Skip to main content
European Commission logo print header

Nanophotonics for ultra-thin crystalline silicon photovoltaics

Final Report Summary - PHOTONVOLTAICS (Nanophotonics for ultra-thin crystalline silicon photovoltaics)

Executive Summary:
Crystalline silicon is the most widespread and preferred material for photovoltaics. Earth-abundant, non-toxic and stable, it presents many assets for a large-scale deployment of solar electricity. It has however an intrinsic drawback, its rather weak absorption of light, particularly of lower-energy photons in the infrared spectrum. This requires the usage of rather thick wafers of silicon, that contribute to a significant share of the final module cost and especially of energy consumption for their production. An efficient way of trapping these low-energy photons would enable making thinner silicon cells as “thick” optically as standard wafers, hence reaching high energy-conversion efficiencies with thinner/lower-cost solar cells.
Texturing of silicon at the sunlight-wavelength scale is known to have such potential of boosting the light absorption of thin crystalline silicon films beyond what can be achieved with standard larger-scale textures. The aim of PhotoNVoltaics was to learn which nanotexture geometries are optically optimal and how to integrate them into thin films of crystalline silicon without damaging their electrical performance. The project brought together 7 partners from 3 different fields – photovoltaics, photonics and nanolithography – and from 3 different worlds – R&D, academia and industry – to study and achieve in 3 years the conditions for a successful integration of photonic nanotextures into solar cells.
The consortium investigated, in tight interaction loops of modelling and experiments, which photonic nanotextures (with which dimensions, shapes, pattern) work best as front-side textures of thin-film crystalline-silicon solar cells. The nanotextures were evaluated not only optically but also electrically, to assess and understand their impact on the solar cell functioning and to find how to mitigate their negative impacts when necessary. Furthermore, periodic, pseudo-periodic and random textures were compared. The results confirm that for light-trapping purposes, periodic structures outperform random textures, but that the controlled disruption of periodicity is beneficial for the optical performance. Interestingly, this means that some tolerance on nanopattern defects is allowed during fabrication. Eventually the project has led to the development of a set of powerful tools for modelling realistic solar cells with a very broad range of patterns, and it defined guidelines for efficient integration of nanophotonic textures into thin-film crystalline silicon solar cells. This understanding was demonstrated by the achievement of ultra-thin cells with boosted short-circuit current densities, without any losses in their electrical properties, resulting in more than a doubling of their energy-conversion efficiency. Furthermore, our results also show that if boosting light absorption beyond the Lambertian limit may be possible in optimised optical demonstrators, it is far more challenging in functional solar cells, where a significant part of the highly-absorbed light is lost in non-active parts of the device, that are however indispensable for the cell functioning. Finally, not only the technical aspects but also economical aspects were studied, to identify the current bottlenecks for deployment into industry of 3 different nanolithography techniques.

Project Context and Objectives:
Over the past few years, many studies have discussed how the “Lambertian limit” of light trapping may be overcome by texturing crystalline silicon (c-Si) solar cell surfaces at the light wavelength scale, rather than at the micron-scale. In fact, various simulations of periodic or non-periodic nanopatterns indicate that significantly higher photon absorption may be achieved compared to micron-scale texturing, especially for absorber thicknesses below 100 µm. However, fewer publications present solar cells, and even fewer present solar cells with clearly higher energy-conversion efficiencies, thus showing that higher photon absorption does not readily translate into a higher electron collection. Nanopatterns are in fact challenging to integrate, because of the electrical degradation they introduce, and because of the optical degradation that they may undergo in later process steps. This integration challenge is yet inevitably to be overcome if thin c-Si foils are set to eventually replace wafers. This requires (i) modelling in order to identify the nanopatterns that provide the best photon absorption, (ii) fabricating these nanopatterns under the various experimental constraints set by the materials, and (iii) integrating such nanopatterns into solar cells without compromising their electrical performances (with the highest short-circuit currents and open-circuit voltage possible). Finally, suppressing the hurdle of light-absorption losses in thin film c-Si solar cells requires (iv) assessing the boundary conditions for an industry-compatible nanopatterning process. Those requirements are the objectives of PhotoNVoltaics.
PhotoNVoltaics is a “Future Emerging Technologies” FP7 project. In other words, it is a high-risk project with ambitious targets and a long-term objective, with a multidisciplinary approach. In fact, solving the issue of an efficient integration of surface nanopatterns into solar cells requires bringing together experts from the diverse fields of photonics, nanopatterning and photovoltaics. It also associates partners from diverse contexts, from the industrial world (SME and multinational), R&D and academia. For the three running years of this project, the seven partners of PhotoNVoltaics have defined the following targets:
1) Fabricate a thin monocrystalline-silicon-based sample in which light absorption surpasses the commonly accepted Lambertian limit of Yablonovitch. The fabrication of optical demonstrators would, at the same time, answer an old scientific question: can nanoscale periodic texturing surpass random texturing?
2) Fabricate the solar cell with the highest improvement in short-circuit current (Jsc) ever reached so far for the wafer-based c-Si technology, with a thickness lower than 40 µm.
3) Identify the conditions of integration of thin c-Si layers and nanopatterning techniques into the PV industry.

Project Results:
see attached pdf
Potential Impact:
see attached pdf
List of Websites:
www.photonvoltaics.org
Coordinator and imec Dr. V. Depauw, valerie.depauw@imec.be
CNRS-INL Prof. Dr. E. Drouard, emmanuel.drouard@ec-lyon.fr
CNRS-LPICM Prof. Dr. P. Roca i Cabarrocas, pere.roca@polytechnique.edu
UNamur Prof. O. Deparis, olivier.deparis@unamur.be
Obducat Dr. R. Jiawook, rizgar.jiawook@obducat.com
Chalmers Prof. Dr. A. Dmitriev, alexd@chalmers.se
Total Dr. P. Prod’Homme, patricia.prodhomme@total.com