In the first 18 months, the BURST project has been running excitingly well. All partners have been working with great dedication on their tasks and collaborating with each other, combining their expertise and capabilities, as well as shipping numerous samples to partner institutes. Work package meetings and general assemblies are well attended and filled with lively discussions of results.
This drive of the consortium has enabled the project to progress well towards BURSTs goals. The most important finding so far is unfortunately also a major setback for the project:
When setting up the project, great hope was placed in inverted pyramid photonic crystals that were promised to enable extraordinarily high absorption, especially of near infrared light. While BURST succeeded in fabricating inverted pyramid structures with excellent quality, the expected high absorption could not be observed. Similarly, careful optical simulations of these structures were not able to reproduce the strong benefits that were initially predicted in the literature and are currently also disputed by independent researchers.
While this is a setback for achieving BURSTs ambitious goals, these observations are an important contribution to the scientific debate on the viability of photonic crystals for light trapping in silicon solar cells.
Achieving less spectacular but still above state of the art optical properties of solar cells can be achieved with various alternative methods, with BURST making progress using three different approaches, that were investigated with device simulations and on test structures, and partly already implementations on solar cells:
A black silicon approach that adds a nanoscale surface roughness to a microscale alkaline etched random pyramid texture, shows great potential optically, but due to the large surface area and high aspect ratio of its features is the most difficult approach to passivate. Using atomic layer deposition, implied open circuit voltages of 735 mV have been achieved, though.
“White silicon” is a similar approach, that combines a microscale random pyramid surfaces with nanoscale features. In this case the nanoscale features are present in a layer of porous SiO2, that can be deposited on top of the passivation layer and anti-reflection coating, thus compared to the black silicon a smaller surface area needs to be passivated and there are less restrictions for the deposition method.
Further reducing the electric surface area is possible with a photonic light trapping structure consisting of a high refractive index TiO2 layer with semi-hemispherical nanovoids, that can be applied on a flat Si surface, but have still been shown to enable large scattering angles for light, and thus showing comparable optical performance as standard textures.
Another major objective of BURST is the reduction of silver usage, again multiple approaches are investigated in BURST:
While the baseline cell process relies on PVD Al, that enables excellent contact to p- and n-type poly-Si contacts without affecting the passivation properties, more industrially viable methods are followed in the form of copper plating, and screen printing.
For the former a deposition and structuring process for a silver-free seed layer has been developed, and fingers with an aspect ratio of 1 have been plated.
For the latter, a silver paste is screen printed for achieving low contact resistivities, followed by copper screen-printing for high finger conductivity. While this approach is not silver free, it greatly reduces silver consumption, relying only on mainstream solar cell fabrication methods. Contact resistances of 2 mΩcm² have been achieved on p- and n-type poly-Si surfaces.
In order to evaluate the developed optical, passivation and metallization building blocks, a high efficiency solar cell process flow has been devised that allows the implementation of single or multiple building blocks. With baseline solar cells, independently confirmed conversion efficiencies of 26 % have been achieved, losses analyzed, and steps for further improvements identified.
Aiming to upscale the lab-type structuring for inverted pyramids, two approaches have been evaluated: A laser beam splitting optic has been developed that enables the fast patterning of etch masks, by allowing to open 667 openings using a single laser pulse. The module has been tested in a laser system, showing good uniformity over typical cell sizes and the potential for patterning rates of 0.015 m²/min. Further, colloidal lithography of polystyrene spheres has been shown to be a possible method to mask atmospheric dry etching of silicon.
To evaluate the environmental impact of the developed technologies, process data has been collected and initial life cycle assessments have been conducted. Further, a criticality assessment, with focus on Ag substitution has been started.
Project results have been disseminated via peer-reviewed publications, and various conference publications. A wider audience has been targeted with an online workshop, a LinkedIn campaign and setting up a project website.
