Sinopsis
The discovery of diamond synthesis by Chemical Vapour deposition (CVD) in the early 1980s opened a large range of new applications for diamond. Among these, and the one that motivated much of the early and current research, is the use of CVD diamond for electronic devices and sensors. It is the aim of this book to review the state of the art on this subject. Within this scope I have aimed to include the widest possible range of devices even at the risk of stretching the conventional meaning of an electronic device. Therefore in addition to the ‘classics’ such as rectifying diodes and transistors, I have included radiation sensors (for a wide range of radiation ranging from photons to high energy particles, heavy ions and neutrons), electrochemical and biological sensors, micro-electro-mechanical systems (MEMS), and a chapter on the recently discovered superconductivity effect in diamond.
When I was approached by Professor Willoughby and his colleagues from Wiley in the spring of 2006, with the proposition to edit a book on diamond electronics, I had serious reservations as there are already several very good books and review articles covering this subject. In addition it was then over 16 years since I had started my involvement with CVD diamond and, realistically, one could not claim that diamond electronic devices had been a rolling commercial success. Despite early claims that diamond would replace silicon as the key electronic material, there are only a few niche applications currently on the market (mostly radiation and electrochemical sensors), a far cry from the industry that would be worth over 2000 million dollars by 2010 as projected in the late 1980s and early 1990s. On further reflection, I concluded that the important advances achieved since the turn of the Twenty-first century had taken CVD diamond technology into a second phase of development and that this progress needed to be properly reviewed. A book like this would therefore not only be a positive addition to the current body of work, but also might help to clarify the future market for diamond electronic devices.
What were the events that led to the current phase of development?
Many of the superlative physical and chemical properties of diamond have been known since the mid 1940s, almost four decades before the momentous experiments by the NIRIM (National Institute for Research in Inorganic Materials) team in Japan, who demonstrated the feasibility of CVD diamond synthesis. It was known that diamond, a wide band-gap semiconductor, was the hardest and most rigid (highest Young’s modulus) material, the most thermally conductive (at room temperature); that it was inert to chemical attack; that (when free of nitrogen) it was transparent in the UV, visible and infrared part of the spectrum (except for the two and three phonon bands between 2 and 7 micrometres wavelength); that it could be used as a radiation detector and that the electron and hole mobilities were very respectable. Later on it was also learned that diamond could be doped with boron to become a p-type semiconductor and rudimentary rectifiers and transistors were demonstrated. All this was achieved using, first, natural diamond stones and later, diamond synthesised by the high-pressure and high-temperature (HPHT) process.
Heroic attempts were made during the pre-CVD era to use diamond in high technology applications outside the abrasive and cutting tool industry that was, and still is, superbly served by natural and HPHT diamond. A good example was the UV diamond window (18.2 mm in diameter and 2.8 mm thick) made out of a natural Type-IIa stone used for the main pressure cell of the Venus space probe launched in August 1978. The price of this window at that time was close to $250 000, equivalent to over $1 million today. Also, small pieces of IIa diamond were sold as heat sinks (or heat spreaders) to mount diode lasers but the price was too high and the size too small to consider these as heat sinks for high-power electronic circuits. Whatever scientists thought of the electronic devices that were made using natural or HPHT diamond, it was clear that there was no hope whatsoever of developing significant commercial electronic devices if the only materials available were small natural stones of irreproducible properties, or similar-sized HPHT diamond mostly contaminated with nitrogen or metal catalysts.
In view of the above, it should come as no surprise that a great deal of excitement was generated when scientists from NIRIM in Japan reported that it was possible to synthesise diamond from the gas phase, at sub-atmospheric pressures in a continuous way on a nondiamond substrate. These achievements were reported between 1982 and 1984, using firstly hot filament and later microwave plasma assisted reactors. Extrapolating (probably too enthusiastically) from what was known of gas phase epitaxial techniques used for the growth of high electronic quality Si and III–V materials, some sectors of the scientific community concluded that we were witnessing the dawn of an era where wafer size electronic quality diamond could be grown. Further extrapolation led to the rather extravagant claims about diamond replacing Si and the very large size of the markets mentioned above. All this resulted in a substantial amount of funding from government agencies throughout the 1990s and a brave attempt by at least one Japanese company to open research sites in the UK and USA for the specific task of developing commercial diamond electronic devices. Alas, well before the century was over, most of the funding (at least in USA) had dried up (being diverted mostly to the wide gap III–V nitrides and SiC), and the industrial research sites in Europe and USA mentioned above were closed down. There was nonetheless a positive outcome from all these activities: very good research was done and the considerable know-how that was acquired has been invaluable for all the further developments in diamond electronics that will be described in this book. More realistic expectations were harboured by different groups of scientists who were interested in the optical, mechanical and thermal properties of diamond, and who believed that CVD technology could open the way to large, affordable windows and heat spreaders, and that CVD diamond could also contribute to the cutting-tool industry with either thick diamond layers or diamond coatings on tungsten carbide tool inserts. These expectations met with considerably more success. High quality diamond windows up to 100 mm in diameter and 1–2 mm thick have been commercially available now for over a decade, and have been used routinely for the transmission of high-power CO2 lasers, broad band transmission from Synchrotron infrared beam lines, and the transmission of MW power from microwave tubes for use in thermonuclear fusion reactors, just to name a few. Whilst not cheap, these windows are affordable. The Venus probe window, for instance, if made by CVD diamond, would cost no more than $10 000. Furthermore, these windows met all the technical specifications that were expected for pure Type-IIa diamond. The notion of a piece of high quality diamond over 100 mm in diameter, 2 mm thick (∼280 carats) that could be routinely available for technological applications at an affordable price would have been considered a wild dream before the advent of CVD diamond technology. In addition, diamond heat sinks of about 1 cm2 are routinely used in the packaging of high-power rf transistors, and the CVD diamond tool industry has found interesting applications in providing a very useful complement to HPHT diamond. All this progress was achieved in just over a decade, which is good going for a new technology. I believe we are currently witnessing a renewed impetus in what I call the second development phase in diamond electronics, and this is due to two major factors.
One derives from diamond being a very adaptable material. New attributes and properties have been discovered over the past couple of decades, such as the large voltametric potential and the possibility of functionalising the surface. These have opened exciting possibilities for biological and electrochemical sensors. Although doping has been, and continues to be, the most serious hurdle in the realisation of diamond electronic devices, the discovery of a surface p-type conductive layer in hydrogenated surfaces, and the possibility of implementing delta doping, have resulted in diamond FETs with very attractive performances. The fact that diamond is a superb ceramic material with a very high Young’s modulus that can be made electrically conductive and can be grown as nanocrystalline layers with smooth surfaces, has motivated considerable interest in the development of Micro-Electro-Mechanical Systems (MEMS).
The other factor is the considerable progress that has been achieved in crystal growth technology resulting, for instance, in the synthesis of very high purity and crystalline quality homoepitaxial single crystal specimens that exhibit considerably greater values of electron and hole mobilities, as well as dielectric breakdown, compared with values previously reported. Size, nonetheless, still remains a problem since even in the most
optimistic case crystal sizes are not much above 8–10-mm squares. There has however been much progress in the understanding of heteroepitaxial synthesis, which offers a glimmer of hope for the synthesis of large size single crystal diamond wafers.
The first part of this book (Chapters 1 to 6) reviews the basic properties of diamond with emphasis on what is new and relevant for many of the new developments. Gordon Davies (Chapter 1) reminds us about the basic physics: lattice structure, phonon spectra, elastic, thermal and other properties, including some of the recent understanding on isotopic effects. In Chapter 2, Jan Isberg reviews the current knowledge of free carrier transport, describing for instance the experimental evidence for the very high reported values of electron and hole mobilities. The topic of defects and impurities in diamond is reviewed by Alison Mainwood in Chapter 3, which gives a comprehensive account of the impurity centres in diamond either as single elements or as complexes, and discusses their possible role in electronic properties including some recent work on shallow dopants. The topic of surface conductivity exhibited by hydrogenated diamond surfaces is treated by Lothar Leyin Chapter 4, a fascinating account of how the understanding of this phenomenon, unique to diamond, has developed. The chemistry of crystal growth, including a description of the new experimental techniques that are being used for the analysis of the growth chemistry, is given in Chapter 5 by Jim Butler, A. Cheesman and Michael Ashford whilst the last chapter in this part (Chapter 6 by Matthias Schreck) summarises the progress in heteroepitaxial growth: the quest for a wafer size single crystal diamond.
The second part (Chapters 7 to 11) is devoted to the various applications where diamond has been used as a radiation detector. This field of application is particularly interesting for diamond because in most devices the active layer consists of intrinsic diamond. By removing the need to use doped active layers the device structures not only become simpler but it also becomes possible to preserve the optimum properties of diamond such as the large electron and hole mobilities, long recombination lifetimes and large dielectric breakdown voltages. This is one reason why radiation sensors have been investigated since the early days of diamond research and were one of the first electronic applications actively pursued since the early 1990s in CVD diamond for the various areas of use: far UV, high-energy photons (X- and gamma-rays) and high-energy nuclear particles. In this part, these applications are reviewed. Because of its wide band gap, diamond has been proposed as an ideal solar blind UV and soft X-ray detector for solar space missions, and because of its tolerance for radiation, as a candidate for detectors in deep-UV photolithography. These applications are discussed by Alan Collins in Chapter 7. There are several attributes that make diamond attractive for radiation detectors in medical radiotherapy applications, an area of application where diamond has been investigated for nearly five decades, well before CVD, and is probably one of the few examples where the only commercially available device is still made from a natural diamond stone. The use of CVD diamond in this application has been studied since the early days, first using polycrystalline and, more recently, single crystal specimens. The progress achieved and current status is reviewed by Mara Bruzzi in Chapter 8. The high predicted radiation hardness of diamond and the possibility of realising large area and affordable detectors using CVD diamond motivated a long and fruitful research collaboration between industry and scientists from CERN (RD 42 Collaboration), which started in about 1994. In many ways, this collaboration was the driving force for the improvement in the quality of CVD diamond, resulting in an increase in collection distance from under 10 micrometres to the current value of over 250 micrometres (in polycrystalline material) and much larger in single crystals. The other outcome of this research activity was the confirmation that diamond is indeed radiation hard, better than silicon by at least a factor of ten. It has been announced recently that CVD diamond detectors will be used as beam monitors in the new Large Hadron Collider, a modest but nonetheless rewarding outcome from these many years of research. Chapter 9 by Harris Kagan and William Trischuk describes the achievements in this field. The detection of heavy ions is one application where diamond seems to excel: the signals are good because a large amount of charge is generated, the detectors can be very fast and, of course, radiation hard. Another close collaboration between industry and the Institute for Research in Heavy Ions (Gesellschaft f¨ur Schwerionenforschung, GSI)
in Darmstadt has resulted in many different applications for CVD diamond detectors (using both polycrystalline and single crystal material), and has also been a driving force for the improvement of material quality and the development of specialised electronics. This progress is described in Chapter 10 by Eleni Berderman and Mircea Ciobanu. We finish this section on radiation detectors with Chapter 11 by Gianluca Verona-Rinati, who describes a neutron detector made from single crystal CVD diamond. This device is able to detect both thermal and energetic neutrons and has undergone prolonged periods of test in TRIGA RC-1, a fusion research reactor, and in the Joint European Torus, a magnetic confinement fusion reactor, exhibiting negligible degradation and confirming the high radiation tolerance of diamond.
Chapters 12 to 15 are dedicated to what are recognised in the electronic industry as the major active devices: rectifiers and high power switches, transistors and opto-electronic devices. The underlying thread in all these chapters is the creative effort that is being made to overcome what seems to be a fundamental limitation of diamond as an electronic material: the difficulty of doping. Chapter 12 by Jan Isberg summarises the efforts in the development of efficient rectifiers and switches for the control of high powers. In Chapter 13, Makoto Kasu describes some of the very impressive results that have been achieved using the surface p-type layer in hydrogenated diamond as the active layer for the fabrication of field-effect transistors (FETs). In these devices, long-term stability still seems to be a limitation. The use of boron doped diamond as the active layer would in principle remove the stability problem. In Chapter 14, Erhard Kohn and Andrej Denisenko describe how the major limitation of low boron activation may be overcome by the use of a delta-doped layer. They extend their brief by commenting also on how this concept may be used for high power rectifiers and other devices, and give a very comprehensive description of the technologies that need to be used for the processing of diamond in the fabrication of devices. Boron incorporates easily into the diamond lattice and, although it results in a relatively deep acceptor level at 0.37 eV, it can be used to make either semiconducting or almost metallic diamond by increasing the concentration. N-Type doping is considerably more difficult. So far the only reliable n-type dopant that has been found is phosphorus, which is more difficult to incorporate and results in an even deeper defect centre near 0.6 eV. As a consequence, n-type doping has not so far been very useful for conventional p–n type devices, either rectifiers or transistors. One exception seems to be light-emitting devices. Although diamond is an indirect band material, by making use of exitonic emission, interesting UV light emitting diodes have been demonstrated as described in Chapter 15 by Toshiharu Makino and Hiromitsu Kato.
Part 4 addresses biological and electrochemical sensor applications. Following the work on surface modification by hydrogen or oxygen, it was discovered that the surface of diamond could be functionalised to accept different types of protein. This has opened a seemingly vast field of potential applications, which, if successful, being mostly in the medical diagnostics field, could result in large markets. This subject is described by Jose Garrido in Chapter 16. Relatively early in the development of CVD diamond applications, it was found that electrically conductive diamond had interesting and promising properties as an electrochemical electrode. It combined a relatively large voltametric potential with the established chemical inertness of diamond. Further work showed that these properties could be used to make electrodes for a number of applications ranging from water purification to chemical analysis. The subject is complicated because CVD diamond prepared by different techniques behaves very differently. The scope of this application with emphasis on electrochemical sensors is described in Chapter 17 by John Foord. Chapter 18, by Joachim Kusterer and Erhard Kohn, concerns the numerous MEMS (micro-electro-mechanical systems) that have been made from CVD diamond using either conventional polycrystalline material or some of the finer grain versions such as nano- or ultrananocrystalline diamond. The group led by Erhard Kohn at the University of Ulm has been at the forefront of this technology and has developed an impressive range of different actuators and sensors, using almost all of the possible advantages that the different types of CVD diamond material are able to offer. This chapter discusses how and in which areas diamond might be able to replace silicon in the MEMS field.
Finally, I thought that we should include in a book on diamond electronics an account of a relatively newly discovered property of boron-doped diamond: superconductivity. If combined with the other attractive properties of diamond the possibility remains open to the development of superconductive devices, although these have not yet been demonstrated. This subject is presented in Chapter 19 by Yoshihiko Takano. The question remains as to whether CVD diamond will ever become a major money earner as an electronic material or whether it will continue to address only low volume niche applications. Some of the serious limitations of diamond are relatively well known: the difficulty in doping, size of single crystal specimens and, not least, the added hurdle of having to compete with the successful advances achieved in the other wide-gap materials: SiC and the III–V nitrides. In a series of publications by Alan Collins over a period of almost a decade from 1989 to 1998, some of the hurdles in developing a successful market for diamond electronics are analysed. These papers make sobering reading and should, I suggest, be compulsory for anyone wishing to invest in this technology. I am not sure if any of the chapters in this book succeed 100% in proving Alan Collins wrong. However I believe they do succeed in showing possible avenues that may be successful in the future. What is also apparent from all the contributions in this book is that there is still a very considerable motivation to continue the research and development activities in diamond electronics. Although diamond is behind the other wide-gap semiconductors in terms of successful applications it still has the allure of being an elemental material with none of the stoichiometry problems and related defects of compound materials. In this regard it may be interesting to note that one of the major CVD diamond producers (E6) has recently started a joint venture (Diamond Microwave Devices) to develop and exploit CVD diamond RF devices. I would not like to speculate as to whether the creation of this company is in response to a product they believe to be close to market or a leap of faith, only time will tell, but I am sure I am joined by the rest of the CVD diamond community in wishing them every possible success. The dream is clearly alive and most certainly continues.
Content
- Basic Properties, Defects and Impurities, Surface properties and Synthesis
- Basic Properties of Diamond: Phonon Spectra, Thermal Properties, Band Structure
- Transport Properties of Electrons and Holes in Diamond
- Point Defects, Impurities and Doping
- Surface Conductivity of Diamond
- Recent Progress in the Understanding of CVD Growth of Diamond
- Heteroepitaxial Growth
- Radiation Sensors
- Detectors for UV and Far UV Radiation
- Diamond Radiation Sensors for Radiotherapy
- Radiation Sensors for High Energy Physics Experiments
- CVD-Diamond Detectors for Experiments with Hadrons, Nuclei, and Atoms
- Neutron Detectors
- Active Electronic Devices
- High-Power Switching Devices
- H-Terminated Diamond Field-Effect Transistors
- Doped Diamond Electron Devices
- Optoelectronic Devices Using Homoepitaxial Diamond p –n and p –i –n Junctions
- Electrochemical and Biological Sensors
- Biofunctionalization of Diamond Surfaces: Fundamentals and Applicatio
- Diamond Electrochemical Sensors
- Micro-Electro-Mechanical Systems
- CVD Diamond MEMS
- Superconductivity in CVD Diamond
- Superconductivity in Diamond
