Thermal fluctuations of the cantilever are a fundamental source of noise in atomic force microscopy. We calculated thermal noise using the equipartition theorem and considering all possible vibration modes of the cantilever. The measurable amplitude of thermal noise depends on the temperature, the spring constant K of the cantilever and on the method by which the cantilever deflection is detected. If the deflection is measured directly, e.g, with an interferometer or a scanning tunneling microscope, the thermal noise of a cantilever with a free end can be calculated from root kT/K. If the end of the cantilever is supported by a hard surface no thermal fluctuations of the deflection are possible. if the optical lever technique is applied to measure the deflection, the thermal noise of a cantilever with a free end is root 4kT/3K. When the cantilever is supported thermal noise decreases to root kT/3K, but it does not vanish.
We propose a simple model to describe the interaction of a forced cantilever oscillation with a specimen in a tapping-mode scanning force microscope experiment in order to make a rough estimation of the forces affecting the surface with each touch down of the tip. Assuming weak damping of the cantilever (quality factor of the cantilever between 100 and 1000) and of the surface, we can estimate the forces to be in the range of those in the contact mode. These forces can vary by orders of magnitude, e.g. 10(-6) to 10(-11) N. To reduce the interaction force we suggest scanning on the low-frequency side of the resonance frequency of the non-contact cantilever oscillation. Increasing the difference of phase between the non-contact oscillation of the cantilever in air and the oscillation during contact introduces strong Variations of the force. The improvement in resolution which can be achieved for soft samples by using the tapping-mode system results from the elimination of shear forces and the possibility of minimizing the force on the surface by varying the set-point of the scanning amplitude. Forces on the substrate will be enhanced by a large substrate stiffness.
Various types of molecular bearings have recently been proposed in the growing nanotechnology literature. Using novel molecular dynamics methods, we have simulated several model graphite bearings. The bearings varied in size from inner shafts of between 4 and 16 Angstrom in diameter, up to 120 Angstrom in length, and outer cylinders of between 10 and 23 Angstrom in diameter, up to 40 Angstrom in length. The turning shaft was either instantaneously started or torqued up to the desired rotational speeds. Frictional properties were size-, temperature- and velocity-dependent. The presence of more than one bearing vibrational mode in some simulations created beats that could possibly adversely affect bearing performance; placing a stretching tension on the bearing suppressed one of the modes and therefore the beats. These and future studies will help evaluate the performance, wear and load-bearing properties of fundamental components of nanomachines such as bearings.
Using molecular dynamics methods, we have simulated several model graphite nanometer-scale laser driven motors. To our knowledge, this is the first study of laser excitation of a nanometer-scare device. The motors consisted of two concentric graphite cylinders (shaft and sleeve) with one positive and one negative electric charge attached to the shaft; rotational motion of the shaft was induced by applying one or sometimes two oscillating laser fields. The shaft cycled between periods of rotational pendulum-like behavior and unidirectional rotation (motor-like behavior). The motor on and off times strongly depended on the motor size, field strength and frequency, and relative location of the attached positive and negative charges. In addition, the two-laser simulations showed much larger motor on times and more stable rotation than one-laser simulations. A mathematical model of the overall process was obtained by employing computational neural networks (CNNs). A CNN was able to 'learn' the mapping from size, charge position, frequency, and strength of the electric field to the motor on and off times. This multidimensional, nonlinear mapping was determined to within an average accuracy of 2% and could be used to determine initial parameters that would lead to better overall performance of the nanomotor.
The atomic force microscope (AFM) is used to map the local elastic properties of substrates by analysis of the force versus tip motion curves. Measurements are presented, which show that gold islands on a rough polypropylene substrate can be distinguished from the surrounding polymer. Quantitative calculations of the elastic deformations of the tip and of the sample, as induced by the AFM, were performed. Surprisingly, the tip deformation is predominant over the sample deformation in a wide regime of forces and of tip radii, which are commonly used in AFM. This fact limits the capability of the AFM to measure local elastic properties. However, with our experimental set-up one can induce a total deformation dominated by the sample deformations.
This paper discusses super-smooth polishing technology researched and developed for excimer lasers, soft x-rays, and other short-wavelength light applications. Short-wavelength light requires surface quality and contour accuracy superior to traditional specifications, as well as free-form contour elements. For this reason, the final target for free-form contours of 500 mm or more was set to 0.08 mu m PV for contour accuracy, and 0.2 nm RMS for surface roughness. To improve surface quality we employed local pitch polishing, utilizing a flexible tool laminated with an elastic layer, which adapts well to various contours. For greater contour accuracy, we developed the CSSP (Canon Super-Smooth Polisher), which polishes 500 mm optical elements, and has a unique structure incorporating an on-machine contour measurement device built onto a single base plate. The CSSP polishing process achieves a contour accuracy of 0.078 mu m PV and a surface roughness of 0.13 nm RMS on a 500 mm fused silica toroidal mirror. We also achieved the targets in the case of both CaF2 and CVD-SiC, materials widely used in short-wavelength light applications.
A new approach to the fabrication of model structures for nanoelectronics with characteristic sizes down to 10 nm is proposed. The approach consists in electron-beam-induced fabrication of self-supporting structures of nanometre sizes in a through slit formed in the substrate, followed by the deposition of a required material onto the structure, which serves as an active layer in a nanometre-scale device. Features of the fabrication steps are discussed. Bismuth nanobridges were fabricated and their voltage-current characteristics were measured, which demonstrated features of electron transport in these bridges connected with their small sizes and inner structures.
Automatic mask alignment in the theta direction using moire sensors is reported, The relation between small angular displacement and linear displacement is clarified. A computer-controlled angular alignment system is developed. A desired alignment position is easily set by a computer. In our experimental condition, angular accuracy of the order of +/-4 x 10(-7) rad is realized by automatic computer control.
A scanning probe microscope (SPM) was used as a tool for nanolithography. Silicon oxide (SiOx) patterns were formed on silicon (100) and (111) surfaces by means of scanning probe anodization, that is, localized anodization induced by an SPM tip. These anodic SiOx patterns could serve as masks for the chemical etching of Si in alkaline solution. Tetramethylammonium hydroxide (TMAH, (CH3)(4)NOH) containing no alkaline metal was used as the etching reagent. The roughness of the etched surface in this solution was one tenth that of the Si surface etched in an aqueous solution of potassium hydroxide which was used in our preliminary study. Consequently, on the Si(100) surfaces, we were able to fabricate grooves 40 nm deep and less than 50 nm wide and walls 30 nm high and 40 nm wide. Due to side etching, the aspect ratios of the structures which could be fabricated on the Si(111) surfaces were much lower than those fabricated on Si(100).
This paper reports a novel approach to the reconstruction of scanning probe microscopy (SPM) images by means of neural networks. The method aims to correct the integrating effect of a finite stylus tip. It is part of a general plan to enhance the performance of SPMs by means of neural networks. A well trained neural network is used in this approach to fulfil the nonlinear mapping from the apparent image to true surface. The results of experiments and simulations show that the reconstructed image tends to be closer the true surface than that measured images, and provides a better lateral resolution of measurements.
For the purpose of fabricating free-form optical elements to an accuracy of 80 nm PV, the CSSP (Canon Super-Smooth Polisher) has been developed. This device finishes workpieces by alternately repeating contour measuring and corrective polishing. In such a system, contour measuring is important because it limits the final accuracy of the workpiece. This paper focuses on the CSSP's on-machine contour measuring method. A contact probe is employed to ensure adaptability to free-form contours with maximum tilt angles of 35 degrees. A unique probe structure is proposed, by which both inclination and motion errors of the probe are simultaneously compensated. Th flaw problem is discussed from an experimental point of view; it was found that the major cause of flaws is dirt on the probe or the workpiece. Dirt also causes unstable irregular figure errors. Thus, the influence of dirt can be estimated from the results of contour measurement. A model of dirt is proposed, and the distribution on a width-height chart of irregular figure error predicted by the model agrees well with the experiment. A cleaning procedure was developed that is effective in reducing the problem of dirt. Using a ceramic air slide and a linear motor, the contact force was controlled to a constant of 2 mN, which is much smaller than the force of 260 mN that causes a yield stress on the CaF2 workpiece. The probe's scanning speed of 4 mm s(-1) was achieved by speeding up this force controller. The margin of force error during contour measurement was under 0.2 mN. A coordinate measuring method using fourteen-axis interferometers was also proposed for compensation of the major mechanical motion errors of the probe and the tables. Some methods of compensation for system errors are discussed. A new method was proposed by which three kinds of angle error, on the x, y, and z axes of measurement, are simultaneously compensated with a repeatability of 0.08 mu rad RMS. The measurement results showed good repeatabilities of 3 nm RMS for a 540 mm line measurement, and 9 nm RMS for a 500 mm aspherical surface measurement.
This paper presents a practical image-processing method to correct the in-plane geometrical distortion of an STM image, and to calibrate it using a regular crystalline lattice and two-dimensional FFT power spectrum. A dual tunneling unit STM with one X-Y stage and two independently controlled tunneling units in the two Z axes has been proposed for comparative length measurement using a regular crystalline lattice as a reference scale. To improve the measurement accuracy, the present image-processing method is applied to the dual tunneling unit STM and the experimental results, in which highly oriented pyrolytic graphite (HOPG) is used as a reference scale for measurement of the 10-320 nm length, show the feasibility of the present image-processing method and the possibility of comparative length measurement using the dual tunneling unit STM.
The scanning tunneling microscope (STM) is known for its high lateral resolution. The unreliability of the scanning and positioning motion of the STM were monitored by interferometer and capacitance recently. The widely used conventional lead zirconium titanate (PZT) tube scanner has good mechanical properties; however, the motion of the tube includes angular motion which generates large error. Even with an optical interferometer to calibrate STM tip motion, the angular motion reduces or eliminates the visibility of interference. We have developed an eight-segmented tube scanner which reduces angular motion for maintaining the visibility of interference and reduces Abbe error. The residual error of the value measured by a high-accuracy interferometer which measured the motion of the newly developed eight-segmented tube scanner was estimated by measuring a standard pattern. The accuracy of the interferometer, which used a balanced detection technique, the principle of which is also reported in this paper, was measured by a capacitance gauge.
LiMo3Se3 is a highly anisotropic, pseudo-one-dimensional conductor. The conductivity is based on infinite, one-dimensional chains of [Mo3Se3](-) triangles. It is also soluble in polar solvents. When it dissolves, the chains disperse to produce an Inorganic, polymer-like solution. We have used this solubility to produce free-standing, micrometer length conducting fibers with diameters as large as 60 nm and as small as the 0.6 nm diameter of a single chain.
Using the two-dimensional frictional force microscope, we observed two-dimensionally discrete friction on the NaF(100) surface with its lattice periodicity, which is explained by the two-dimensional stick-slip model quantitatively. By the raster scan, such a discrete friction composes the lattice periodicity image of the NaF surface. This means that one of the AFM contrast mechanisms of the lattice periodicity of the ionic crystal is the discrete friction. On the other hand, with a normal load of 14 nN the discrete friction disappeared, so the number of contact atoms is estimated as a few atoms. Thus, even under a near single-atom friction regime the atomic scale friction is explained by the two-dimensional stick-sip model quantitatively.
This paper describes a positioning system for ultraprecision machine tools. This system has a hydrostatic slide table driven by a hydrostatic lead screw (36 mm diameter, pitch 10 mm) and a laser interferometric positioning system. This means that the table is supported by hydrostatic bearings in any direction absolutely without any mechanical contact. Quite high accuracy of straight motion and positioning is realized and maintained; straightness and angular deviation are better than 50 nm and 2 mu rad respectively, positioning accuracy is about 10 nm within a 200 mm stroke. In addition, the effect obtained by a self-controlled restrictor for screw and by a short-pitch (pitch 6 mm) lead screw is mentioned. The stiffness of the hydrostatic screw in the feed direction can be statically infinite with this restrictor. By the use of this lead screw, positioning accuracy is improved and a 0.1 nm positioning step is possible.
For the purposes of grinding non-axisymmetric aspheric surfaces, a large five-axis controlled ultraprecision grinding machine has been developed for the first time in the world. As the machine is equipped with a profile measurement system, profile accuracy can be improved by feedback of profile error. A toroidal surface of 500 mm x 100 mm made of CVD-SiC has been machined. It has always been hard to grind CVD-SIC in the ductile mode with metal-bonded grinding wheels because of its hard and brittle properties, but grinding in the ductile mode has been accomplished by an improved truing and dressing method. As a result, surface roughness of less than 5.0 nm RMS was attained. By correcting profile error with the profile measurement system, profile accuracy of about 0.5 mu m was obtained after four corrections.
For interferometric measurements of step height standards with different roughness on either the upper or base levels, a shift of the reflection plane and hence a different height compared with a mechanical stylus measurement is assumed. This paper investigates an evaporated aluminium step height standard consisting of six steps with different roughnesses of the upper levels. Comparison measurements between a high-resolution stylus instrument, an interference microscope with two different objectives and two high-precision interferometers have been made. The agreement between the measurements of the two interference microscopes and the phase shifting interferometer was very good. The maximum deviation was less than 9 nm for steps up to 8 mu m height. Comparing the optically and the mechanically measured step heights, no systematic height differences were observable.