Researchers are now making transistors on paper, rather than silicon, The Economist reports:
The silicon in a transistor has two separate roles. One, when it is doped with small amounts of other elements, is as a semiconductor. This is a material that permits the limited movement either of electrons (which are negatively charged) or of positively charged “holes” in the crystal lattice where an electron ought to be. Silicon’s other role, when it is pure, is as a dielectric—a material that can be penetrated by an electric field, but not an electric current. It is silicon’s role as a dielectric that Dr Fortunato and Dr Martins propose to replace.The two researchers built their transistors by coating both sides of a sheet of paper with semiconductors made of oxides of zinc, gallium and indium, rather than silicon. They then deposited aluminium onto the coated paper to connect the resulting components together. One side of the paper carried the control currents while the other carried the output currents. The paper thus acted as the dielectric between the components of each transistor, as well as being the substrate for the circuit, in the same way that the base of a silicon chip acts both as substrate and as dielectric.
This approach lets the transistors be both flexible and cheap to produce. They can be made at room temperature, unlike a silicon chip, and paper is a lot less pricey than electronics-grade silicon. They also seem reliable. Dr Fortunato and Dr Martins tested their prototypes for two months without detecting any fall in performance.
Paper transistors, and circuits based on them, are not, it must be said, going to replace silicon chips as the microprocessors in computers any time soon—if only because they are nowhere near as miniaturised. But the two researchers have already used them to make a simple, disposable memory circuit, which they will describe in a forthcoming issue of Applied Physics Letters. Such paper-based “chips” would be much cheaper than the cheapest chips available today, and could be used in radiofrequency identification (RFID) tags on such things as packets of food on supermarket shelves—the cost of RFID chips is one of the factors preventing their widespread adoption. Baggage tags, banknotes with electronics embedded for security and even postage stamps that can be read by smart franking machines are other possible uses. Electronics may even come to rely on paper, rather than eliminating it.