0
    Nova
    Tech + EngineeringTech & Engineering

    Diamond Factory: Expert Q&A

    On July 7, 2009, James Butler answered questions about diamonds, both natural and the engineered kind.

    Nova

    Q: Once a seed diamond becomes a grown man-made diamond, can it then be used to generate more seed diamonds? Dave Defino, Stroudsburg, Pennsylvania

    James Butler: Yes, and it can often lead to improved quality in the newly grown CVD [chemical vapor deposited] diamond. Often the quality and defect structure of the CVD grown layer is limited by the defects inherent in the seed crystal. If one gets the growth process right, some of these defects will not propagate or disappear as the new CVD material grows thicker. Similar effects were observed in many other crystal systems, including silicon, where 50 years ago the "best" quality was barely one-inch-diameter, with many impurities and intrinsic defects. Now, nearly perfect, pure 16-inch-diameter silicon wafers are available.

    Q: Why do diamonds conduct heat energy so well but not electricity? Charley Cox, Pensacola, Florida

    Butler: In metals, heat is conducted by the electrons, which also conduct charge (electricity). In diamond, heat is conducted by the lattice vibrations (phonons), which have a high velocity and frequency, due to the strong bonding between the carbon atoms and the high symmetry of the lattice. These strong bonds between the carbon atoms also give rise to many other of the extreme properties of diamond, such as hardness, chemical erosion resistance, electrical insulating strength, and optical transparency. As discussed below in the next answer, adding impurity atoms (dopants) can make diamond electrically conductive.

    Q: I saw the NOVA scienceNOW special on diamond electronics. I understand that pure diamond is a fantastic electrical insulator, and that it becomes a conductor when boron is added during the manufacturing process. The program suggested it could be used as a switch (like a transistor, I presume). Is diamond only useful as a substrate or insulator like silicon? Can it be made to "switch?" If so, how? Jack F., Alpharetta, Georgia

    Butler: The building blocks of most electronics are resistors, capacitors, inductors, diodes, and switches. Diamond materials can be useful for most of these components. The simplest way is the use of diamond as a heat spreader—to conduct the heat away from a local hot spot, which degrades the performance of a device built from another material, e.g., silicon or gallium arsenide. Pure diamond is an excellent insulator, but when charge carriers (electrons or holes) are injected into diamond, they can move with extremely high velocities. One can incorporate impurity atoms, such as boron or phosphorous, during the growth of diamond, and these atoms can donate an extra electron (in the case of phosphorous), creating an n-type semiconductor, or accept an electron (in the case of boron) from the valance electrons of carbon to create a p-type semiconductor.

    There are many ways to design a switch for electricity in a semiconductor, but the most common is a transistor. Transistors are fabricated from combinations of p- and n-type semiconducting and insulating materials, all of which are now possible with diamond. The combinations of the extreme properties of diamond—thermal conductivity, insulating strength, high charge carrier velocities, low dielectric constant, etc.—suggest that diamond should out-perform nearly every other semiconducting material system for electronic applications. IN PRINCIPLE! The reality is that there are many other factors involved in developing and implementing a technology: cost, manufacturing infrastructure, investment, and knowledge base. I think it is fair to say that diamond materials need a lot more research, knowledge, and technology development before they can be considered a mature semiconducting material.

    Q: Dr. Butler, Thank you for a peek into this new technology. As an older EE [electrical engineer], I am extremely enthused about the potential impact that diamond technology will have in the electronics industry. How long do you think it will be before the technology is used in practical applications? Marianne, Portales, New Mexico

    Butler: Diamond materials are already used in many practical applications: passive thermal management in advanced microwave and laser diode devices, optical windows for industrial laser machining, cutting tools for automotive and aerospace applications, electrodes for waste water cleanup, etc. How rapidly diamond materials are accepted into new (and old) technologies will depend on many factors, but I think it will mainly depend on the development and growth of the industrial base for CVD diamond. This will depend on both investment and the development of the economic drive for such investment. The insertion of CVD diamond materials into electronics is happening, albeit slowly, because of the availability of competing materials (Si, SiC, GaN), cost (which depends on infrastructure and demand), and knowledge (more research and technology development is necessary). Please see my answer to the previous question for more specific comments about diamond as a semiconductor.

    Q: Can diamonds be used in solar cells to prevent the efficiency degradation when they heat up? If so, would this be cost prohibitive? Anonymous

    Butler: Currently, it costs a fair bit of energy to grow high-quality diamond, so cost is likely to be an issue in large area/volume applications. I suspect there are many ways diamond might be used in solar cell technologies, but I am not conversant with the specific issues relating to efficiency degradation.

    Q: I understand that scientists are using the CVD process to grow carbon nanotubes as well. Can the same chemical vapors be used for both the process of artificial diamond growth and carbon nanotube growth? If not, what is it that requires these processes to be different? Greg, Virginia

    Butler: Yes, the chemistries for the growth of diamond and carbon nanotubes can use very similar reactants and apparatus, but the actual growth conditions are very different. Diamond CVD generally involves a large excess of hydrogen (and atomic hydrogen) and a hydrocarbon, usually methane. Carbon nanotube growth is best when very little hydrogen is present and generally uses acetylene, ethylene, or graphite as the carbon source.

    Q: The piece indicated that boron-laced diamond may someday be used to replace thousands of pounds of silicon transistors, hinting at lower energy requirements to accomplish large tasks due to lower-weight, more efficient transistors. If diamonds are to be included as part of the "green energy infrastructure," how "green" are they? For instance, how much electricity (e.g., heat required to create the carbon plasma and the associated carbon emissions) is required to produce one ounce of transistor-grade diamond? Dean Smurthwaite, Denver, Colorado

    Butler: Multiple silicon devices have to be used to switch high voltages and currents in applications like high-speed trains. Because the individual Si switches cannot withstand the high voltages, many devices have to be connected so that the voltage across each individual device does not exceed the limit of silicon. Complex circuitry called snubbers involving inductors and capacitors is required to protect the individual silicon switches from being destroyed by the full voltage. It is this circuitry that can add a ton or more of weight to each rail car. Devices that could withstand the full voltage would eliminate most of this extra weight, and diamond is a material which might enable this.

    There are other relevant factors such as the potentially higher switching speed of diamond-based devices, which would improve the compactness, weight, and efficiency of coupling various energy sources to the power grid. For these and other reasons, diamond could have a significant impact on "green energy infrastructure." The basic cost of producing CVD diamond exceeds $50 a gram, but material costs are generally insignificant in the production of complex, high-technology devices.

    Q: What temperature and pressure were used to grow the diamonds, and what gases were used? Ralph Schmidt, Stuart, Florida

    Butler: Most companies do not reveal their proprietary recipes. In general, the substrate temperature is between 400 and 1200°C (typically 850°C), and total gas pressure is between 0.007 to 1 atmospheres (5 to 760 torr) (typically 15 to 250 torr in plasma or hot filament enhanced growth processes). The common gases used are hydrogen (94 to 99.9% of the total) and methane, with occasionally some rare gas (argon) substituted for some of the hydrogen. However, almost any hydrocarbon source can be used and many oxygen containing gases can be used as long as the proper ratio of H/C/O is employed. Gases containing other atoms, e.g., N, B, P, Si, can be used as sources of impurity or dopant atoms. One can even grow diamond in air with an oxygen-acetylene torch under the proper conditions (including safety considerations).

    Q: How will the increase of synthetic diamonds impact the diamond market, a market renowned for its domination by companies like De Beers, and how will normal jewelers tell the difference between a synthetic one and a real one if the difference is microscopic? John Zandler, Phoenix, Arizona

    Butler: Synthetic (CVD) diamond will probably have very little impact on the diamond GEM market for many reasons: (1) Human nature seems to prefer "natural" gems with their unique imperfections; (2) the CVD diamond will be more valuable in technology than in competing with natural diamonds in the gem market; and (3) nearly 100 million natural diamonds are mined each year, and the total CVD production is a drop in the bucket compared with that. All competent gem labs have the capability to distinguish natural from CVD grown diamonds due to the presence or lack thereof of defects inherent to each type of diamonds. Good gemologists are also being trained in what characteristics to look for. The best of the CVD diamonds often stand out for being too perfect.

    Q: How does the cost of the process profiled on NOVA scienceNOW compare to conventional crystal growing of industrial diamonds? It looked as if there was a tremendous amount of energy used in the relatively small-scale operation that was profiled. Mark Palmer, Springfield, New Jersey

    Butler: Energy is one of the major costs in the CVD growth of diamond (capital investment and labor are the others). The process requires the dissociation of molecular hydrogen into atom hydrogen, the initiating reactant for both the gas phase and surface chemistry, by a plasma or hot surface, or in the case of combustion flame growth, the use of expensive chemical energy. I am not familiar with the costs of the high-pressure, high-temperature (HPHT) synthesis of industrial diamonds, but I suspect they are inherently very similar. The primary advantage HPHT might have right now is the existing capital investment and infrastructure.

    Q: Part of my duties as an electron microscopist was examining diamonds. We found that natural diamonds had pits, which looked like growth rings you find on wood. Have you found these growth pits on the diamonds you manufacture? I worked for Engelhard Industries, and we were attempting to make man-made diamonds to compete with the GE products. Milton Kranitz, Roswell, New Mexico

    Butler: Fluctuations in the growth process involving temperature, reactant or impurity gases, etc. can give rise to variations in defect concentrations observable optically, in luminescence, or in electron microscopies. These are some of the clues used by gemologists to determine the origin of diamonds. The as grown (and unpolished) surfaces of CVD diamond materials display local growth morphologies which are characteristic of the CVD growth process, but these are not directly present in a polished CVD gem.

    Q: What is the limit in size for items made from grown diamonds? Carl Kingery, Newport News, Virginia

    Butler: Size is only limited by the technology and size of the reactors for diamond materials. Single crystal diamond growth currently depends on the size of the single crystal seed available, and the engineering of the reactor. The largest CVD grown diamond I have heard reported is currently about 13 carats, but there are efforts to grow larger for use as diamond anvils to study the high-pressure physics and chemistry of planetary cores.

    Q: What kinds of by-products are created when making the diamonds? You said you use methane. Does that convert into something safer? Could this be used as a way to remove greenhouse gases? Penelope B., Portland, Oregon

    Butler: The by-products of the CVD growth of diamond are primarily waste heat from the plasma and substrate, hydrogen, and any unused reactant hydrocarbon gases. Since the CVD process is a net conversion of gaseous carbon to solid carbon, one might consider it as a very minor consumer of greenhouse gases.

    Q1: What do you see as the potentials for diamonds in semiconductor devices? How does it compare against GaAs [gallium arsenide] or other semiconductor materials in terms of performance and cost?

    Q2: Has there been any discussion of using diamond films for solar panels, for example for mission-critical hazardous conditions. What do you think of the potential?

    Q3: How many years away do you see CVD diamonds being used in large-scale applications? Is there large-scale government support on the research? Sean Shi, New York, New York

    Butler: Diamond has great potential as a semiconductor device, but much research and development is required before its full potential is realized. When one looks at the relative figures of merit of various semiconductor materials for a range of semiconductor device applications, diamond out-performs the other by orders of magnitude.

    BUT, there are also limitations to exploiting diamond as a semiconductor. A fundamental limitation is with the dopants to make p- or n-type doping (see also my answer to the third question above). The current dopants, boron and phosphorus, are not very active at room temperature (20°C) and require higher temperatures to perform well. Cost, available size, and the maturity of fabrication technologies are also important issues. For the foreseeable future, I doubt if diamond will replace other semiconductor materials in current applications, but diamond is likely to enable new uses in harsh and high-temperature environments.

    Diamond certainly is a material to consider for mission-critical hazardous conditions. There was significant U.S. government and industrial support for diamond research several decades ago, which provided the basis for the rapid development of the field to where it is today. But there has been very little research support in the last decade in the U.S., and the field is becoming dominated by research and production in Europe and Asia.

    Major funding for NOVA is provided by the David H. Koch Fund for Science, the NOVA Science Trust, the Corporation for Public Broadcasting, and PBS viewers.