When the Berlin Wall came down the world rightly rejoiced. But for those who lived in the former East Germany (DDR), the ensuing years proved a struggle as they tried to earn a living in the capitalist economy of the new Germany.
PicoQuant is one of a small but significant group of companies that emerged from the aftermath of German re-unification and proved that science and innovation are universal. Its founders all grew up in the DDR and built an organisation that takes the best parts of the old system and connects it to the technological frontiers of global industry. They took a good idea, added a bit of cleverness and hard work and built the leading supplier of picosecond diode lasers. There are plenty of other companies chasing them, though, and when this technology emerges from the laboratories into the mainstream OEM business there will be plenty of companies looking to take over that lead.
PicoQuant’s story starts at the DDR Academy of Sciences which, like many DDR organisations, was having to slim down after reunification. The German government was encouraging the scientists there to start hi-tech businesses and offering generous seed capital funding. A company, Biosquant, was formed to develop a couple of research projects, one involving photonics and the other division involved in water purification. Biosquant burned through its start-up funding without much sales success and eventually closed.
The photonics division had more technical success by discovering how to make pulsed diode lasers. When Biosquant folded, Rainer Erdmann, who had managed the photonics group, recruited four colleagues and founded PicoQuant in 1996 to exploit the pulsed diode techniques.
Getting diode lasers to emit very short pulses is not easy. A small number of companies have managed to do it, but the techniques are not covered by any patents. Rather it is achieved using electronic alchemy, and each company has its own proprietary ‘secret sauce’ that makes its devices work.
Uwe Ortmann, head of sales and marketing at PicoQuant, explains: ‘The key people in the company were electronic engineers and they came up with a very cunning way of making pulses and putting them on diode lasers to make a normally continuous laser pulse in short bursts at a very high repetition rate. You cannot just take any diode. If you take a batch of ten diodes and you can make some of them pulse it does not mean the others will pulse. Often we can only get a fraction of a batch of diodes to pulse. Each diode is capable of pulsing, but it is a question of whether it is good enough to pulse. We buy in the diode chip and put the electronics on it. The key is matching the electronics to the diode. If you overpower them you can destroy the diode. The know-how we have is in managing this. When the company started, patenting was very expensive, so we have not patented our knowhow, we have just kept it secret.’
While Biosquant had been amply funded in a bid to pump-prime former DDR business ventures, PicoQuant received a very small amount of outside support and had to survive on a shoestring budget until it made its first sales. At the time PicoQuant started, the standard short pulse high repetition lasers were the Ti:sapphire or dye lasers, but they were extremely expensive and required a lot of maintenance. A solid-state solution sounds like a revolution, but of course the catch is that the power output is measured in mW and the peak power is about 1W, hardly enough for most industrial applications.
On the other hand, there are a whole range of applications for which that is more than enough, such as remote sensing, information transfer and biochemistry. The first users were physical chemists working in time-resolved fluorescence, something which later really took off when it was applied to biology. About 80 per cent of the early customers were university research laboratories, who were quite happy to just have a bare laser and would build their own machine around it.
Sales came very easily because the actual unit cost of the laser was so much less than the rival technologies, even though they were much better known. It could be purchased on a laboratory’s routine budget rather than being regarded as a major capital purchase that needed to go through a full institutional budgeting process or require a grant application to an outside body.
Left: A basic fluorescence lifetime imaging set-up, using a time-correlated single photon counting unit. Right: An image of the autofluorescence of cells of Lillies of the Valley.
For the first few years the company simply sold the light source, but its expertise was in electronic design, so its first development of the product line was to include the timing electronics. This was released as the TimeHarp 100, then the TimeHarp 200 and as further development step, the PicoHarp 300. The ‘Harps’ are now a major package product line. The electronics became multichannel with the HydraHarp brand.
The Sepia multichannel laser was first developed for an optical tomography OEM customer, but is now offered with a variety of colour laser heads both to scientists and OEMs. PicoQuant tries to stay close to the research community to discover what the scientists are wanting from its device. This has worked both ways, and when scientists heard about the Sepia device they were very quick to find new applications for it.
Ortmann says that there could be many applications that will interest OEMs in the future, particularly in the biomedical field, but at the moment there is still a tremendous amount of fundamental research that has to be done to discover the physical properties of biological systems. For example, optical tomography machines could replace x-ray mammography machines in the future. But what is holding things up is research into the optical properties of tissue under various conditions, rather than the availability of the light source.
The next stage for the company was clearly to offer a complete system in fluorescence spectroscopy. Ortmann says: ‘When you have an excitation source and you have the detection electronics, it’s not very hard to close the cycle. Of course, you need good optics and software, but then you are very close to a spectroscopy system. If you can use your own detector and source out of the same hand then this is a good position for a manufacturer. Nowadays we balance the company on three legs.
‘Now everyone can buy a single molecule sensitive device and they do not have to be a physicist or a laser expert. You can just switch on the device and measure single molecules. This was a big improvement as far as researchers are concerned.
‘Of course we have OEMs in this area, but the biophysics is still not so advanced that you are going to sell thousands of these things. For example, if every hospital was going to buy an optical tomography machine, then OEMs would be interested. We are working in a niche on a much smaller scale. It is still an interesting international market for companies like us serving the research community.
‘If optical tomography did take off then we would prefer to be an OEM supplier, because there are clinical and regulatory requirements that we would not be able to service. Companies like GE can do all this stuff much better than we can; we are not big enough or strong enough.’
He says that once this technology is mainstream, he still expects PicoQuant to have a role as a supplier. The light source is a small component of the OEM machine and the volumes will still be comparatively small. He thinks the big OEMs will find it hard to justify repeating the research effort that went into making the diode lasers pulse and will prefer to source components from someone who has already discovered the secret.
Since the company started in 1996, there have been some significant developments in electronics and semiconductor laser technology. This has allowed PicoQuant to gradually extend its range of lasers into new wavelengths and consequently open up new applications.
Ortmann says: ‘When we started we could only make red diode lasers down to 640nm and that is not really the ideal wavelength for fluorescence. You really want to excite below the emission side; that means you normally excite at 400nm and there were no lasers of that wavelength. Since then we have had gallium nitride diodes, where we can go to wavelengths that are really useful to chemists. For example, there are UV LEDs that are now going down to 265nm and, of course, chemists want to go down to the intrinsic florescence of proteins, for example. We were very quickly able to pulse these LEDs and so we are now able to get down to the deep UV. We are also developing faster modules with greater capability.
‘For OEMs it is not important that you can do everything, you just have to know where to source things and that the company you are buying from is, like PicoQuant, big enough to sustain your requirements for many years into the future.’
PicoQuant has grown to about 44 people, and includes some people from the original DDR Academy of Science, but headcount has started to grow much more quickly since it entered the systems business. It has also required a change in the way the company operates. When it just sold lasers to scientists, there were no customer training or service needs; now its engineers have to go out and meet the customers of its systems products and solve their problems.
Ortmann says: ‘We never needed to do installation or service for lasers and timing systems, but once we went into the systems side we had to employ people to go out and train users and perform maintenance.’
He has also built up a sales channel around the world with resellers in several countries, and a subsidiary office in the USA. These representatives are fully trained, but when it comes to the technical aspects of a sale or detailed support, PicoQuant prefers to deal with these directly from Germany. Ortmann says: ‘The people we have here in the sales force are very good at the science; they are not selling catalogue products. If someone asks for something from us and they would be better using a rival technique, we tell them. We like to have the reputation that you can get help here; it pays off in the long term. We have a sales network with three people in the US, and for most areas we like to do as much as possible direct, but you cannot have everything direct, so we train resellers. We need to have people who understand the local sales techniques and paperwork and actually administer the sales.’
The company has managed to keep profitable in its first 12 years, but that does not mean a few people at the top of the company have become rich while the rest eat cabbage. Maybe in homage to its roots in developed socialism, PicoQuant runs itself more like a co-operative than a capitalist company.
Ortmann says: ‘We invest the money of the company partly in the people and partly in the buildings and equipment. If there is any money left over at the end of the year, then everyone gets bonuses, and that means everyone. We do not have a system where someone gets 50 per cent more because they have a PhD rather than a college degree, because the person with the college degree could be doing a more important or harder job. The five founders of the company, who own it, are from East Germany and have some feelings left over from the ‘social times’ they grew up under.’
PicoQuant has found its base in Berlin to be very comfortable as it expands. It has high-tech offices in the WISTA Campus in Adlershof, surrounded by technology and new media companies, as well as research centres. Berlin also has several universities from which to recruit and living costs are low compared to other technology centres in Germany. It has achieved a leading position in what is a very small market, but is determined to stay at the leading edge when picosecond diode laser technology comes out of the laboratory into the mainstream.