Build a better droplet producing machine and the world's pharmaceutical companies will beat a path to your door.

Young campus company, Allegro Technologies has just opened up in a Trinity College incubation unit and is commercialising novel equipment that can measure out tiny amounts of liquids. A typical tap water drop might contain about five microlitres of water but the new device can deliver a droplet of just one microlitre, about a cubic millimetre in size, and about as small as a drug company testing new pharmaceuticals would want, explained Dr Igor Shvets, a lecturer in the Department of Physics at Trinity College.

"Our technology allows us to get as small as one nanolitre, three orders of magnitude below the current limit," Dr Shvets explained. This is about as much liquid ink required to make a single fullstop on this page. "We have developed a system in the college and have applied for a European patent."

This is potentially a big deal for the pharmaceutical companies, he explained. These companies constantly develop new drugs not knowing whether they are good for anything nor whether they are toxic. They retain huge libraries that typically contain millions of untested compounds. They are produced at a very high cost and for this reason only very small amounts of a compound are synthesised, so they are considered a very precious resource.

They also take up a large amount of space. Compounds are often kept in cold rooms in well plates. These typically contain 384 individual wells in a plate measuring two by four or three by five inches. A computer-based referencing and retrieval system is also needed so that individual compounds can be recovered from the library.

Development of any one substance might cost $500 million (€516 million) and take five to 10 years to bring to market, Dr Shvets said. "A lot of this goes into the pre-clinical activities," including seeing what the drugs do in the cell, determining whether they are toxic and then seeing if there is a potential clinical application.

"The most common way to do this work is to identify a protein or enzyme in the body" and then see if a compound reacts with it. "Basically all drugs are based on blocking some reaction within the cell," he explained. "There is no way to predict in advance how things will react."

The process of first isolating a cell chemical such as a protein or enzyme and then searching for reactions with the pharmaceutical library is known as "high throughput screening".

There are cost implications however. "A pharmaceutical company may need to go through a large number of targets before identifying something that will work," Dr Shvets said. The high cost of producing library compounds is such that a single reaction might cost from $1 to $5 and perhaps two million reactions would have to be done during screening.

This is why chemical droplet size is important to these companies. The library compounds are "an invaluable asset for the company" he said, so they strive to use as small an amount as possible during testing.

This is where small droplet technology becomes important. Most firms currently use a relatively old technology based on a syringe pump to produce five microlitre droplets. "Every year they have reduced the volume and now they are looking for a new technology which would allow them to decrease below five microlitres."

Dr Shvets sees the main market for droplet technology at the 100 nanolitre to one microlitre range, well above what his droplet machine can achieve. It is also important that the device be able to handle a wide range of fluid types if it is to serve as a clinical device handling anything from blood to sugar water and oil-based liquids. The current generation of devices uses very similar technology to that found in an ink jet printer but these do not readily handle a variety of fluid types, he said. A successful system ideally would also have a disposable delivery tip assuming that mixing of the test solutions is not wanted. He said that ink-jet-based systems are too expensive to be disposable. "The technology that we have developed allows us to go well below a dollar per tip."

The project crosses the divide between physics and medicine and the company's staff reflect this. Its research partners include Dr Shvets and Dr Jurgen Osing a physicist, and Professor Dermot Kellegher, Trinity's professor of clinical medicine.

The company is currently raising venture capital in the region of £2 million to £3 million "to sustain the company over the next two years. The company will be quite heavily focused on research and new technologies are expected to emerge".

One key area for them will be "microarrays", the technology that promises to put an entire laboratory on a single microchip. It is closely linked to droplet production because it involves applying tiny drops of DNA and other proteins to a surface and then exposing the surface to body fluids or intercellular chemicals. The individual droplets react with these substances to deliver information such as the diagnosis of an illness.

A modified droplet device would support microarray technology, Dr Shvets said, by automating droplet placement but also by delivering much smaller droplets of a consistent size.

Dr Shvets believes there is a "great future" for this work given the success of US firm, Genetics Microsystems Inc which launched in 1997 with a $10 million capitalisation and which was bought out two years later for $100 million. Market growth in this area is expected to reach dot.com company proportions, he said, but unlike a dot.com Allegro Technologies would actually retain an asset, its intellectual property.