Edited by Leland Teschler
|
Epson researchers added a layer of polysilicon by shooting the material out of an inkjet nozzle for this prototype thick-film transistor (TFT). They created a gate, source, and drain using a conventional lithography process. |
|
Inkjetted circuit elements are potentially simpler to make than those created with conventional lithography, as illustrated by a comparison of the steps involved with the two processes. |
So say researchers at Seiko Epson Corp. in Japan. They have devised a practical way of synthesizing liquid silicon that shows promise as a material for building transistors.
The new material comes out of research conducted for Japan's New Energy and Industrial Technology Development Organization. Key to this development was the synthesis of silicon in a form that would dissolve without using a solvent containing impurities. Seiko Epson worked with chemical supplier JSR Corp. to develop such a material based on cyclopentasilane (CPS), a compound of silicon and hydrogen. Researchers radiate the material with UV light to polymerize it. In this form it is less prone to evaporation, a problem with untreated CPS. Researchers then inkjet or spin-coat the material onto a substrate as a film. Then it is baked at over 500°, turning it into amorphous silicon, and finally annealed with a laser for better electrical conduction.
The attraction of the new process is that it holds the promise of eliminating the high-vacuum lithography now used to create ICs. The inkjetting or spin coating takes place in a nitrogenous atmosphere. There are fewer process steps and the manufacturing equipment is potentially less expensive and relatively small. The fab facility, in turn, could be smaller than conventional plants as well.
But there are several obstacles to be dealt with before the inkjetting process is ready for production. One difficulty is in picking a substrate. Researchers fabricated demonstration transistors on glass which uses a layer of the CPS film which is 300-nm deep, considered too thick for practical production. Researchers would also like to create inkjet nozzles capable of ejecting more minute droplets of the material. The ultimate goal is to shoot droplets on a nano-level for creating 0.5-µm features.
