CORDIS provides links to public deliverables and publications of HORIZON projects.
Links to deliverables and publications from FP7 projects, as well as links to some specific result types such as dataset and software, are dynamically retrieved from OpenAIRE .
Exploitable results
There are a number of publications that have been claiming for years that the contribution of UV light enhances diffusion mechanisms in Si. Without precise explanations about the claimed photoelectrical interactions explaining this mechanism significantly enhanced (and therefore faster) diffusion processes are claimed for identical processing temperatures.
In order to understand if UV radiation could effectively enhance diffusion, the FLASH consortium decided that it would be worth to have a systematic view to this topic. Jipelec offered the possibility to execute experiments in a prototype RTP reactor that is equipped with additional excimer lamps and that allows to switch on these UV lamps independently from the conventional lamps.
After careful and systematic planning of experiments between FhG-ISE, CEA Genec and RWE Schott Solar an experimental matrix on etch- polished, mirror-polished and double-side polished Cz-Si wafers was prepared applying the following three diffusion sources: spin-on diffusion from liquid dopants, diffusion from APCVD deposited PSG (phosphor silicate glass) layers with an undoped SiOx capping layer and diffusion from pre-diffused (POCl3) wafers.
The difficult topic of temperature measurement during RTP processing was addressed first to make sure that any comparison between samples that were processed with additional UV light and samples that were processed with the normal spectrum of the RTP lamps would be reproducibly executed at the same temperature. It turned out that open loop processing was mandatory in this experiment to achieve unambiguous results.
The fundamental investigation about the influence of UV radiation on the diffusion process has been executed in a prototype reactor of Jipelec as shown in the schematic below.
The excimer UV lamps emitting at a wavelength of 222 nm are located above the wafer. The power density of UV lamps impinging on the top wafer surface is in the range of 60 mW/cm2. This is in 2 orders in magnitude lower than the power coming from the THLs. However, according to publications of authors that reported about UV-enhanced diffusion mechanisms, this intensity would definitely suffice to observe UV-enhanced diffusion. This is not the case. UV-enhanced diffusion does not occur, and can definitely be dropped.
It was possible to specify the requirements of the RTP prototype reactor and to realize a basic design for the system. This helped Qualiflow-Jipelec and ACR to start the detailed design and construction of the prototype equipment. The detailed specification given by RSS addressed all issues that are relevant for the equipment to be competitive with the latest generation of conventional process equipment and the respective conventional process steps.
In parallel very profound investigations started to identify the process limits of RTP compared to conventional processes. A series of experiments were conducted by RSS and FhG-ISE. Very systematically designed experiments were set up and executed to see the limits of RTD, RTO and RTF.
At the same time this form of co-operation guaranteed that the results of the development are relevant for industrial solar cell production scenarios, provided that the prototype platform AOP would be able to result in the required throughput to be competitive with conventional equipment and provided that the performance of AOP would be comparable or better to the single wafer RTP reactor installed at FhG-ISE. Continuous exchange technical information and informal discussions between the partners allowed to take everyones experience into due account.
These studies identified that it is possible to achieve comparable solar cell efficiencies with RTD than with advanced industrial conveyor belt diffusion furnaces, such as e.g. the furnace running at ET. However, short processing time is required to meet the throughput goals that were set in the project for the prototype RTP reactor.
This may provide a drawback because the diffusion time had to be shortened to less than 15 seconds and therefore the diffusion temperature had to be increased clearly above 950 degrees Celsius, assuming that a high P surface concentration (>1020 P atoms/cm3) and an emitter depth of ~0.3 µm are required to allow for the formation of metal contacts from Ag pastes.
A large number of systematic bilateral experiments between RSS and FhG-ISE showed that diffusion temperatures of 950 degrees Celsius or above are leading to reduced minority carrier lifetimes and solar cell efficiencies for multicrystalline and EFG Si material. This disadvantage for RTP excludes that the prototype test platform AOP could be used in advantageous way for the diffusion of those materials. A different set up of the equipment would be required for that.
However, for mono-crystalline wafers as for instance Cz-Si solar cells it could be shown, that very short processing times (1-5 seconds) are acceptable without degrading the material or solar cell performance. On the other hand a performance advantage could also not be demonstrated yet.
It can be clearly stated that RTP has no advantages with respect to bulk improvement (minority carrier diffusion length) or gettering of impurities. Thus possible performance advantages would have to arise from a better temperature homogeneity during diffusion or better process reproducibility. Other advantages would be a higher throughput or better process yield. The AOP prototype developed in the project was characterized in terms of its potential in terms of processing time and of the transport system.
Rapid Thermal Oxidation was investigated as well in bilateral experiments between FhG-ISE and RSS. Soon both partners agreed also on this topic that RTO or generally thermal oxidation would not be beneficial for simple solar cell processing sequences on multicrystalline or EFG Si, due to the need for bulk passivation in these materials. Mono-crystalline materials like Cz-Si remain instead a good choice for which RTP in AOP might be advantageous over conventional solar cell processing. Further work on AOP might be required to address this topic again. However, also for Cz-Si wafers RTO has to be seen as an additional process step that should bring at least an efficiency improvement of 0.2 % absolute or more to be of economic relevance.
FLASH was the first project to address RTF in detail, performing firing processes in classical single wafer RTP platforms and translating those processes to a concept for a high throughput prototype RTP reactor developed within this project. This is important, considering that industrial conveyor belt firing furnaces do not have the large freedom to shape the temperature profile and gas injection profiles in the same way as RTP systems are able to do that. Thus RTP proved to be a unique tool to study the influence of individual parameters of the firing process on the solar cell performance of the finished solar cell.
Searching for OpenAIRE data...
There was an error trying to search data from OpenAIRE
No results available