An interview with the ISFET inventor Professor Piet Bergveld, by Chris Toumazou and Piet Bergveld Pantelis Georgiou 40 years of ISFET technology: From neuronal sensing to DNA sequencing

apparatus as part of my collection of classic we followed were still concerning tubes!). I medical equipment at the University [of Twente]. remember that at a certain moment I thought After World War Two the physicians there to myself: Well, then you can also make a pill, also started operations to take away part of or a with an oscillator and a that changes complete lung, and I remember my father an FM signal with pH and pressure for talking about a patient that had died instance, which can then be remotely received unexpectedly after an operation. In fact nobody using an FM receiver. So, in 1963 I developed knew beforehand who could withstand such an a telemetric lab-in-a-pill, based on a operation and who could not. And I remember transistorised Hartley oscillator in the group that they therefore decided to build next to the of Measurement Technology, to measure existing mechanical and electrical workshops, pressure and temperature in the intestines where I spent much of my free time, of patients. The ultimate goal was to also a new workshop for making biomedical measure pH, but on studying pH glass measurement instruments. These instruments electrodes I discovered that they were not yet were made to measure, in the laboratory, patient small enough to be incorporated into my pill.

blood parameters such as CO2, oxygen and pH. Nevertheless, this was my fi rst acquaintance With these they could then decide whether a with pH electrodes. certain patient could be operated on or not. I think that that experience, at the age of about The invention of the ISFET after 10–12 years, inspired me and made me decide the creation of the MOSFET that developing new measurement equipment for medical applications would be my profession in Q: What was the fi rst idea concerning the the future. ISFET? A: From 1964-65, I did my Master’s degree at An early vision: invention of Philips in Eindhoven at the Natlab in the group a lab-in-a-pill of the late Dr. Klein, who became Professor later on at the Technical University of Delft. Q: Did you study biomedical engineering at a At that time they were developing ECG, This article presents a personal account of the ABOVE: Professor Piet Bergveld university? EMG and EEG amplifi ers, still based on life and scientifi c journey of Professor Piet holding one of the very fi rst A: When the time came to go to university in the application of vacuum tubes, and they had Bergveld, the inventor and founding father of the ISFET wafers 1960, I decided I was keener on electrical started to look at using transistors, with BOTTOM RIGHT: One of the Ion-Sensitive Field Effect Transistor (ISFET). fi rst 1 inch wafers as produced engineering than medicine. I felt I had enough junction fi eld effect transistors in the fi rst stage, The interview gives a unique overview of how by the technician Arie Kooi, experience from the mechanical, electrical and because of their high input impedance. It ISFET technology has evolved over the years, containing 6 . The source instrumental workshops from the sanatorium, was my task to develop a differential amplifi er and drain connections are at and the challenges faced during the development the side of one single ISFET, to and that it would be better for the patient that I that made use of the high impedance transistors from its initial use in neuronal sensing to the be able to mount them on the be an engineer rather than a medic. That was for differential measurements, but solving in technology we see today, which has huge small end of a Perspex holder, the reason why I studied electrical engineering the meantime the problem of their bad with connections at the side. potential in the current era of genetic at the University of Eindhoven. This output resistance. I did this by technology. university was founded two years applying cascades of JFETs and before I enrolled as a new bipolar transistors and reached Introduction to the biomedical fi eld technical university in the in this way rejection ratios Netherlands, although there of the order of 10,000 for Q: How did you get into the biomedical fi eld? was no biomedical a differential amplifi er. This A: I was already involved in this area because as engineering, which I preferred. was the subject of my Master’s a child I lived from 1948–1960 with my family at But I got a good education with thesis. the TBC health-resort Sanatorium Beatrixoord, a fair engineering approach and I It was also the time that in the southern part of the province Friesland. My always did my practical work on a Philips had developed their father was the economic director, so we lived in medical subject which I thought out very fi rst , which I heard the campus of the hospital which was in the forest myself, and I looked for a group which was about because I worked in the Natlab. But as the TB patients were, before World War Two, willing to accept this for the credit points. in those early days you could hardly fi nd two treated only with clean air. At that time the only This was at the time that AM radio was MOSFETs with the same set of properties, thing that they could measure was taking Röntgen replaced by FM and transistors also came into so it was not yet useful for the new images of the lungs. I still have the old Röntgen existence (although the courses in electronics differential amplifi ers. ❯❯ doi: 10.1049/el.2011.3231 Electronics Letters Dec. 2011 – Special Supplement: Semiconductors in Personalised Medicine S7 But then, having had experience with reactions on glass for sensing pH from my lab-in-a-pill work at the university, I proposed to Dr Klein that we could potentially use the silicon oxide of these MOSFETs directly as a glass membrane, similar to the pH probe. This was when I came up with the idea for creating an ISFET. Dr Klein’s response was: “We at Philips do our utmost to encapsulate chips as well as possible against moisture and you want to put it directly in contact with the liquid? Forget about it.’’ To which I replied “Well, you should never say something like that to somebody from Friesland, because then you are sure that he will do it anyway.” [laughs] It is clear that at that time this initial idea for using the silicon oxide layer of a MOSFET as a sensing layer was not well received, and was not pursued by Philips. Many years later I discovered that somebody from Philips Natlab had written a small piece on my idea in an obscure German journal called Neues aus der Technik; done to prevent someone, including myself, from patenting it.

Waiting for a fi eld to emerge; ABOVE: A photo of an Initially I created the amplifi ers and the fi rst extended gate FET experimental set-up with an colleagues created the signal processing Parkinson’s disease, but also necessary for ISFET placed on the fl exor tibialis of a locusta instrumentation, in addition to my fi rst physical executing my ISFET plans. Q: How did you proceed? ABOVE RIGHT: Cover of sensor, which was a tremor sensor, fi rst based on A: I then moved to work at the University Bergveld’s doctorate thesis a microphone and later on so-called pixie Creation of the fi rst ISFET which shows how axons and of Twente, in Enschede. This third technical microelectronics may meet beams. for neuronal sensing university of the Netherlands was founded in each other in the future For the measurement of the electrical 1961, and started lecturing in 1963, so I was one BELOW: One of the ISFETs brain activity we used very long glass Q: How did you develop the fi rst ISFETs? mounted on a Perspex holder, of its fi rst academics. to be able to place it on a nerve needles with a 50 micron-thick platinum wire A: Some years later, in 1967, the late Prof. O.W. I was still highly interested in biomedical or muscle passing through the tip, which I then had to make Memelink from Philips was appointed at our engineering but, of course, it did not yet exist in into a shielded cable to connect to the amplifi er. department to start a chair on Solid State those days. My fi rst boss, Prof. Ir. M.P. The problem was, however, you could not Electronics. He decided to found a clean room Breedveld, however, was interested in starting it measure fairly steep action potentials due to the facility focused on the fabrication of MOSFETs. up in the future, and this was why I chose capacitance of the line. With my knowledge from I was his fi rst client; I said: “I can use your Twente. Nevertheless, in the fi rst years the my Master’s degree at Philips, I realised you MOSFETs. Bring them here, but don’t bring the courses that I had to give were pure electrical, could fi x this problem using a fi eld effect gate metal,” as I wanted to try out sensing ions mechanical and chemical engineering. But in the transistor, built in the tip of the electrode as a remotely in the brain based on my original evenings I studied physiology, biology and single chip, with the 50 micron wire directly ISFET concept. [bio]chemistry, because I was sure that in the bonded to the gate of the transistor. This would Why did I want to use this? In the Parkinson’s years to come somebody from the local then allow you to use two unshielded wires to the project I was successful with the measurement hospital would come to the university electronics. That was the subject of my fi rst of electrical signals, and my colleagues needed and ask: “Can you help us?” paper in 1968. With the source and drain to compare these signals with simultaneously Inevitably, I expected they then terminals and wires, following the gate recorded tremors to fi nd out what the would come to me, as I was signal and connected low-ohmic, I signifi cance was. I began to think to familiar with both medicine and could connect 5 metres of unshielded myself: Well, now I can measure the engineering. wires and the system could still potential, but what does this potential And indeed it happened. A measure up to 10 kHz of electrical actually mean? neural surgeon by the name of neural activity, without interferences By measuring potentials in general, one Dr van Hoytema developed a from the main supply. In fact cannot deduce what the fundamental cause is. It new stereotactic way to help this system was what later on was can result from a battery, a generator, all kinds Parkinson’s disease patients by called the extended gate FET! of electrochemical processes, etc. To really coagulating certain parts of the Thus I had created a system that know something about how an axon works I brain. He asked us whether we could measure the channel resistance of needed to measure sodium and/or potassium, could develop a measurement system an FET at a remote place. With this you did and to do so locally I needed the ISFET. for him so that he could measure specifi c not need to integrate all of the electronic After some initial experiments with the taenia activity in a certain area of the thalamus which circuitry at that spot, only the high input coli, a small smooth muscle of a guinea pig would then indicate the place where the problem impedance transistor. The rest of the electronics which shows spontaneous activity, I decided to had to be, and he could then coagulate a smaller could be operated remotely. This approach was continue the experiments with the fl exor tibialis volume than without knowing the exact place to handy for the neural probes developed for of a locusta (grasshopper), which could be ❯❯ coagulate. S8 Electronics Letters Dec. 2011 – Special Supplement: Semiconductors in Personalised Medicine doi: 10.1049/el.2011.3231 activated on demand by stimulation of a aqueous solution. That was not done at that point, otherwise they become unstable, and are ganglion in the chest. time. Fortunately, Prof. Jay Zemel from the of course very fragile, which thus limits University of Pennsylvania in Philadelphia, their application. ISFETs do not have PhD thesis: the ISFET and the OSFET USA, wrote me enthusiastic letters “I had, in my opinion, these drawbacks. concerning my fi rst papers, and this Q: Can you tell us about the fi rst results? convinced the scientists at my something great for First commercial A: Besides the intended measurement of ion department that my work was application activity around a nerve or muscle, I wanted to possibly not too bad for a PhD thesis. electrophysiologists and there create a sensor that could measure the electrical Since then Jay Zemel became and was no interest at all. That Q: How did you handle these potential without measuring any ionic activity, still is our family friend, and I call hesitations to apply ISFETs? to be able to decouple the two when measuring him my scientifi c father. was a disappointment to me” A: Having not had a good response nerve or muscle activity. Therefore I used from the electrophysiologists grease on top of the gate to stop it measuring An initial disappointment, initially, I came into contact with ions. I called it an OSFET because it was a ISFET technology was not Cordis, a company from Miami in the USA MOSFET without the M. It was not sensitive to well received at fi rst with a factory in the northern part of any chemical; it was just a potential sensor. the Netherlands, who at that time worked in When I put on the grease, I saw typical bipolar Q: Did electrophysiologists embrace the new the pacemaker industry. They contacted me action potentials. However, when I did not use possibilities for making measurements on nerves because the director had told them that the the grease, thus using the pure silicon oxide in and muscles? business of making pacemakers alone was contact with the muscle, I observed a A: Unfortunately not, and that was a diffi cult and that they had to add something monophasic response. From that I realised that disappointment. When I fi nished the initial else. They asked me if they could make what I was measuring was something unique. It papers about OSFETs and ISFETs I thought that for measuring in vivo parameters, and ultimately could not be an action potential, because action this should be interesting for electro- also build them into the pacemaker to control its potentials should by defi nition be bipolar physiologists, because at that time they could functioning. except at the beginning and end of the nerve. only measure action potentials with platinum Cordis therefore developed, together with us, That is why I speculated that what I was wires, but with these electrodes they could not an ISFET-based catheter with an integrated measuring with the ISFET must be potassium measure DC and low-frequency components. For reference electrode with a polyHema plug. We outfl ow from the axon. those measurements they needed fragile helped them especially to develop a reactor for

This was the primary work of my PhD thesis micropipettes. the deposition of Al2O3 as the gate material. called The OSFET and the ISFET, where the On the contrary, I had developed a unique All the parameters of the specifi c sensors were ISFET has the ion sensitive quality and the measurement system, based on my ISFETs, stored in the connector which contained a OSFET has the potential sensitive quality. So, which could successfully measure from DC to PROM, so that the measuring equipment could the one and the same device which is made 30kHz, which I wanted electrophysiologists to read the specifi c sensor after only one insensitive to ions measures bipolar action use. I remember writing them letters: Dear calibration [comparison with blood sample pH potentials and, in the ion sensitive case, it electrophysiologist, what do you think about this as measured by a laboratory system]. With these measures monophasic pulses, due to ionic new way of making measurements on nerves, types of sensors they could measure, for the fi rst activity. axons etc? I got hardly any reaction from them, time, that your blood pH fl uctuates with your At that time there were no PhD students at the or no reaction at all. breathing, and that holding your breath results in university. So I had to do my PhD research and My aim has always been, from my childhood a direct acidifi cation of your blood. write up my results during the evening, because as I told you, to develop new measuring systems This was important because, during an during the day I had to conduct my lecturing which could be helpful for medical doctors for operation, an anesthetist taking care of the duties teaching, among others, courses in their diagnostic tasks, and so also for patients. patient needed to critically decide when to stop electromechanics and electronics. Now I had, in my opinion, something great for the mechanical ventilation to allow the patient I had no supervisor, which is the reason why electrophysiologists and there was no interest at to breathe on their own. If the patient is not able all of my initial papers are authored just by all. That was a disappointment to me. to reach stable breathing their blood pH will myself. Later on, however, I asked Prof. In general I was confronted with: Why should drop dramatically. With the ISFET catheter pH Memelink whether he would be my promoter. you develop a new pH sensor? We are content sensor you could simply monitor the patient’s He was a little bit reluctant, because many with our glass electrodes. That’s what chemists blood pH behavior during the mechanical BELOW: The very fi rst micro- colleagues were still dubious about the fact said. But they did not realise that glass fl uidic chip for rapid in situ ventilation and then how they are doing without that I was using chips in direct contact with an electrodes can only be miniaturised to a certain titrations any mechanical assistance. If the pH dropped ❯❯

doi: 10.1049/el.2011.3231 Electronics Letters Dec. 2011 – Special Supplement: Semiconductors in Personalised Medicine S9 We made the fi rst multiple ISFET chip with integrated electrodes around the gates (see preceding page). This allowed us to measure differentially and also use the electrodes as a proton actuator by coulometry, and we were able to in situ titrate a liquid on top of the device. Therefore we mounted a lid of silicon with a channel and an inlet and outlet on top of the device. In fact this was one of the fi rst microfl uidic chips, way before microfl uidics was ever introduced to the fi eld. We showed with this device that you could do an in situ titration within one second. If you calculate back the titrated volume, based on the distance between the gold electrode around the gate of the ISFET and the gate itself, then you will fi nd a volume of a few nanolitres. Thus this experiment was in fact already microfl uidic and nanotechnology in one and the same chip, far then you could use this as an indicator to ABOVE: A differential urea making sharp ones in needle form to probe the before its time. identify that the patient could not breathe sensor meat, cheese and any other product where pH is unassisted and reconnect them again to the a critical parameter. First differential ISFET mechanical ventilation system. It’s a pity that I (and also others) compared for chemical feedback Cordis originally thought they could sell initially the ISFET too much with glass these catheters for all operations. But one of membrane electrodes. That resulted in Q: How did you further develop this the problems was that it was sometimes prone comparison concerning the quality of unique system? to blood clotting, which would prevent you measurement, focused on the long term A: We then developed what we called a dipstick from getting a correct sensor response. This stability, e.g. the drift. ISFETs should drift sensor, where we had two ISFETs both with an was not caused by the ISFET surface, but by more than the laboratory type of glass electrode around the gate and a backside the encapsulation, which cannot be smooth electrodes which is true, but one should have electrode. This could be used for a differential enough. This biocompatibility problem made it compared them with miniature glass electrodes, measurement, and we used one electrode as a too risky to sell the product. That was about at which also have more drift than the larger ones. reference electrode and one as a proton actuator, the same time that Cordis discovered that the Moreover, the drift may also be caused by the to insert protons in the solution. To be sure that pacemaker industry was not so profi table any miniature reference electrode, as well as by only protons were injected by the actuator, we more due to the rapidly growing number of the encapsulation, but this was hardly later changed the gold electrode to one of companies and advancements with the investigated. Not to be caught in this iridium oxide. technology. discussion, I decided that ISFETs should This was one of the fi rst solid-state sensors So, the money dried up in the pacemaker be promoted more regarding their specifi c that could be used for a dynamic measurement. industry, after which they decided to stop this properties, being tiny solid-state devices We used this for all kinds of experiments, for project. The development department of Cordis and having a very small response time, and as instance to determine the acidity of tomato became Sentron, and Sentron went on to make such not being in competition with glass juice for Heinz. Based on the dipstick device this type of sensor with an ISFET and a electrodes. we also developed an enzyme sensor with reference electrode for many other applications urease as an example. As an electrical engineer, than the medical. ISFETs and their fast response: I knew how you could use feedback if you invention of fi rst microfl uidic needed to linearise a normal urea ISFETs in non-aqueous solutions chips enzyme sensor. We immobilised urease on one of the Q: What were the advantages of using ISFETs in Q: Which applications did ISFETs, whereas the other applications? you have in mind? other was used as the A: The researchers from Sentron used ISFETs, A: After my fi rst PhD reference, with a layer for instance, to make sensors for the paint students, Nico de Rooij “I decided to start manu- of inactivated urease. industry. At that time, nobody knew that the and Luc Bousse, who Then we used the process of making paint had something to do worked on the facturing my own ISFET differential output with pH because it could not be measured. But fundamentals of the new signal of the system colleagues in the paint industry asked Sentron to device, I had no PhD amplifi ers in 1980” to control the current give them a sensor to try. They then discovered students working on the through the actuator that in the paint industry it is interesting to know ISFET theory only. This was electrode of the enzyme about the pH. The same was then found for meat because at that time (around ISFET in a feedback loop. and cheese. In the meat and cheese industry one 1980) I was more interested in the Urea reacted with the urease knew that measuring pH was important, but application of ISFETs and specifi cally wanted to enzyme, making OH− ions which was then measuring it with a glass membrane electrode is make use of their fast response time. At the same automatically compensated by the injection of dangerous, because if the glass breaks, then the time I wished to get rid of the use of a reference H+ ions from the actuator. The result was a whole production line needs to be stopped and electrode. Therefore we introduced differential linear relationship with the urea, independent everything thrown away. That’s what Sentron did ISFET measurements, with a normal platinum or of the background buffer, which is a unique (and still does) with these ISFET sensors: gold electrode as the common (ground) lead. way of measuring urea. ❯❯ S10 Electronics Letters Dec. 2011 – Special Supplement: Semiconductors in Personalised Medicine doi: 10.1049/el.2011.3231 The ultimate perfect ISFET theory oxide-liquid interface. Proteins work with a ABOVE LEFT: 40 years of ISFET the Fermi level. So if you can make a MOSFET through a comparison with proteins buffer capacity and a double layer capacity, the history, lying at the table of where the gate consists of a mixed conductor, Prof. Bergveld’s living room fi rst to deliver charge, and the second to store ABOVE RIGHT: The very fi rst such as iridium oxide and Perovskite oxides, Q: Have you further investigated the possibilities charge. Similarly I realised that the ability of the ISFET next to the current DNA which conducts with both electrons and ions, of ISFETs in combination with proteins? hydroxyl groups at the surface of a gate oxide to SNP chip held by Prof. Bergveld and you connect the source and drain to the (right) and Prof. Toumazou (left) A: When I started to use ISFETs for take up protons or deliver them can be seen as amplifi er, and the gate of the mixed oxide to immunosensors, the original speculation was that the buffer capacity of the surface, whereas the ground, then the system will fi nd an equilibrium. because proteins are charged molecules, you double layer capacity may store the charges. We But the output of the amplifi er will be a function could bring them to the top of an ISFET and you could easily prove that Ta2O5 has a very high of the type of gate. If it is another material with should be able to measure the presence of buffer capacity and is therefore favorable to use another work function then it will be different. charged molecules with an immune reaction on for ISFETs as the gate oxide. It shows a nearly So, if the work function of the gate changes as a top of the ISFET, a change in the charged state. Nernstian response (59 mV/dec), hardly function of the liquid, so will the output of the That was unfeasible however because the infl uenced by changes in the concentration of ISFET amplifi er. double layer of the gate oxide, approximately 10 any ions, just protons. We therefore used Ta2O5 In this way we measured oxygen and Å, is a lot smaller than the size of protein as the gate material from that point on. hydrogen peroxide without a reference molecules. In other words: the charged protein electrode. This is quite a different system to the molecules are too far away from the ISFET Misunderstood; proposing ISFETs classical ISFET, if I may call my original surface, which therefore will be controlled only without a reference electrode ISFET design that. It is not based on the by the pH of the background solution. So in measurement of a surface potential, but on a order for the ISFET to sense a protein on top of Q: I read about the criticism concerning your change in the material properties of the gate it something was needed to trigger the proteins early statement that ISFETs should also be material. In fact, such a device is still an to release protons. That was the reason that we operating without the use of a reference original MOSFET in which the gate is in direct used our, what we called, ‘ion step procedure’: a electrode. Can you comment on that? contact with a liquid. We therefore called it an stepwise change in sodium or potassium A: In one of my fi rst papers I proposed that you EMOSFET, where the E stands for electrolyte. chloride in the liquid that was going in a tube could also measure ionic activity without a Applying the device you can also send a alongside the ISFET with a membrane with reference electrode. This was not well received current through that gate into the liquid to a immobilised proteins. Due to the ion step, by the electrochemists and their community; they remote platinum wire in order to control the protons were released from the proteins, and that said: That’s very stupid from that man because a redox state of the gate material. So you can was not only by diffusion but by an active potential source should always be connected by make a feedback loop, with an oxidising agent process that could be directly measured by the two wires, in order to measure its value. I knew changing the threshold voltage of the gate ISFET as a pulse. that of course as an engineer but they had material, and the amplifi er sending a current That was when I realised that ISFETs were a misunderstood the proposed concept. We knew that reduces it again, keeping the material at lot more suited to dynamic measurements than that the ISFET’s function is based on the the same state of oxidation (constant static measurements, as I had already showed principle that at the surface there is a potential, threshold voltage). The value of the current is with the in situ titrations. This is in fact similar which is a function of pH. One side of this directly related to the concentration of the to what is found in biology, where equilibria are potential source is the gate capacitance and the oxidising agent. less important as a signal, just the changes are of other side is the liquid, which thus has to be I published the basic idea too early I think. importance. One measures therefore, for connected with a reference electrode to the other Especially chemists who read it were all against instance, action potentials instead of the side of the gate capacitance, e.g. the silicon. it saying this man is crazy! I remember I was at equilibrium cell membrane potential. What I was referring to, and which I proved a conference and I was presenting something We as engineers know that if we get a black later on, was based on the relationship of the about this. Then someone from the audience box and we don’t know something about it, you ISFET threshold voltage and the work function stood up and said: “I don’t have a question, but kick it and then the response tells you something of the gate material. If you make a MOSFET this is not true,” and then sat down, which was about what is inside. While studying the acid- with a platinum gate, a gold gate, copper gate, or very impolite, I think. This is the reason that I base behavior of proteins I realised there was an any other material, you see a difference in the only picked up the idea again much later and analogy with the acid-base behavior of an threshold voltage because of the difference in published details about the EMOSFET. ❯❯ doi: 10.1049/el.2011.3231 Electronics Letters Dec. 2011 – Special Supplement: Semiconductors in Personalised Medicine S11 ISFET array used for detection of DNA base pair incorporation based on the sequencing concept invented in 2001 by Prof. Toumazou. Prof. Bergveld provided a saliva sample which was tested using an ISFET-based disposable cartridge from DNA Electronics Ltd to detect a gene in his DNA known as N-acetyl-transferase (NAT), which performs a ‘housekeeping’ function in humans and regulates cell function. The results shown to the left are a signature of the presence of this gene in the ‘ISFET Father’s’ DNA detected in a lab-free setting in under 30 minutes by the same technology he invented 40 years ago. The same technology has already been scaled to the world’s largest array capable of sequencing a whole human genome. As we move forward, we anticipate that ISFET technology is going to make a huge impact in the biological space Public outreach for ISFETs; selling EMI Enschede, but nevertheless it is anyway ABOVE LEFT: Prof. Bergveld’s whereby dynamic changes in signal are required DNA tested from a saliva my instrumentation from His Master’s Voice!” as I was called ‘The sample in a lab-free setting to be detected on the nano-scale, leading to a Father of the ISFET’. using ISFET technology from new era for personalised medicine. Q: Did you or your university commercialise the DNA Electronics Ltd Acknowledgement: Chris Toumazou and ABOVE RIGHT: The nucleic ISFET? ISFETs and what the future holds: a acid amplifi cation test (NAAT) Pantelis Georgiou would like to thank A: Back in the 1970s it was not customary for chemical camera? performed showed a positive Professor Bergveld for sharing his memoirs of universities or staff members to apply for patents. response for the N-acetyl his life’s work, and for his help in compiling transferase gene In our case Cordis received the opportunity to Q: What are your expectations for the this feature article. apply for the patents and the university would future application of ISFETs? receive royalties. It was my intention to publish A: Because the acid-base behaviour of proteins Biography as much as possible to promote the further and the gate oxide of an ISFET are similar in Piet Bergveld was born in Oosterwolde, the development and application of ISFETs. their existence, an ISFET is in fact the most Netherlands, on January 26, 1940. He received However, it sometimes takes much more time perfect contact for proteins, for information amaster’s degree in Electrical Engineering than you expect when you start a certain retrieval. But then the measurement procedure from the University of Eindhoven, the Nether- development to get people to use your should be dynamic. lands, in 1975 and a PhD from the University technology. You can try to infl uence it, but when In one of my last offi cial keynote lectures at of Twente, the Netherlands, in 1973. The people are thinking in a different way it gets very the MicroTAS symposium 2002 in Nara, Japan, subject of his dissertation was the develop- challenging. Well, my story is for instance: when I introduced a 2D ISFET array, based on the ment of ISFETs and related devices (the actual I started with the ISFET business, people asked Philips SeeMOS technology. I proposed that, invention of the ISFET);which since then has me if they could get some ISFETs. I said: yes, with only a few adaptations and making use of been also investigated by many international we can send them some ISFET chips, but then the existing CCD readout system, one could research groups of universities and industry. thought what could they do with them? If they easily realise what I called a ‘chemical camera’. Since 1965 he has been a member of the had no problem bonding the chips to a carrier It meant that redistribution of charges as a result Biomedical Engineering Division of the Fac- and packaging them then that was fi ne, but what of, for instance, an immune reaction or any other ulty of Electrical Engineering (University of about the necessary electronic systems? specifi c protein binding phenomena at the surface Twente) and was, in 1984, appointed as Full I wanted to convince people that they could of the array should be monitored as the mirrored Professor in Technology. He is one use ISFETs in many application fi elds, however, dynamic charge rearrangement in the silicon. of the project leaders in the MESA Research so I decided to start manufacturing my own Because I had already planned my Institute. His research subjects still concern ISFET amplifi ers in 1980. At home I made retirement, I hoped that somebody should pick the further development of ISFETs and bio- boxes with plugins for a number of ISFET up this challenging idea. It’s up to you whether sensors based on silicon technology as well as amplifi ers to allow people to use ISFETs. The or not you think that you could do this at your physical sensors, both for biomedical and ISFETs were sold worldwide by Nico de Rooij, department at Imperial College. If yes, I will be environmental applications, resulting in more who had become professor at the University of very happy, because although my original idea than 450 papers and 35 PhD theses. He was Neuchâtel in Switzerland. I called my company in the 1970s was to measure the dynamic ionic Research Dean for the Faculty of Electrical ‘Electromedical Instrumentation’ because at that activity around nerve and muscle fi bres in order Engineering from 1994 to 1998 and received, time I had written a book for nurses called to understand its meaning and help doctors to in 1995, the Jacob Kistemaker Award from the Electromedical Instrumentation, aiming to teach understand and recover deviant behavior, the FOM Institute. He was founder and chairman nurses some basic electronics for biomedical system is of course also very important for any of the international steering committee of the instruments focused on patient safety. I was other similar dynamic biological system, of annual microTAS Symposia from 1994 to selling the amplifi ers mainly to university labs which DNA sequencing is a clear example. 2000. In 1997 he was appointed as a member and some industries. of the Royal Netherlands Academy of Arts The result was that several research groups Concluding remarks and Sciences. He received the Microsystems which used my instruments mentioned in their In conclusion of the interview Prof. Chris Leadership Award during microTAS 2002 in papers that the amplifi er they had used was Toumazou presented Prof. Bergveld with the Nara, Japan, retired from the University of from EMI, which everyone confused with the latest technology developed by his group, which Twente in February 2003 and received the well-known British company Thorn EMI. I said was a realisation of the visions he had already knighthood of the Dutch Lion in April 2003. for fun “Well, it is not from Thorn EMI but from mentioned. A CMOS-based microchip with an S12 Electronics Letters Dec. 2011 – Special Supplement: Semiconductors in Personalised Medicine doi: 10.1049/el.2011.3231