||This article may contain original research. (October 2007)|
A chemical computer, also called reaction-diffusion computer, BZ computer (stands for Belousov–Zhabotinsky computer) or gooware computer is an unconventional computer based on a semi-solid chemical "soup" where data are represented by varying concentrations of chemicals. The computations are performed by naturally occurring chemical reactions. So far it is still in a very early experimental stage, but may have great potential for the computer industry.
The simplicity of this technology is one of the main reasons why it in the future could turn into a serious competitor to machines based on conventional hardware. A modern microprocessor is an incredibly complicated device that can be destroyed during production by no more than a single airborne microscopic particle.
In a conventional microprocessor the bits behave much like cars in city traffic; they can only use certain roads, they have to slow down and wait for each other in crossing traffic, and only one driving field at once can be used. In a BZ solution the waves are moving in all thinkable directions in all dimensions, across, away and against each other. These properties might make a chemical computer able to handle billions of times more data than a traditional computer. An analogy would be the brain; even if a microprocessor can transfer information much faster than a neuron, the brain is still much more effective for some tasks because it can work with a much higher amount of data at the same time.
Historical background 
Originally chemical reactions were seen as a simple move towards a stable equilibrium which was not very promising for computation. This was changed by a discovery made by Boris Belousov, a Soviet scientist, in the 1950s. He created a chemical reaction between different salts and acids that swing back and forth between being yellow and clear because the concentration of the different components changes up and down in a cyclic way. At the time this was considered impossible because it seemed to go against the second law of thermodynamics, which says that in a closed system the entropy will only increase over time, causing the components in the mixture to distribute themselves till equilibrium is gained and making any changes in the concentration impossible. But modern theoretical analyses shows sufficiently complicated reactions can indeed comprise wave phenomena without breaking the laws of nature. (A convincing directly visible demonstration was achieved by Anatol Zhabotinsky with the Belousov-Zhabotinsky reaction showing spiraling colored waves.)
Basic principles 
The wave properties of the BZ reaction means it can move information in the same way as all other waves. This still leaves the need for computation, performed by conventional microchips using the binary code transmitting and changing ones and zeros through a complicated system of logic gates. To perform any conceivable computation it is sufficient to have NAND gates. (A NAND gate has two bits input. Its output is 0 if both bits are 1, otherwise it's 1). In the chemical computer version logic gates are implemented by concentration waves blocking or amplifying each other in different ways.
Current research 
In 1989 it was demonstrated how light-sensitive chemical reactions could perform image processing. This led to an upsurge in the field of chemical computing. Andrew Adamatzky at the University of the West of England has demonstrated simple logic gates using reaction-diffusion processes. Furthermore he has theoretically shown how a hypothetical "2+ medium" modelled as a cellular automaton can perform computation.
The breakthrough came when he read a theoretical article of two scientists who illustrated how to make logic gates to a computer by using the balls on a billiard table as an example. Like in the case with the AND-gate, two balls represents two different bits. If a single ball shoots towards a common colliding point, the bit is 1. If not, it is 0. A collision will only occur if both balls are sent toward the point, which then is registered in the same way as when two electronic 1's gives a new and single 1. In this way the balls work together like an AND-gate. Adamatzkys' great achievement was to transfer this principle to the BZ-chemicale and replace the billiard balls with waves. If it occurs two waves in the solution, they will meet and create as a third wave which is registered as a 1. He has tested the theory in practice and has already documented that it works. For the moment he is cooperating with some other scientists in producing some thousand chemical versions of logic gates that is going to become a form of chemical pocket calculator. One of the problems with the present version of this technology is the speed of the waves; they only spread at a rate of a few millimeters per minute. According to Adamatzky, this problem can be eliminated by placing the gates very close to each other, to make sure the signals are transferred quickly. Another possibility could be new chemical reactions where waves propagate much faster. If these teething problems are overcome, a chemical computer will offer clear advantages over an electronic computer.
An increasing number of individuals in the computer industry are starting to realise the potential of this technology. IBM is at the moment testing out new ideas in the field of microprocessing with many similarities to the basic principles of a chemical computer.
- "Introducing the glooper computer" - New Scientist article by Duncan Graham-Rowe (Restricted access)
- L. Kuhnert, K. I. Agladze, V. I. Krinsky (1989). "Image processing using light-sensitive chemical waves". Nature 337 (6204): 244–247. doi:10.1038/337244a0.
- Adamatzky, Andrew and De Lacy Costello, Benjamin (2002). "Experimental logical gates in a reaction-diffusion medium: The XOR gate and beyond". Physical Review E 66 (4): 046112. doi:10.1103/PhysRevE.66.046112.
- Andrew I. Adamatzky (1997). "Information-processing capabilities of chemical reaction-diffusion systems. 1. Belousov-Zhabotinsky media in hydrogel matrices and on solid supports". Advanced Materials for Optics and Electronics 7 (5): 263–272. doi:10.1002/(SICI)1099-0712(199709)7:5<263::AID-AMO317>3.0.CO;2-Y.