H. Nussbaumer

Band‐band impact ionization and solar cell efficiency

The effect of impact ionization has been taken into account in the calculation of the maximum solar cell efficiency in the thermodynamic limit. A red shift of the optimum band gap is observed with respect to the Shockley–Queisser result. A maximum solar cell efficiency of 60.3% at Eg=0.8 eV is predicted, which compares with 43.9% at Eg=1.1 eV for the usual case. In order to obtain a significant increase in η, an impact ionization probability of at least a few percent is required. A most pronounced effect is observed for bandgap energies below about 1 eV.

A simple and effective light trapping technique for polycrystalline silicon solar cells

Abstract Mechanical grooving using a standard dicing saw in combination with bevelled blades is shown to be a promising texturing technique for polycrystalline silicon surfaces. A minimum total reflectance of R = 5.6% at 950 nm and an average R = 6.6% between 500 and 1000 nm have been obtained on SILSO silicon. The reflection coefficient obtainable is limited by blade tip curvature and a non-perfect surface after damage etching.

Polycrystalline silicon solar cells with a mechanically formed texturization

Polycrystalline (Wacker SILSO) silicon has been mechanically textured using a conventional dicing saw and beveled blades for V‐groove formation. The minimum optical reflectivity achievable is limited by the blade tip radius and surface roughness after damage etching. Solar cells were prepared using a conventional diffusion and screen printed metallization. Grooved cells without an additional antireflection coating (jsc=31.8 mA/cm2, Voc=536 mV, FF=69%, η=11.8%) showed a 20% increase in jsc and a 1.1% absolute efficiency improvement as compared to a nongrooved reference cell with an antireflexion coating (jsc=26.4 mA/cm2, Voc=547 mV, FF=74.1%, η=10.7%). In grooved cells the efficiency is found to be limited mainly by the fill factor due to a nonoptimized front grid design.

Industrial LCP selective emitter solar cells with plated contacts

The investigation of different selective emitter (SE) approaches [1–3] is a current trend in solar cell manufacturing. The incorporation of a local high phosphorous doping underneath the front contact grid allows for the use of high-sheet resistance illuminated emitters that combine low recombination and improved blue response. Further efficiency increase compared to the standard screen-printed solar cell is achieved via plated contacts [4–5] that feature better aspect ratio and optical properties [6], higher line conductivity and smaller width [5] compared to screen-printed contacts. In this paper we present detailed technological requirements for next-generation front side metallization as well as experimental results of the RENA high-efficiency metallization cluster consisting of Laser Chemical Processing (LCP) and Ni-Ag light-induced plating (LIP). It becomes clear that efficiency on cell level is not the only figure of merit for a successful product and that the combination of SE with plating has a much higher potential for increasing cell efficiency than the metallization of SE via screen-printing.

Optical behavior of textured silicon

The optical behavior of polycrystalline SILSO silicon wafers with a mechanically V‐grooved surface has been studied between 300 and 1500 nm. The texturization was carried out by a conventional dicing saw using beveled blades. For a 35° V‐grooved surface and a nonmetallized backside the optical path length in the weakly absorbing part of the spectrum (1100–1200 nm) was found to be enhanced by a factor of 33 as compared to a nongrooved wafer. The enhanced reflectance in the nonabsorbing spectral region for the former is analyzed and explained. The different loss contributions due to a nonideal grooved structure are discussed.

Hall mobility minimum of temperature dependence in polycrystalline silicon

Molten zone recrystallized as well as sheet grown polycrystalline silicon has shown a minimum in the temperature dependence of the Hall mobility. In order to explain this experimental finding a new model is proposed, which is based on negatively charged grain boundaries for the p-type silicon material under study. This results in a potential well at the grain boundaries instead of the more generally observed potential barrier. A key feature in the model is that the space charge density at the grain boundary depends on the Fermi level position and therefore on temperature. In addition, the change in the measured Hall mobility before and after hydrogen passivation of the grain boundaries is discussed.

PV Installations Based on Vertically Mounted Bifacial Modules Evaluation of Energy Yield and Shading Effects

Bifacial solar modules promise an increased energy yield, compared to systems with standard, monofacial panels, and also offer new opportunities with regard to the installation. One particular approach is the vertical mounting of PV modules, which is reported to be an effective measure to avoid soiling or dust deposition and is an option to obtain a broadened energy generation profile. In spite of the general interest in this type of installation, the amount of published data is very limited, especially with regard to arrays, for which pronounced shading effects can be expected. In this work we present an analysis of the energy yield and the respective losses for arrays of vertically mounted bifacial solar modules with varied installation conditions.

THE SWISS PV WALL SYSTEM TO MAXIMISE SELF-CONSUMPTION IN A SINGLE BUILDING ELEMENT

With decreasing subsidies for PV energy fed into the grid other economic incentives are needed to sustain the growth of PV power. An effective solution is to increase self-consumption for grid-connected, residential PV systems by means of generating hot water by the use of a heat pump. This widely used concept of thermal energy storage has been adapted for a facade element with integrated PV modules for the first time. The so called Swiss PV Wall System (SPWS) contains all elements needed in the facade, behind the PV modules with a power of 1 kWp, like a PV inverter, a heat pump and a hot water tank. Simulation showed that virtually 100% of the PV energy can be used either by the heat pump to produce hot-water or the households electricity needs. A functional model of the SPWS was realized and first measurements showed that the heat demand of a one family household may be covered for the majority of the months during the year. Economic analyses showed that the cost per kWh heat is at about 0.215 CHF and thus lower than standard solar thermal and some fossil solutions.

Cleaning in Crystalline Si Solar Cell Manufacturing

The large scale of production of modern PV manufacturing as well as the cost pressure demand a different approach to cleaning processes in semiconductor and PV applications. The subject of this presentation is, to highlight aspects of similarities and differences. Total added cost/m2 of Si are estimated for typical PV manufacturing conditions and compared to semiconductor applications. Typical technical solutions are reviewed. They are compared to the anticipated technical and cost requirements in the near future according to PV roadmaps and cell concepts which are evident today. Starting with typical cleaning processes during the wafering (sawing, cleaning, separation) up to cell processing (texturing, diffusion, coating and plating) the main cleaning processes are presented and their specifics are indicated. Finally recontamination and conditioning in production lines are reviewed.

Application of SunsPL for fast laser chemical process development

Selective emitters will soon become a standard feature of industrial silicon solar cells. RENA offers a laser chemical process (LCP) to locally ablate the antireflection coating and form a highly doped emitter at the opened area. The front side metallization is then applied by light induced plating (LIP) of nickel and silver. However, it is still a challenge to optimize this process sequence for precursors coming from different production lines. For this purpose we use photoluminescence imaging (PL) as a valuable tool to reduce the impact of the quality spreading of the precursors on the evaluation of our experiments. With PL it is possible to calculate an implied open circuit voltage or dark saturation current as well as a pseudo fill factor before metallization. Instead of comparing different groups of an experiment only by their final cell parameters, we compare the quality parameters obtained with PL after each process step. Like this we can exactly quantify the impact of each production step on the final solar cell performance. In this paper we will demonstrate how using PL improves the significance of our experiments and allows fast process optimization with a small amount of wafers. In general this procedure for process control can also be applied for other process steps.