Jan. 14, 2016


Trapping Light in Ultra-Thin Cu(In,Ga)Se2 Solar Cells

  • Fig. 2: Cross section of an ultra-thin Cu(In,Ga)Se2 (CIGSe) solar cell with SiO2 nanoparticles on top of the molybdenum (Mo) back contact with conformal over-growth by the other layers; image taken by scanning electron microscopy.Fig. 2: Cross section of an ultra-thin Cu(In,Ga)Se2 (CIGSe) solar cell with SiO2 nanoparticles on top of the molybdenum (Mo) back contact with conformal over-growth by the other layers; image taken by scanning electron microscopy.

Authors: M. Schmid1,2, M.-C. van Lare3, G. Yin1, A. Polman3

The world’s growing energy demand is one of the greatest challenges our society is currently facing. Fossil fuels have limited resources and the related CO2 emission contributes to substantial climate changes. Renewable energies are therefore of highest importance to ensure world energy supply and to help keep our planet clean.

Abundant Energy from the Sun
The sun is providing us with a nearly infinite amount of energy. Within two hours the sun is sending more energy to earth than we consume within one year. The exploitation of solar power has attracted growing interest in recent years and photovoltaics (PV) already make a significant contribution to energy production in industrialized countries. To increase competitiveness of PV on the market, high efficiencies at low costs have to be achieved. Reducing material consumption of solar cells has lead from the first generation of wafer-based devices to the second generation of thin-film solar cells. Amongst these, solar cells with absorber material made from copper, indium, gallium and selenium, so called chalcopyrite (Cu(In,Ga)Se2, short CIGSe), hold the current stabilized record efficiencies. Besides the high efficiency and good stability, these solar cells however come with the challenge of containing rare and expensive indium. Reduced absorber thicknesses are therefore desired.

Cu(In,Ga)Se2 Solar Cell Ultra-Thin
Generally, a CIGSe solar cell is built up as follows: a substrate, e.g. glass, is coated with molybendum (Mo) as a back contact onto which the CIGSe absorber layer is deposited. On top of the absorber a cadmium sulfide (CdS) buffer layer and a stack of intrinsic and aluminum-doped zinc oxide (i-ZnO and AZO) layers are deposited to form the electric junction and to extract the generated charge carriers. The inset on the bottom left of figure 1 shows such a schematic device cross section. In the standard configuration the absorber layer has a thickness of 2 – 3 µm and basically all of the incident photons with an energy larger than the band gap (i.e.

with a wavelength smaller than about 1.2 µm or 1200 nm) are absorbed. The light to electricity conversion is limited such that one photon can at best generate one electron-hole pair. The external quantum efficiency (EQE), which gives the ratio of the collected electron-hole pairs to the incident photons, therefore is always smaller than one and the difference to one reflects losses. These losses can be of electrical nature or due to incomplete absorption of the incident photons. The black curve in figure 1 shows the quantum efficiency of a CIGSe solar cell with an absorber layer of less than 500 nm thickness. At longer wavelengths we observe a significant drop showing that photons of these wavelengths are no longer efficiently absorbed in the ultra-thin CIGSe. We therefore aim at increasing the absorption which requires effective means of light trapping in the ultra-thin CIGSe.

Trapping Light with Tiny Structures
Enhancing the absorption probability can be achieved by extending the effective path length of light travelling through the absorber layer. Textured surfaces and interfaces are one way for increasing scattering and prolonging the light path. Nanoparticles are even more promising since they can redirect light to even larger angles leading to efficient light trapping in thin layers. In addition to their scattering abilities, nanoparticles can also locally enhance the electric field strength. By concentrating light in their vicinity they open up another opportunity to enhance light absorption in very thin layers. The effects of light scattering and concentration are known for metal nanoparticles under the terminus “plasmonic effect”. They relate back to the collective movement of free charges which are present in metals. Yet, light scattering and concentration can also be achieved by nanoparticles made from dielectrics, like for example oxides. These dielectric nanoparticles come with the big advantage of having no parasitic absorption, which means they are lossless compared to their metallic counterparts. Furthermore, for the case of inorganic materials, they are characterized by high chemical and thermal stability. This stability is particularly beneficial when integrating them into devices like CIGSe solar cells where high process temperatures determine the fabrication. An isolating material finally has the benefit of avoiding recombination when integrated into the active layers of a solar cell. We therefore focus on the integration of dielectric nanoparticles into ultra-thin CIGSe solar cells to enhance the absorption.

Nanoparticles for Current Increase
At Helmholtz-Zentrum Berlin (HZB, Germany) a CIGSe solar cell with absorber layer thicknesses of 460 nm and an efficiency of 11.1% could be fabricated. The short circuit current density, which is a measure for the current generated from the absorption of photons, amounts to 28.2 mA/cm2 for this device. The value compares to 36.6 mA/cm2 for the thick record CIGSe solar cell. The Young Investigator Group NanooptiX at HZB is investigating how to enhance the short circuit current density by integrating nanostructures with sizes of the order of the wavelength. In cooperation with colleagues from the Center of Nanooptics at AMOLF, Netherlands, nanoparticles fabricated by nanoimprinting were investigated. Nanoimprint lithography is based on patterning a template layer with a “stamp”, which can then be used as a mask for growing nanoparticles. Firstly, titanium oxide (TiO2) nanoparticles were fabricated at AMOLF on top of a completed ultra-thin CIGSe solar cell. The device slightly improved, but the effect was mostly due to reduced reflection when introducing the TiO2 nanoparticles. Therefore, in a second attempt, silicon oxide (SiO2) nanoparticles were directly integrated at the back contact of the CIGSe solar cell. The schematic cross section of SiO2 nanoparticles at the Mo/CIGSe interface is represented in the inset at the top right of figure 1. The realistic cross section taken with a scanning electron microscope is shown in figure 2 (eye catcher). The nanoparticles depicted in black, which were printed onto the Mo back contact (purple), are conformally over-grown by the CIGSe absorber (red), followed by the other layers. The quantum efficiency measured on this device is given by the red curve in figure 1. Compared to the black curve for the case without nanoparticles a significant increase in the wavelength range from 700 to 1100 nm can be observed. This enhancement in absorption translates to an increase in short circuit current density by 2 mA/cm2. The efficiency increased to 12.3%. By additionally coating a layer of anti-reflection nanoparticles a further increase to 13.1% efficiency could be obtained and the short circuit current density was enhanced up to 32.1 mA/cm2. Simulations show that the effective absorption has the potential to be further improved by optimizing the geometry of the SiO2 nanoparticles.

Novel Concepts Benefit from Nanostructures
A detailed investigation of the enhancement assisted by three-dimensional optical modelling reveals the absorption in each individual layer. In this way not just the enhanced absorption in the CIGSe can be retraced but also the remaining losses identified. Despite the improvement, a significant part of the light is still lost in the poorly reflecting molybdenum back contact. A next step therefore is to move to transparent back contacts made from so called TCMs, i.e. transparent but conductive materials. The reduced absorption losses can be exploited by reflecting transmitted light back into the solar cell where it finally will be absorbed. Alternatively, the transmitted light can also be used for other purposes. In a tandem solar cell, a combination of two absorber materials with different band gaps, light that cannot be absorbed by the upper (top) cell is transmitted to the lower (bottom) cell where it will be converted. Such a device will benefit from an ultra-thin top cell which passes sufficient light to the bottom cell for it to work efficiently. At the same time, light that can be converted by the top cell has to be trapped there. This is where the wavelength selective activity of nanoparticles plays an important role to achieve optimum light distribution between the two solar cells. Improved exploitation of the solar spectrum by concepts beyond planar single junction thin-film devices is what brings us to the third generation of photovoltaics. Increased efficiency and reduced material consumption ask for novel device concepts where nanostructures play an important role.

1Nanooptische Konzepte für die PV, Helmholtz-Zentrum Berlin für Materialien und Energie GmbH, Berlin, Germany
2Department of Physics, Freie Universität Berlin, Berlin, Germany
3Center for Nanophotonics, FOM Institute AMOLF, Amsterdam, The Netherlands

Martina Schmid

Nanooptische Konzepte für die PV
Helmholtz-Zentrum Berlin für Materialien and Energie

More articles on:

PV efficiencies since 1976:

Reference publication:

M.-C. van Lare*, G. Yin*, A. Polman, M. Schmid “Light coupling and trapping in ultra-thin Cu(In,Ga)Se2 solar cells using dielectric scattering patterns” ACS Nano 9, 9603 (2015), DOI: 10.1021/acsnano.5b04091, *equal contribution

Register now!

The latest information directly via newsletter.

To prevent automated spam submissions leave this field empty.