Technology News
10/01/2000
 |
null
IBM researcher "turns on" first quantum computer
While practical quantum computers are still very much a scientific endeavor, a presentation at the recent Hot Chips 2000 conference clearly confirms what has been a rather phenomenal theoretical prediction. Specifically, data presented by Isaac Chuang (Fig. 1), research staff member at IBM's Almaden Research Center, San Jose, CA, confirms predictions made earlier this year — by Prof. Richard Cleve of the University of Calgary in Canada — that these remarkable systems will be able to solve problems that are difficult even for today's supercomputers.
 |
Figure 1. Isaac Chuang holding a quantum computer ? a glass tube containing specially designed molecules that can solve some of the most difficult mathematical problems exponentially faster than the fastest conceivable classical computer.
Chuang, who leads a team of scientists from IBM Research, Stanford University, and the University of Calgary, has a system that is a 5-qubit ("quantum bits") quantum computer at IBM's Almaden Research Center. Its 5-qubits are five fluorine atoms (19F) within a molecule specially designed so the fluorine nuclei's spins can interact with each other (Fig. 2). These are programmed by radio frequency pulses and detected by nuclear magnetic resonance instrumentation (the latter similar to NMR technology used in hospitals).
Using the molecule, Chuang's team solved in one step a mathematical problem for which conventional computers require repeated cycles. The problem is "order-finding" — finding the period of a particular function. This problem is typical of many basic mathematical problems that underlie important applications, such as cryptography.
 Figure 2. Structure of the penta fluorobutadienyl cyclopentadienyldicarbonyl-iron complex used in IBM's 5-qubit quantum computer. |
Order-finding can be described by considering a large number of rooms and an equal number of randomly placed one-way passages, some of which may loop back upon themselves into the same room. It is certain that at some point, a person moving through the rooms and passages will return to the starting room. The problem is to calculate — with the least number of queries — the minimum number of passages through which one must travel before returning to the starting room.
The highlight of Chuang's presentation at Hot Chips 2000 was that while a conventional approach would require up to four computing steps, the 5-qubit quantum computer solved any case of the problem in one step. Quantum computers depend on the often-nonintuitive quantum-physics interactions of atoms and nuclei, compared with the flow of electrons through transistors and circuits used in conventional electronic computers. Quantum computers get their power by taking advantage of certain quantum physics properties of atoms or nuclei that allow them to work together as qubits to be the computer's processor and memory.
According to Chuang, the crux of this work lies in the use of the quantum Fourier transform (QFT), which allows one to efficiently determine the unknown periodicity of a function given as a black box. "In our experimental work, the spins of five fluorine nuclei in a molecule subject to a static magnetic field acted as the quantum bits. These bits were manipulated and read out using room temperature NMR techniques," he said. "The bottom line is that our work was made possible by the synthesis of a five spin molecule with excellent spectral properties, and by the development of new methods and experimental techniques. These include the invention of a more efficient and effective temporal averaging scheme for initial state preparation, and the construction of strategies to control the dynamics of five spins using a four-channel spectrometer, and by the invention of a technique to simultaneously rotate multiple spins at nearby frequencies."
While Chaung's 5-qubit quantum computer is the most powerful one demonstrated to date, the caveat is that several-dozen-qubit-quantum computers are needed to solve significant problems and practical application still faces daunting challenges.
"This result gives us a great deal of confidence in understanding how quantum computing can evolve into a future technology," Chuang said. "It reinforces the growing realization that quantum computers may someday be able to live up to their potential of solving, in remarkably short times, problems that are so complex that the most powerful supercomputers can't calculate the answers even if they worked on them for millions of years."
Chuang says, "Quantum computing begins where Moore's Law ends, about 2020 when circuit features are predicted to be the size of atoms and molecules. Indeed, the basic elements of quantum computers are atoms and molecules."
Looking forward, Chuang foresees that the first applications are likely to be as a co-processor for specific functions, such as database lookup and finding the solution to a difficult mathematical problem. — P.B.
Interest in isotopically pure silicon growing
Eagle-Picher Technologies, Joplin, MO, is quadrupling its capacity for isotopic enrichment of silicon. Why? Researchers at MIT and other labs have demonstrated that the pure isotope 28Si has a thermal conductivity that is 60% greater than natural silicon at room temperature, and 40% greater at 100°C. This increase in silicon's thermal conductivity could have a profound effect on the ability of ICs to dissipate the heat that they generate.
Naturally occurring silicon is 92% 28Si, but removal of the other 8% — 29Si and 30Si — has a significant effect on the thermal conductivity. This has also been demonstrated with isotopic enrichment of 12C diamond, where removal of 1% 13C increases the thermal conductivity by 50% at room temperature and by an order of magnitude at very low temperatures. Even more dramatic results have been shown with an isotope of germanium, 70Ge. The basic mechanism is that the lattice vibrations (phonons) and electrons that conduct the heat can travel with less scattering if all of the atoms are identical.
Isonics Corp., Golden, CO, a customer of Eagle-Picher, is driving much of the development in this area. The company recently announced the funding of three new research programs to investigate the benefits of 28Si. Isonics manufactures wafers (up to 200mm dia. by the end of 2000) from material supplied by Eagle-Picher and, according to Steve Burden, VP of semiconductor materials at Isonics, the company's only previous source was located in the former Soviet Union. With another source, the R&D can accelerate. The new programs are with the Power Semicon-ductor Research Center at North Carolina State, Southern Methodist University, and the National Renewable Energy Laboratory (Golden, CO). This brings the total number of institutions working with 28Si supplied by Isonics to 12. This number includes IC manufacturers such as Cypress and AMD, so it is not an area of purely academic interest now.
The cost of 28Si wafers is of course a concern. Burden says, however, that the wafers could easily pay for themselves by using conventional cooling methods instead of more expensive cooling techniques — such as thermoelectric or cryogenic coolers — as microprocessor speeds reach several gigahertz in the next few years. — J.D.
Fully automated gas-standard calibration
Engineers at Innovative Lasers Corp., Tucson, AZ, have developed a novel gas calibration concept dubbed Automated Gas Standards (AGS) that is applicable for qualifying newly installed gas lines, for environmental monitoring, and, more generally, for any semiconductor manufacturing application requiring trace gas analysis in ultra-high-purity (UHP) gases.
"Our AGS technology is specifically designed for fully automated and rapid on-line verification of gas sensor performance, facilitating critical decisions involving gas quality much faster and with a more accurate set of data," says Markus Wolperdinger, director of product development at Innovative Lasers.
Through the ability of the system to switch automatically between a number of gases (i.e., external sample, calibration, and zero gas), the AGS concept overcomes the need in conventional calibration systems to repetitively disconnect and reconnect gas lines between the calibration system and the trace gas sensor. AGS systems operate on the principle of flow dilution, in which a well-controlled amount of a trace gas from permeation tubes is blended and diluted with a "zero-gas" of known purity.
"The conventional manual approach is time consuming and can lead to serious contamination of instrumentation, as well as the gas lines since it requires frequent modifications of gas lines connected to the gas sensor and the calibration system," says Wolperdinger.
Briefly described, the fully software-controlled AGS systems provide the following:
- Gas dilution and mixing via UHP mass flow controllers (MFC) dynamically maintain the gas flow and mix in all branches of the system with essentially zero dead volume, to guarantee maximum precision and repeatability of trace gas concentrations and to maximize the response of the AGS system to user-selected flow and concentration changes. All components utilized in the AGS system are individually traceable to NIST standards to ensure the highest accuracy of calibration gas concentrations and overall system performance.
- Automatic routine switching between external sample gas, calibration gas, and internal "zero gas" with minimal user input and without modifications to the gas line system is facilitated by system control software that is intuitive and easy-to-use.
 Moisture in nitrogen ramp-up and dry-down sequences generated with an AGS system and measured with a commercial quartz-crystal-microbalance-based moisture analyzer. |
Engineers at Innovative Laser have tested AGS systems with different commercial trace gas analyzers over a wide dynamic range from single-digit ppb to high ppm levels. The overall accuracy of trace gas concentrations generated with AGS systems is typically better than ±2.5%. With a sequence of four individual ramp-up and dry-down sequences of moisture concentrations (200ppb to ~1.5ppm) generated in nitrogen dilution gas over >36 hrs, the reproducibility of moisture set-points generated by the AGS system was better the ±2% (see graph).
"We expect that our AGS instruments will lead the way to a new approach in trace gas-calibration applications with their ability to provide access to instant sensor performance verification without requiring modifications to gas supply lines once the instrument is installed," says Wolperdinger. "As a consequence, one can anticipate significantly reducing the risk and financial impact of delayed and incorrect decisions concerning process management." — P.B.
DNA molecules may nanofabricate future computers
Using DNA's ability to precisely match its constituent pieces, scientists at Lucent Technologies' Bell Labs are working to attach DNA to electrically conducting molecules to assemble rudimentary molecular-scale electronic circuits. The eventual application may result in computers that are 1000x more powerful than today's systems.
 The DNA sequence showing how DNA "electronic tweezers" form; the blurred portion of the image shows the range of motion during the entire nanofabrication process. |
The basis of this work is reflected in the August 10 announcement in the British journal Nature where Bell Labs researchers describe the ability to design pieces of synthetic DNA whose opening and closing stages can be precisely controlled; the process has been dubbed "a synthetic molecular motor." The controllability of the process is that the V-shaped structure of single-stranded DNA molecules can be "fueled" by an additional strand to "zip" close and then opened with a "removal strand" that wrests the fuel strand away from the structure (see illustration). The opening and closing, much like electronic tweezers, can be repeated by successively adding fuel and removal strands to a solution containing the DNA.
This work was led by Bell Labs physicist Bernard Yurke and included Andrew Turberfield, a physicist at the University of Oxford, who spent a recent sabbatical year at Bell Labs, physicist Allen Mills, post-doctoral fellow Friedrich Simmel of Bell Labs, and Rutgers University graduate student Jennifer Neumann.
Because the researchers could not observe the DNA motors with available microscopy, they relied on fluorescence to detect the closing and opening actions. A pair of dye molecules was attached to the ends of the DNA "motors" and when laser light excited the dyes, the amount of fluorescent light indicated the distance between the two ends.
The DNA motors are an example of nanoscale self-assembly, where molecules are mixed together in a solution to find each other and automatically combine to construct a device.
Yurke said, "This may lead to a test-tube based nanofabrication technology that assembles complex structures, such as electronic circuits, through the orderly addition of molecules."
Bell Labs physicists devise plastic laser
Physicists at Lucent Technologies' Bell Labs have developed the first electrically powered injection laser from an organic semiconductor. They now foresee that such "plastic" laser technology may lead to more widespread use of lasers. The work of Bertram Batlogg, head of Bell Labs' solid state physics research department, and Christian Kloc, Hendrik Schon and Ananth Dodabalapur was recently reported in the July 28th issue of Science, a journal of the American Association for the Advancement of Science.
Injection lasers are attractive because they are compact and can operate with small power sources, which expands their possible applications. Previous organic lasers had been powered by light sources, such as other lasers, which limited their applications.
To make the plastic laser, Bell Labs scientist Kloc first grew high-quality crystals of tetracene — an organic molecule with four connected benzene rings that conducts electricity well (see image). Then, when the researchers injected electric current to excite the tetracene to emit light, the light bounced back and forth between mirrors in the material, eventually producing beams of intense yellow-green light.
"Previously, researchers in the laser community thought organic materials would never be able to carry the large current necessary for electrically driven plastic lasers," said Batlogg.
Tetracene, however, is among the purest organic semiconductors, which leads to the desired electrical properties needed for a laser. "The tetracene crystal remains transparent just before intense beams of light are formed," Batlogg said, "so it absorbs very little light, which enhances the lasing effect."
Since organic materials are less expensive than the most common semiconductor ones (such as gallium arsenide) in lasers, organic materials may decrease production costs of individual lasers. Alternatively, it may be possible to have several hundred lasers, instead of just one, for the same price in a machine, such as an optical storage device or laser printer, which would speed access or display of stored information.
"These research results open up a whole new set of possibilities for electrically driven lasers," said Federico Capasso, physical research VP at Bell Labs. "They not only would be inexpensive to manufacture, but they could be tailor-made to produce a wide range of wavelengths, each of which could have specific applications. They can be driven by today's silicon circuitry and may someday be combined with plastic transistors to further reduce production costs and potentially lead to lightweight, flexible products."
A few months ago, Batlogg and his colleagues reported in Science and Nature that they used single crystals of pentacene — a molecule with one more benzene ring than tetracene — to build organic field-effect transistors (J.H. Schon, et al., Science Vol. 287, 2000, p. 1022 and J.H. Schon, et al., Nature, Vol. 403, 2000, p. 408.)
Because the current configuration of the Bell Labs plastic laser operates at a visible wavelength, it is not yet appropriate for optical communications. — P.B.
The history of quantum computer research
When quantum computers were first proposed in the 1970s and 1980s by theorists such as the late Richard Feynmann of California Institute of Technology, Pasadena, CA, Paul Benioff of Argonne National Laboratory, IL, David Deutsch of Oxford University, England, and Charles Bennett of IBM's T.J. Watson Research Center, Yorktown Heights, NY, many scientists doubted that the computers could ever be made practical. But in 1994, Peter Shor of AT&T Research described a specific quantum algorithm for factoring large numbers exponentially faster than conventional computers, fast enough to break the security of many cryptosystems. Shor's algorithm opened doors to much more effort aimed at realizing the quantum computers' potential.
Significant progress has been made by numerous research groups around the world. Isaac Chuang is currently among the world's leading quantum computing experimentalists.
In 1996, Chuang co-invented, with Neil Gershenfeld of MIT, the quantum computing method on which he is working.
In 1998, he demonstrated that the NMR technique could implement basic elements of a factoring algorithm and a database-search technique. He is currently exploring the prospects for scaling up the NMR approach to solve more complex problems. He also led teams that demonstrated the world's first 2-qubit quantum computer in 1998 at University of California at Berkeley and a 3-qubit quantum computer in 1999 at IBM-Almaden.
The order-finding result announced at Hot Chips 2000 is, reportedly, the most complex algorithm yet to be demonstrated by a quantum computer. Earlier this year, scientists at Los Alamos National Laboratories announced they had achieved quantum coherence in a 7-qubit molecule. While this is a necessary condition for achieving a quantum computer, they have not yet used the molecule as a 7-qubit quantum computer to solve a problem or to implement a quantum algorithm. — P.B.
Coming soon: Your WIP is calling
According to Douglas Scott, strategic marketing VP at PRI Automation, Billerica, MA, in the semiconductor manufacturing factory of the future, you won't need to know where your WIP is.
 |
"If it needs help, it will call you," says Scott. In his scenario for ~2005 "you are enjoying a mountain hike. At the fab, your engineering lot has reached the checkpoint you specified for reviewing some critical metrology results. It calls you. You pause on the trail to review the results and adjust some control parameters for the next process step. To verify the impact of your proposed change, you run a process simulation, and tweak the adjustments a bit. You transmit the adjustments back to the fab, send your experiment on its way, and resume your hike."
How are we going to get to this point? "Through CIM," says Scott. "By fully integrating the fab-wide material handling systems with the many different software components to create a highly intelligent and productive manufacturing environment: a 'silicon machine' capable of producing large quantities of complex semiconductors in the least amount of time and at the lowest cost."
Moisture in nitrogen ramp-up and dry-down sequences generated with an AGS system and measured with a commercial quartz-crystal-microbalance-based moisture analyzer.
Tech Briefs
Microlithography industry experts attending the International Sematech 157nm Pellicle Risk Assessment Workshop support concentration of pellicle development efforts on soft-polymeric pellicles for emerging 157nm lithography. The attendees also highlighted the need for back-up research into hard pellicles using modified fused silica. "No pellicle solution today meets all the requirements for 157nm lithography," said Gerhard Gross, director of lithography at International Sematech. "There are risks associated with making a decision this early in the game, but it is necessary to help the industry move ahead. Without a clear direction now, we won't be ready in time." Meeting participants selected a primary and secondary path from among three options: soft, hard and no pellicles. In each case, the favored option was selected by more than 75% of the respondents.
To support expected strong demand for its ArF excimer laser scanners, Canon is building a plant to increase its supply of calcium fluoride (CaF2) lens material. Through its fully owned Optron Inc. subsidiary, Canon has obtained 9 acres of land in Yuuki City, Ibaragi Prefecture, and will begin constructing a new 107,500 ft2 factory for the intensive production of CaF2 glass by mid-2001. The new factory will install an electrical furnace for the removal of impurities from the raw calcium fluoride ore. The estimated cost of the new facility is US$40 million (4 billion yen). Canon plans to begin large-scale commercial manufacture of its second-generation, 200mm/300mm ArF excimer laser scanner in 2001. A 300mm-capable first generation ArF tool has been at work at Selete since late 1999. Canon will produce its new scanner at its Utsunomiya Optical Factory in Japan.