Table of contents

A summary is given of some research carried out at Texas Instruments over the past few years in the area of quantum semiconductor devices and very small semiconductor structures. The ultimate goal of this work is the creation of a revolutionary digital integrated circuit technology which will permit the downscaling of minimum feature sizes in integrated circuits to continue well into the next century, while maintaining non-cryogenic operation.

Advances in micromechanics, the construction of mechanical structures using microfabrication techniques, have come swiftly over the past five years as more and more engineers have entered into research on this topic. A major emphasis has been on the development of surface micromechanics a term employed to describe the fabrication, using thin-film-deposited materials, of structures on a plane surface-usually that of a silicon wafer. Rapid progress has been made in research on surface micromechanics because it has concentrated on processes and materials that have been refined to an extraordinary degree to enable the incredibly high performance of today's integrated circuits. Micromechanisms having appreciable complexity have already been built and demonstrated. The dimensions of these mechanisms are generally smaller than 0.1 mm. Rotating micromotors, powered by electrostatic drive, have recently been demonstrated and other actuation techniques show promise for applications to structures that are useful to engineers and scientists. A growing body of research is being directed toward the controlled motion of these tiny structures, a field that has been named microdynamics. Research on mechanical properties of the materials used in surface micromechanics is needed because most such materials have only been characterized electrically. Determining the process dependence of these mechanical properties is an ongoing challenge. Skill in exploiting the special microfabrication techniques developed for integrated circuits can speed evaluation of properties and structural ideas. The results attained thus far have served to arouse considerable interest in researchers from an increasingly widening range of disciplines.

A Japanese national project in the field of nanotechnology is now in action. This project is based on the innovative research effort ERATO program. The principles of the ERATO program and the activities of the project are described in this paper. The Yoshida Nano-Mechanism Project is one of the projects in the ERATO (Exploratory Research for Advanced Technology) program of JRDC (Research Development Corporation of Japan). JRDC is a statutory corporation of the Japanese Government, and is run by the Science and Technology Agency. JRDC administers and fully funds the ERATO program. The research fields and project directors are selected and supported by JRDC, and the researchers are employed by JRDC for each project. The project has funds of US$12 million to be used over the five-year lifetime of the project. The project started in October 1985. New scientific approaches to the fields of measurement, control and processing with nanometer precision are being studied. Physical phenomena and mechanical properties of matter in the nanometer region will be studied. Some specific goals have been chosen in the form of instruments or components to be built. The development of these instruments or components is important, but more important will be the technologies developed in the building of the instruments and the understanding derived from using these instruments. The project is now approaching its climax.

This paper demonstrates the capability of X-ray interferometry for the measurement of the detailed parasitic motions that may occur in very high precision slideways. X-ray interferometry has been used for the first time to obtain simultaneous fringes by motion of the sensing crystal in two orthogonal directions, giving (with two independent X-ray sources) fringes corresponding respectively to the lattice spacings of (111) and (220) crystal planes. A monolithic X-ray interferometer was designed for this experiment, with two sets of leaf spring supports for the moving blade, so that it could be driven in the two orthogonal directions by magnet/coil transducers. A phase plate was positioned and moved so that three spatial-phase signals could be obtained for each positional measurement of each of the two axes, providing measurement accuracy of the order of 20 pm for the conditions of this experiment.

The limitations to the accuracy achievable by optical interferometers when they are used to measure sub-nanometric path differences are discussed, and techniques for realising the optimum performances from these instruments are described. An optical design for a two-beam interferometer is proposed that would maintain the accuracy of the instrument when resolving with subnanometric sensitivity and, in addition, facilitate the alignment of the optical and mechanical axes of the measurement system.

A one-axis mechanism, driven by an AC linear motor and guided by a rolling ball guide, has been constructed. The micro-dynamics of the rolling guide shows that the group of balls of the guide behaves as a rigid spring within a displacement of about 100 nm, and normal rollings take place for a displacement greater than 100 mu m. These different characteristics are used to control the stage during coarse and fine positioning. A coarse positioning with 100 nm resolution, 200 mm s^-1 maximum velocity and 250 mm stroke, and a fine positioning with better than 1 nm resolution without backlash are achieved. The switching from the coarse positioning to the fine positioning is studied.

A surface-texture and profiling instrument for sub-nanometre surface measurement based on the NPL Nanosurf 2 is described. It has a horizontal dynamic traverse range of 50 nm to 50 mm, which is achieved by the use of an ultraprecise slideway having low environmental susceptibility. Ergonomic and functional design differences, and the reasons for these, are discussed, together with the results so far obtained from the first instrument, called Nanostep.

In order to obtain single-point diamond tools (SPDT) with cutting edges of sharp or small negative land for ultra-precision turning of soft materials, machining of the SPDT by an argon ion beam of 1.0 keV has been investigated. The edge radius of the SPDT was measured using a profile SEM (scanning electron microscope having two secondary electron detectors). By this processing method, mechanically lapped or polished SPDTS which have a mean edge radius of 70 nm were sharpened, and their edge radii became of the order of 20-30 nm. Moreover, is is found that the processing method can be useful for improving the surface topography of the SPDT formed by mechanical lapping removing micro-chippings on the cutting edge of the SPDT, and improving the straightness of the cutting edge of the SPDT. The processing method, also, can result in the formation of a small negative land of sub-micrometer width on the cutting edge of the SPDT.

The topography of the (110) surface of p-doped GaAs has been investigated by scanning tunnelling microscopy for various voltages between the tunnelling tip and the GaAs sample. At +2.5-+3.5 V sample bias the topography is predominantly characterised by a row structure parallel to the (110) direction. However, at a sample voltage of +1.5-+1.8 V the morphology is dominated by rows parallel to the (001) direction. A continuous transition between the two structures is observed at intermediate bias voltages. An attempt is made to explain this observation in terms of the electronic states involved at the surface.

The combined developments of sub-micrometre integration in IC technology and high speed data handling in digital image processing have provided the means of introducing a new concept of 'vision' on a nanometre scale,which we call 'nanoscopy'. Using advanced optical imaging techniques under computer control and processing, high resolution image synthesis of certain objects can be carried out so long as they can be defined by separable functions. Measurements of position two to three orders of magnitude below the diffraction limit can be made. Nanoscopy is illustrated by three techniques applied to defect analysis in semiconductor materials. Nanoscopy techniques are non-contact, non-destructive, and give large amounts of quantified information concerning the 3D distribution of defects. Because they give nanometre resolution within a relatively large field of tens to hundreds of micrometres, the location of position or of particular features is simplified, which can be a problem with super resolution techniques such as electron microscopy and scanning tunnelling microscopy.

The electronic transport through three-dimensionally confined semiconductor quantum dots' is investigated and analyzed. The spectrum corresponds to resonant tunneling from laterally confined emitter contact subbands through the discrete three-dimensionally confined quantum dot states. Momentum non-conservation is observed in these structures.

The scanning tunneling microscope (STM) has proven to be an unsurpassed instrument for studying surfaces with sub-nanometer resolution in all three spatial dimensions. The same circumstances that permit the STM to image with atomic resolution also allow it to function as a highly localized probe capable of exerting human influence upon regions of only nanometers in extent. Demonstrations of the STM's ability to modify matter at the nanometer scale give evidence of its promise as a tool for nanofabrication applications. In addition to presenting some recent results of STM modifications of Au and graphite substrates, this paper introduces a system for classifying the various mechanisms by which an STM may induce modifications, and classifies published and new experimental observations within this framework.

Four novel schemes are introduced for controlling the motion of a Si mu -actuator directly using a low-power optical signal: (i) photoelectric leakage current, (ii) electric field screening in a semi-insulating layer, (iii) modulating the leakage current of a p-n junction and (iv) modulating the permittivity of a dielectric gas. All use a parallel-plate capacitor with a flexible Si cantilever as one plate. Experiments with the photoelectric device demonstrate for the first time light-controlled actuation, over a distance approximately 0.1 mu m in a time <0.5 s. A figure of merit, defined as a product of useful force and displacement over the product of required power and actuation duration, is introduced for evaluating and comparing different mechanisms that are used to achieve mu -actuation. The figure of merit for direct optical mechanisms are: mu approximately=0.34 for the photoelectric actuator, mu approximately=1 for the electric field screening controlled actuator, mu approximately=1 for the p-n junction leakage current controlled actuator, and mu approximately=2.5*10^-6 for the permittivity modulated actuator. Direct optical mechanisms are faster and use less power than thermal techniques ( mu approximately=10^-5) and purely electrostatic actuators ( mu approximately=0.1).

Recent advances in surface scanning microscopes (STM, AFM) and nanometre topographic instruments warrant a new look at the criteria of performance required and how to achieve them. In this paper the basic system is examined with particular emphasis being placed on the role of damping in the mechanical system. Two concepts are introduced, integrated damping and differential spatial damping. Also, the use of scanning velocity to couple the instrument and surface coordinate systems is demonstrated.