Solar Cells Info

by Steve Bush, 20 September 2007
Source: Electronics Weekly.com
http://www.electronicsweekly.com/Articles/2007/09/20/42228/

Durham University has made significant improvements in thin-film solar cells. “We have done a lot of work measuring recombination at grain boundaries and on the control and increase of grain size,” principal investigator Professor Ken Durose told EW.  Durose and his team are working on thin-film cells, three years into the Government’s sustainable power ‘Supergen’ project.

“Our project is geared to trying to make solar energy cheaper,” he said. “That is why we are looking at thin-film. Bulk silicon is too expensive.”  The cells are cadmium telluride-on-glass – a technology which is the subject of research worldwide and is also in production.  CdTe cells require several thin-film semiconducting layers and, being thin-film, the layers are polycrystalline rather than single-crystal.

Multiple crystals mean energy loss. “Electron-hole pairs recombine at the crystal boundaries,” said Durose.   Of particular interest to Durham is the ‘window’ layer – a thin cadmium sulphide layer generally used in front of the CdTe to catch carriers that would otherwise be lost at the CdTe surface.

The University has looked at the basic mechanism by which atoms add as the cell is constructed.  “We have developed a process to control nucleation,” said Durose. “You want grains to be big, so nucleation density has to be low, but final coverage high.”

Durham’s coverage is now high enough to have eliminated one of the other bugbears of thin-film cells – pinholes which cause local short circuits.  “The process is looking very good,” said Durose, unfortunately “details of it are a bit secret.”   Durose can now control grains in the window layer to order. “In a comparison between our CdS and standard CdS, we can increase CdTe solar cell open circuit voltage by 160mV to 700mV,” he said.

Copper or chlorine?
To effectively extract carriers from a CdTe/CdS solar cell, is it desirable to form a p-n junction between the two materials.  One way of doing this, said Professor Ken Durose of Durham University, is to weakly p-dope the CdTe.

Copper is initially the most effective dopant and, said Durose, US researchers have achieved 850mV open circuit voltage from cells using this technique.  However, copper is prone to diffuse over time, spoiling the cell.  Durham and some other researchers are instead using chlorine. “You don’t get the high efficiency of copper, which holds the record 16.6 per cent efficiency for this kind of cell,” said Durose, “Chlorine can give around 11 per cent, but the efficiency stays higher for longer than copper.”  This said, copper researchers now claim to have undisclosed methods of holding dopant atoms effectively.