Word of caution: "''We are very pleased with our semiconductor finger technology which has increased the average conversion efficiencies of our best monocrystalline PV cells to 18%"
The so called "semiconductor finger" technology it is in fact the adoption of a solar cell design that includes a "Selective Emitter (SE)" feature. SE is necessary for a very high efficiency cell, especioally the "dirty" cell designs and manufacturing procedures that involves the "dirty" but cheap screen printed contacts. For the efficiency to increase the surface emitter net carrier concentration has to be as high as possible (just below the P solubility level, or above E+21/cm3)yet low enough between the front grid fingers.
I hate to bore you with tech details. The SE concept has been tried by various labs for well over 30 years, yet no good (read cost effective) solution has been found as yet. The UNSWA SE scheme (where a n-type dopant is included into the front grid Ag-based screen printing paste prior to firing it) has been already tried by BP, and the result was only a very marginal efficiency increase, as: (i) the possible firing temperature is too low for the n++ layer to form, and (ii) lateral diffusion of this dopant widen the grid line.
Just wanted you to know: don't get too exited about the possibility of any near-term 18% c-Si solar cell efficiency using low grade Si substrates as the company would want you to believe.
Mike
