline silicion (mu c-Si:H) bottom cells, we obtained an initial efficiency of 13.6% and an efficiency of 11.3% after light-soaking. We also present a quadruple junction device with an a-Si:H top cell, a low Ge-content a-SiGe:H second cell, and mu c-Si:H third and bottom cells. In this device configuration, we obtained an open-circuit voltage as high as 2.57 V. The performance of these cells was limited by not yet optimized current matching, leading nevertheless to an initial efficiency of 10.1%. A brief roadmap towards quadruple-junction devices with stabilized efficiencies of 14% is also outlined. (C) 2014 Elsevier B.V. All rights reserved.

One avenue to enhance CdTe cell performance is to improve the optical transmission of the transparent conductive oxide (TCO)/window layer stack. In this paper, we examine soda-lime float glass coated with an Al-doped ZnO layer and a buffer layer. The possible advantages of using a ZnO-based TCO include reduced surface roughness, improved transparency, and an integrated buffer layer that can be optimized for use in a CdTe PV device. Device processing was modified to address the chemical and thermal differences between the ZnO-based TCO stack produced by Saint-Gobain and the TCOs previously used at the National Renewable Energy Laboratory (NREL). These process modifications produced ~ 8% efficiency for devices without a buffer layer. Incorporation of buffer layers has already produced devices with ~ 11% and > 12% efficiency for CdTe deposition temperatures of 570° and 500°C, respectively.

The influence of a lattice-matched GaAsSb capping layer on the structural properties of self-assembled InAs quantum dots (QDs) grown on InP substrates is studied on the atomic scale by cross-sectional scanning tunneling microscopy. While lattice-matched In0.53 Ga0.47 As -capped QDs are clearly truncated pyramids, GaAs0.51 Sb0.49 -capped QDs grown under the same conditions look like full pyramids and exhibit a larger height, indicating that capping with GaAsSb reduces dot decomposition. Since there are no differences in strain between the two capping layers, this behavior is most likely related to the surfactant effect of Sb, which stabilizes the growth front and avoids adatom migration. © 2010 American Institute of Physics. U7 - Export Date: 2 August 2010 U7 - Source: Scopus U7 - Art. No.: 074309

Silicon heterojunction solar cell technology (HJT) takes advantage of ultra-thin amorphous silicon layers deposited on both sides of monocrystalline silicon wafers, enabling excellent silicon wafer surface passivation resulting in high device power output and in addition to efficient use of thin wafers. A full cell processing platform was developed in our laboratory, enabling to achieve > 22 % cell efficiency. Advanced concepts for metallization and interconnection are under study, from fine-line printing combined with SmartWire interconnection to Copper plating. Importantly, we show that the HJT technology intrinsically enables high bifaciality of the cells (> 95 %), and further demonstrates a low thermal coefficient (< 0.2 – 0.3 %/°C). The high performance of heterojunction cells and SmartWire interconnection based modules allow for very low cost of electricity for Heterojunction based solar systems, with a potential below 6 Euro cents per kWh in Europe, bringing photovoltaics as a very competitive electricity source. It further provides upgrade potential towards 24 % cell efficiency.

In thin-film silicon “micromorph” tandem photovoltaics devices, the fill factor (FF) of the microcrystalline silicon (μc-Si:H) bottom cell is typically higher than that of the amorphous silicon (a-Si:H) top cell. Consequently, in order to maximize the efficiency, the bottom cell should be slightly limiting the current in short-circuit condition. Using an intermediate reflector layer (IRL) is a good solution to increase the photocurrent in the top cell without changing the thickness of the active layer. In this contribution, we evaluate different IRL by measuring current-voltage (I-V) characteristics of tandem cells under various irradiance spectra (from red-rich to blue-rich), to vary the short-circuit current matching conditions (from top-limited to bottom-limited). This technique is applicable to multi-junction device of any other photovoltaic technology. With this experiment, a fair appreciation of the IRL relevance can be made, regardless of the sub-cells current matching under AM 1.5 G. By this study, we show that a non-appropriate deposition regime for the silicon-oxide-based IRL can affect the stability of the bottom cell of micromorph devices. We then show that with appropriate deposition regime the degradation rate of micromorph devices can be drastically reduced. Combining such an IRL with all recent development in our laboratory, we obtained a tandem efficiency of 11.9 % under AM 1.5G.

Improving micromorph devices performances nowadays requires a current density increase: good devices typically exhibit an open circuit voltage times fill factor product of one (V oc ×FF∼1.4V×0.71=1). Their short-circuit current density (J sc ) value thus dictates their efficiency (expressed in %). Maximizing it with reasonable cells thicknesses necessitates therefore combining robust cell design with adequate light management through transparent electrodes and intermediate reflectors engineering. We will first show how record micromorph devices (13.5% initial and >11.5% stabilized efficiencies) are prepared on optimized single-layer ZnO electrodes. Such electrodes requirements will be discussed: 1) Strong and wide light scattering is needed on the entire useful wavelength range. Large features grant high total currents (>26mA/cm2) while sharp ones allow for high top cell currents (>13mA/cm2). 2) Sufficiently small or smooth substrate features permits high quality cell growth, providing good cell design (typically V oc over 1.4V). 3) Good conduction and transparency for electrodes (requiring ∼50cm2/V/s TCO mobility) should preserve sheet resistance close to 20Ω/□ (for FF>70%) with low absorption. We will then focus on pushing further micromorph devices potential. Either textured intermediate reflectors can fulfill the bottom cell needs, or double-texture substrates can be implemented: light scattering at large wavelengths is here achieved via nanoimprint lithography (a versatile approach to glass-texturing), topped by small and sharp ZnO features guaranteeing high top cell current. By combining excellent TCO with smart under-structures, thin devices delivering high currents with excellent efficiencies are within reach.

Multi-junction photovoltaic devices include two or more component sub-cells which are electrically interconnected in series. At any power point, the current output of the total device is limited by the sub-cell with the smallest current density. Therefore, the maximum efficiency is reached when the sub-cells have equal current densities at their respective maximum power points. In this case the sub-cells are so called "power matched". We report an experimental procedure in which the current voltage characteristics of tandem solar cells can be measured under various irradiance spectra, i.e. under various shortcircuit current matching conditions. This permits the probing of the optimized short circuit current mismatch, where the sub-cells are power matched, which is essential to define design rules for the tandem stack. The method applies well to devices where one of the sub-cells is metastable. We show that, in the case of thin-film silicon tandem cells, the optimum mismatch changes significantly after light induced degradation. Consequently, the degradation factor of such devices is shown to depend not only on material quality but also on the initial short circuit current matching. This experiment also provides relative quantification of the fill factors of each sub-cell. Our example suggests that a high bottom cell deposition rate can be detrimental to the fill factor of the top cell in the case of thin-film silicon tandem cells deposited in superstrate configuration. (C) 2013 Elsevier B.V. All rights reserved.

In Thin Film Silicon (TF-Si) solar cells light harvesting schemes must guarantee an efficient light trapping in the thin absorber layers without decreasing the silicon layers quality and consecutively the p-i-n diodes electrical performance. TF-Si solar cells resilience to the substrate roughness is reported to be possibly improved through optimizations of the cell design and of the silicon deposition processes. By further tailoring the superstrate texture, amorphous silicon / microcrystalline silicon (a-Si:H/mu c-Si:H) tandem solar cells with an initial efficiency up to 13.7 % and a stabilized efficiency up to 11.8 % are demonstrated on single-scale textured superstrates. An alternative approach combining large and smooth features nanoimprinted onto a transparent lacquer with small and sharp textures from as-grown LPCVD ZnO is then shown to have a high potential for further increasing TF-Si devices efficiency. First results demonstrate up to 14.1 % initial efficiency for a TF-Si tandem solar cell.

Recent improvements in light-emitting diode (LED) technology has allowed for the use of LEDs for solar simulators with excellent characteristics. In this paper, we present a solar simulator prototype fully based on LEDs. Our prototype has been designed specifically for light soaking and current–voltage (I(V )) measurements of amorphous silicon solar cells. With 11 different LED types, the spectrum from 400 to 750 nm can be adapted to any reference spectrum—such as AM1.5g—with a spectral match corresponding to class A+ or better. The densely packed LEDs provide power densities equivalent to 4 suns for AM1.5g or 5 suns with all LEDs at full power with no concentrator optics. The concept of modular LED blocks and electronics guarantees good uniformity and easy up-scalability. Instead of cost-intensive LED drivers, lowcost power supplies were used with current control, including a feedback loop on in-house developed electronics. This prototype satisfies the highest classifications (better than AAA from 400 to 750 nm) with an illuminated area of 18 cm × 18 cm. For a broader spectrum, the spectral range could be extended by using other types of LEDs or by adding halogen lamps. The space required for this can be saved by using LEDs with higher power or by reducing the maximum light intensity.

We presentanewtechniquetomeasurecomponentcurrent–voltage(I–V) curvesofindividual sub-cellsintegratedinamonolithicmulti-junctionsolarcell.Thisnewapproach,comparedtoall previouslyreportedones,iswellsuitedforthin-filmsiliconp–i–nstructureswheretheso-called shiftingapproximation,whichsupposesthatilluminationonlyshiftsthe I–V curve withoutchangingits shape, isnotvalid.Moreover,theproposedmethodisparticularlyresistanttoproblemsrelatedto electricalshunts.Theprincipleofthismethodliesincouplingthelevelofaselectivelightbiaswiththe level ofmeasuredelectricalcurrentinordertofixthevoltageofaselectedsub-cellwhilesweepingover the currentaxis.Whenoneofthesub-cellshasafixedvoltage,itisthenpossibletogetthe I–V characteristicsofthesecondone,shiftedbyafixedvoltagevalue.Thismeasurementprocedureis simple andrequiresnomodeling.Theaccuracyofthemethodisevaluatedbynumericalsimulationsof a thin-filmsiliconp–i–nphotodiode.Ourtechniqueisthensuccessfullyexperimentallytestedona speciallypreparedthree-terminalamorphous/microcrystallinesilicontandemsolarcell.

Dual-junction solar cells have the potential for higher conversion efficiencies than single junctions, thanks to a better utilization of the solar spectrum. However, their optimization is more complex because the monolithic interconnection makes the individual subcell electrical measurement difficult or impossible. This study presents a new technique to measure the open-circuit voltage VOC of each individual subcell of a dual-junction device with a method that circumvents the need for completely selective illumination. The method is based on the description of the total device VOC as a function of two independent variables determined by two distinct illumination spectra. Without completely selective illumination, one can measure only a limited range of this function; however, an unambiguous extension gives values corresponding to the conditions of ideal selectivity, which correspond in turn to the VOC of the individual subcells at 1-sun illumination. This method does not rely on any specific electrical or optical model of solar cell. Method is successfully tested with a simple measurement setup based on commercially available luminescence diodes and voltmeters in combination with specially prepared three-terminal dual-junction thin-film silicon solar cells.