Light and Elevated Temperature Induced Degradation of Silicon Solar Cells

Lead Research Organisation: University of Manchester
Department Name: Electrical and Electronic Engineering

Abstract

The importance and urgency of reducing carbon dioxide emissions has received much publicity. Electricity generation is responsible for 38% of carbon emissions world wide. Of all sources of global warming electricity generation is probably, technologically, the most easily replaced by carbon free sources. Electricity from sunlight using the photo-voltaic effect, which we will refer to as solar PV, was very much a niche application as little as 15 years ago. However in the last decade silicon solar PV technology has developed with astonishing speed so that today it is the cheapest form of electricity generation in most countries within 45 degrees of the equator. Equally importantly the cost of manufacture is decreasing by 24% for each doubling of production volume, much faster than most products.

At the moment Solar PV provides only 2.6% of the world's electricity (in kWh) although a higher percentage in some countries (eg 7.9% in Germany, 5.4% in India). There are a number of factors which delay the take up of this technology. The biggest difficulty is intermittency in countries like the UK where peak load does not match peak solar output necessitating pumped storage hydro or other rapid start up generation which adds to the cost. In tropical and sub-tropical countries solar generation matches the load much better and it is these countries in which electricity demand is increasing most rapidly. However in general there is a reluctance to invest in Solar which in part is due to Solar being regarded as an unproven technology and questions regarding long term reliability of a capital intensive system with a costing based on a projected life of >25 years.

It is well known that silicon solar cells degrade. There are two commercially important mechanisms. One is due to a reaction involving boron and oxygen which happens very quickly reducing the efficiency by ~2% in the first 24 hours of operation. This is well enough understood for specialists to be on the way to developing ways of minimising the effect and demonstrating stability. The other mechanism is called "light and elevated temperature degradation" (LeTID). It takes months or sometimes years to produce a degradation of between 2 and 5%. The higher the light intensity and the higher the temperature the faster the degradation although there are large variations between different materials and solar cell designs which are not at all understood despite much behavioural data.

The aims of this project are to develop a fundamental understanding of the degradation mechanism, to test proposed methodologies for reducing or eliminating LeTID and to use our understanding of the degradation mechanisms involved to develop meaningful accelerated life tests. Experimental work will be done in Manchester using test devices fabricated by us in Manchester and by the University of New South Wales (Australia). The prime techniques used will be optical, chemical and electrical measurements in Manchester and the Australian National University (Canberra) supported by modelling work at the University of Aveiro (Portugal). These will include lifetime spectroscopy, Deep Level Transient Spectroscopy and variants, admittance spectroscopy, low temperature photo-luminescence, time resolved photo-luminescence, Raman spectroscopy, hydrogen measurements and Secondary Ion Mass Spectroscopy. Materials and devices samples will be supplied by two manufactures active in the silicon solar field.

Planned Impact

Electricity generation and ground transport which, in principle, could be electrified, constitute 60% of carbon dioxide emissions originating from human activity. Solar silicon PV already provides the cheapest electricity in regions with high insolation embracing most areas within 45 degrees of the equator. However today only 2.6% (measured in kWh delivered) of electricity is generated by PV. Substantial adoption of solar PV for future capacity is constrained by intermittency in some countries particularly where peak load does not match peak output but a key factor is the concern regarding long term reliability and unknown degradation in the field over the projected 25-30 year life. No real time life data exists for modern cell designs and accelerated life tests lack consistency. It is in relation to these issues that the work proposed in this contract will have its most important impact.

The cost of PV is expected to follow the steep learning curve of the last 38 years (price decreasing by 24% for each doubling of production) for at least the next decade. The two percent increase in efficiency that this work promises is important in terms of cost but the key benefit of understanding degradation mechanisms is because it instils confidence in accelerated life testing and may well open a route to eliminating degradation over the system lifetime. The expected life of a PV system based on PERC cells is 30 years however because the fundamental cause of the LeTID degradation mechanism is unknown and real time experience is very limited, financial institutions rate the capital investment in PV solar "high risk". The result is that financing is the biggest single cost of a 25 year industrial scale solar system. Successful completion of this project will contribute very significantly to confidence in the durability of Si PV and overcome a major barrier to wider take up as well as lower financing costs.

The impact of understanding degradation within the PV industry is very considerable. On the 18th of May 2019 we published "Identification of the mechanism responsible for the boron oxygen light induced degradation in silicon photovoltaic cells" in J. Appl Phys. The Journal publicised it as a "feature" article and subsequently identified it as one of the most "read" papers of 2019 having 10,149 downloads during the last two weeks of May. Ensuing correspondence showed that it aroused worldwide interest from both manufacturers and installers as well as the scientific community and has great potential impact. In our view understanding LeTID will be of even greater significance to the community.

It is our expectation that aspects of this work will be patentable. That together with presentations at IEEE PV and EU PVSEC conferences and publication in scientific journals and trade magazines will be routes to advertise the work and facilitate impact. We have as industrial sponsors GCL Solar (Portland USA) and TNO/ECN (Netherlands) and they, together with our consultant Bob Falster, will guide us to industrial beneficiaries.

Academic impact is dealt with in the Academic Beneficiaries section of this form
 
Description The project has been badly affected by COVID but we have started to interact with our industrial partners, in particular sun-edison solar who have provded a hugh selection of solar silicon wafers. We have found that indium and gallium doped solar cells do not degrade and the company is considering adapting their processes to prevent degradation
First Year Of Impact 2020
Sector Energy
Impact Types Economic

 
Description Collaboration with ANU 
Organisation Australian National University (ANU)
Country Australia 
Sector Academic/University 
PI Contribution The photovoltaic group at ANU have been sending us samples processed so as to demonstrate the LETID effect
Collaborator Contribution This is an important collaboration as LETID is a noriously difficult effect to pin down and ANU have a world lead in producing materials with a consistent LETID effect evident in them
Impact one joint publications listed so far (with Ziv Hamieri)
Start Year 2021