Renato Turchetta of RAL shines a light on imaging in the non-visible spectrum
30 January 2012
Renato Turchetta, CMOS Sensor Design Group Leader at the Rutherford Appleton Laboratory shines a light on non-visible imaging applications..
Q: Please briefly describe your background in digital imaging.
A: In the early '90s I was developing detectors for particle physics experiment at CERN, while working at the University of Strasbourg in France. In the field, there was a general interest in 'spinning off' the sophisticated detectors developed for particle physics into other domain and mainly X-ray digital imaging. My first step into this field was then to develop the readout IC for a photon counting detector for digital mammography at the Italian synchrotron Elettra. The detector worked well and the principle of the detector (edge-on silicon microstrips) was exploited commercially although by a group external to the collaboration.
Products based on this technology are still sold today. After that project I started looking at different technologies for imaging and came across the first papers about CMOS image sensors, which had just been published by the JPL group. I immediately found the concept had great potential and starting thinking of ways of applying similar concepts to scientific applications. The first idea was about a CMOS structure that could efficiently detect charged particles. This idea was then patented and it is now at the heart of CMOS image sensor working in direct detection for transmission electron microscopy.
After this first application of CMOS image sensors, I continued looking at different applications for this technology. By then I had moved to the Rutherford Appleton Laboratory, one of the largest UK national laboratories and a place rich in applications and needs for advanced detectors. It is a good place where to develop new detector technologies. From here it is also easier to link to industry, working with them to find a way of applying our technology to their needs.
Although detection of charged particles and X-ray digital imaging remain at the heart of my work, nowadays my group and I also work in developing the CMOS image sensing technology in other fields like mass spectroscopy and UV detection.
Q: Your presentation covers image sensors for non-visible wavelengths, what are the key differences in the technology used?
A: The first difference is to be found in the penetration of non-visible photons. Moving towards longer wavelengths in the infrared domain, the penetration rapidly increases demanding thicker substrates. In the other direction, towards UV and X-rays, at first the penetration becomes very shallow and then gradually increases. Shallow penetrating radiation demands high-quality backthinning for backside illumination, while infrared and, in some respect X-rays as well, demands substrates thicker than those used in the visible range.
The second difference is that when moving away from the visible spectrum, optics becomes gradually unaffordable or impractical. For X-rays this means sensors have to be as large as the field of view, and this can be very large, e.g. in medical imaging!
The third difference is when the wavelength of the photons, decrease, their energy increase correspondingly, thus making single photon detection easier, allowing both photon counting and spectroscopy.
Q: What are the main applications, and how to they compare in market size compared to visible spectrum imagers?
A: Needless to say that applications tend to be more exotic and less consumer-oriented! They are however very diverse, and this is also true for market size. Regarding this latter, at one end of the spectrum we find scientific applications where there size of the market can be limited to only a few, high-performance devices.
A variety of applications that I would tend to label as 'industrial analysis' require a relatively small numbers of units per area, we are talking tens or hundreds. Regarding the applications where I work, the highest volumes are to be found in medical imaging where several thousands of sensors are needed every year.
Q: Can you outline the emerging technologies in this field?
A: I have been working for a long time in medical imaging. In large area applications, like mammography or chest radiography, TFT technology has been the major player for many years now. CMOS technology is starting to emerge and it could become the major players in the next few years.
Q: What are the main research challenges right now?
A: Some of the challenges stem from the differences between detecting non-visible photons, while working with a technology (CMOS) where investments are mainly to detect visible light! For example, while backside illumination (BSI) technologies are being developed for small pixels, good BSI for scientific applications still remains a dedicated process, or better to say post-process. It would greatly help the work of designers if there was an integrated approach with CMOS foundry and backthinning.
In many applications, sensors need to be large and designing a large sensor, up to the size of a single wafer, has its specific challenges and requirements, different from implanting the same architecture on a smaller scale. Achieving high-speed is also a frequent requirement. Last year I already presented an architecture which can achieve equivalent speed of tens of millions of frames per second, and in my group we are also working on other architectures suitable for different applications.
Q: What interesting projects are you involved with at the moment?
A: Some of the most exciting projects I am involved at the moment are on X-ray imaging. In one of them we are working on a high-resolution sensor that coupled to an optimised scintillator can achieve unprecedented spatial resolution. In another project, we are working on developing a wafer-scale sensor for medical imaging. The goal is to achieve lower noise and faster frame speed than other competing technologies.
Q: Finally, we are pleased to have you on board for the conference this year, what are you hoping to gain from the conference?
A: I am always curious to learn what other people are up to. Imaging technology evolves very rapidly and I find it is essential to keep up to speed with the latest innovations. Innovations can come from many directions and it is important to keep the eyes open and try to see where new technology developed for one application can be applied into others, even if apparently very distant. For example, 4T technology was greatly boosted by their application in consumer products, but it is also an excellent way of achieving the very low noise required in many scientific applications.
Renato Turchetta will be providing a presentation on Day 2 of Image Sensors 2012 entitled "CMOS image sensors for non-visible applications: seeing the invisible".
>> View the full programme here
>> View a full list of speakers here