Cell Processing

Amorphous silicon is one of the most effective materials in passivating silicon interfaces. At Fraunhofer ISE, highly passivating amorphous silicon coatings were developed by an industrially applicable Plasma-Enhanced Chemical Vapour Deposition (PECVD) process. Thin-film stacks of amorphous silicon and SiOx display excellent passivation quality, indicated by effective charge carrier lifetimes ranging from 900 to 1600µs and resulting surface recombination velocities between 9 and 3cm/s-1. The demonstrated temperature stability opens up new application opportunities also for amorphous silicon films in the industrial production of highly efficient solar cell structures, which will be further discussed in this paper.

The principal paths to cost reduction for the photovoltaics industry are increasing the efficiency of solar cells and the power density of modules, together with the reduction of the specific consumption of silicon. Following the slowdown in the ever-increasing growth of the PV market earlier this year, and the reduction in the market cost of polysilicon, wafer producers and most cell producers moved back to the 180µm generation substrates. It may take some time for manufacturers to tackle the technological issues that need to be addressed in order to successfully decrease wafer thickness further. In this article, some of the issues related to the production of thinner and thinner cells are outlined and discussed.

Crystalline silicon solar cell fabrication involves many wet chemical process steps. Like most processes in solar cell manufacturing, many of these wet chemical processes were transferred from the semiconductor industry. In contrast to microchip fabrication with maximum throughputs of 100 wafers/hour, state-of-the-art solar cell equipment relies on several 1000 wafers/hour. Furthermore, specific processes have been developed for the texturisation of the wafer surface. Therefore, there is a need for dedicated methods of characterization of these wet chemical processes. Fraunhofer ISE has developed several analytical methods such as titration, ion chromatography and near infrared (NIR) spectroscopy for the complete analysis of the chemical composition of wet chemical processes baths. These methods were compared considering the inline/online capability, measurement cycle and running costs, with the result that NIR spectroscopy was identified as a complex but very powerful tool for process characterization, as outlined in this paper.

Formation of the pn-junction for charge carrier separation is one of the key processes of a modern high-volume solar cell production. In silicon wafer-based solar cell technology this is achieved by diffusion of phosphorus atoms in boron pre-doped wafers forming a sub-micron shallow n-type emitter in a 200µm-thick p-type base. In this contribution we discuss both the characteristics of emitter doping profiles and the diffusion process itself as required for optimal solar cell conversion efficiencies. In addition we give an overview on state-of-the-art industrial diffusion technologies and conclude with a brief outlook on their evolution.

In any solar cell process, the metallization step is critical as it often sets conditions and limitations for the other process steps. The main metallization technique used today in Si solar cell production is screen-printing of metallic pastes, namely Ag pastes for the front side, Al pastes for most of the rear side, and Ag or Ag-Al pastes for the solder pads at the rear. While these techniques are clearly robust and convenient, they have limitations. Therefore alternatives are being investigated. A technique that is presently finding its way into production is two-step metallization with Ag plating. Another more radical approach is to avoid printing altogether, instead using some kind of ablation followed by plating. For the rear, the full Al-BSF is being replaced by dielectric passivation and local Al-alloyed contacts. Back-contacted cells are increasingly being introduced in production, and they pose very specific challenges to metallization. For the sustainability of Si photovoltaics, it is crucial that the future metallization solutions only make use of abundantly available and non-toxic materials.

The use of perfluorinated gases such as NF3, CF4 or SF6 for PECVD (plasma enhanced chemical vapor deposition) chamber cleaning has a much higher impact on global warming than does the use of onsite-generated F2. This holds true even when supposing that in the future much more effort is paid for the correct abatement and a leak-free supply and take-back chain. This paper will discuss the steps available to the PV industry for control and reduction of carbon emissions in the chamber cleaning process.

The etching technology currently used in the solar industry is mostly based on wet chemical processing. Plasmaenhanced dry chemical etching at atmospheric pressure is an alternative to the existing technology, especially when combined with similar process technologies, for example plasma-enhanced deposition techniques at atmospheric pressure, to provide a continuous in-line processing of crystalline silicon solar cells. This paper presents the use of plasma chemical etching using Fourier Transform infrared (FT-IR) spectroscopy to monitor different silicon wafer processing steps as an alternative to the widely used wet chemical processing approach.

Inline processing, one of the fastest-growing production processes for crystalline silicon solar cells, uses continuously operated belt furnaces to achieve higher overall throughput compared with traditional batch processing. A second, major advantage of inline processing is improved manufacturing yields through reduced breakage of today’s thinner, increasingly delicate wafers. This is accomplished by eliminating several handling steps unique to batch processing techniques. This paper describes the influence of ECN-Clean, as developed by Mallinckrodt Baker and ECN in 2006, whose application increases the efficiency of solar cells produced using inline processing by approximately 0.3 percent absolute, compared with standard inline processing.

A vast majority of silicon solar cells are manufactured using silver paste that is screen printed onto the front side of the wafer and fired to form the front-side contact. Though this method is well established within the industry, it continues to present several areas for potential efficiency improvements. The Fraunhofer Institute [1] has, among others, studied the potential of using electrodeposition of silver on top of the front side silver paste as a way to improve the front-side contact and increase cell efficiency. These results have shown cell efficiency increases of up to 0.4% absolute. This type of improvement has captured the interest of many manufacturers, but there has been a hesitancy to adopt electrodeposition as there is uncertainty as to what they can expect on their cells. Since efficiency gains are dependent upon many factors that can be unique to an individual cell, this paper provides a much-needed exploration of the potential effects of electrodeposition of silver in a way that isolates its effects from that of other factors.

The deposition of thin films is a key technology for a large variety of technical and scientific applications. Among them is the deposition of silicon nitride (SiNx) to passivate the surface of silicon solar cells. The SiN film serves several purposes. It is a broadband anti-reflection layer, it serves to saturate dangling bonds and/or other surface states of the silicon, and last but not least, it is a protection layer to prevent alkali ions and other impurities from diffusing into the silicon causing perturbations of the performance of the solar cell. This multitude of properties to be fulfilled at the same time often causes difficulties in assessing the effect of a single process parameter, let alone the task of optimizing the SiN film in all required aspects at the same time. The aforementioned technical features of the SiN film provide the very property that largely determines the aesthetically pleasing appearance of a cell, and hence a PV module, as the colour of the module is determined by the cell composition. In order to complicate things further, there are numerous deposition techniques being applied both on a scientific level as well as in production environments.