This article presents off-axis modulation transfer function (MTF) measurements in the image field covered by the azimuthal angle θ = ±20 deg and a defocus range of Δz ∈ [−40 μm, +40 μm] with the slit based MTF measurement setup at PTB. The experimentally determined standard deviation of a set of N = 10 MTF repetition measurements under a certain set of repeatability conditions is employed to estimate the standard uncertainty uexp with respect to the repeatability of measurements in the setup. The importance of measurement repeatability and measurement reproducibility in the assessment of measurement accuracy and measurement uncertainty for interlaboratory comparisons is discussed.
We present an interferometric form measurement system that is capable of measuring the form of optical surfaces from flats to moderately curved freeforms. A Fizeau interferometer is scanned over the specimen measuring its topography and the distance to the surface in subaperture measurements. The angle between the interferometer and the specimen is adjusted for each position and additionally relevant stage angle errors are measured with two tiltmeters. An optical surface with a spherical form (50 mm diameter, 10 m radius of curvature) is measured, and an uncertainty budget yields an uncertainty of 69 nm ( k = 2 ) for its topography. The determined radius of curvature agrees well with the nominal specifications and a measurement with a coordinate measurement machine. Furthermore, a form measurement of a car’s side window is presented.
Despite the strong demand for high-precision Modulation Transfer Function (MTF) measurements for industrial and scientific applications, currently there is a lack of MTF calibration services traceable to national standards. Trioptics GmbH, in cooperation with the National Metrology Institute in Germany (PTB), is establishing an accredited laboratory for traceable MTF measurements of optical lenses aiming at an expanded uncertainty of 𝑈MTF = 0.01 for coverage factor 𝑘 = 2. To reach this goal, influences of misalignment and mechanical stage positioning errors need to be considered. We present an experimental approach to investigate the influence of the detector’s positioning accuracy on the MTF measurement uncertainty depending on the imaging properties of individual lenses under test. For samples with different optical properties (effective focal length and aperture), the MTF is measured at various positions in a 2D grid spanned over image field and focusing direction. At each position, the derivatives of MTF in field and focusing direction are calculated. Knowing the overall mechanical uncertainty of the detector, this data can be used to determine local uncertainties of the measured MTF. In contrast to the classical way of defining measurement uncertainty only by benchmarking the accuracy of the measurement equipment, we now include MTF-sensitivity of the sample under test. Hereby, a more precise and application-oriented analysis of the MTF accuracy for a measurement system can be achieved.
This contribution presents simulation and measurement results considering the identification and preclusion of aliasing effects in Modulation Transfer Function (MTF) measurements of unknown specimens. An adaptation of the commonly applied slanted edge algorithm is employed to retrieve the MTF from tilted Line Spread Function (LSF) images, which are recorded for different microscope objective magnifications and tilt angles. Measurement results demonstrate the capability of the slanted slit oversampling technique to detect and resolve aliasing issues. The variation of magnification objectives, tilt angle, binning value, camera detector and evaluation algorithm reveals systematic MTF error contributions due to these parameters.
Flatness metrology is needed in the optical industry as well as for synchrotron mirrors, for precision engineering and for telescope mirrors. Calibrated optical flats are a key element for quality assurance in many fields and applications of science, research, and industry. The National Metrology Institute of Germany (Physikalisch-Technische Bundesanstalt, PTB) offers interferometric flatness calibrations for specimens up to 300 mm in diameter and deflectometric flatness calibrations up to 900 mm for elongated specimens. We are continuously developing our flatness metrology further to meet future requirements, e. g. to achieve uncertainties of a few nanometers as well as to offer measurements of large flats up to 1.5 meters. In this contribution we focus on three current activities. Firstly, we show our new large form measuring system for smooth curved specimen, which can also be used for flatness measurements of specimens with diameters up to 1.5 meters. Secondly, we demonstrate, how sub-millimeter lateral resolutions can be achieved with small angle deflectometry. Thirdly, we present an international round robin comparison of an optical flat. The flatness on an aperture of 300 mm is to be measured and the participants shall reach an expanded uncertainty of less than 15 nm (k=2). This will also ensure a reliable traceability also for subaperture measurements of optical flats with larger diameters.
The Physikalisch-Technische Bundesanstalt is currently developing a form measurement system for optical surfaces with diameters up to 1.5 metres, and local radii of curvature larger than 10 m. Moreover, the surfaces can have a radial symmetrical form, or a freeform shape. This contribution presents an optomechanical simulation of the measurement system to give hints about the achievable accuracy. The simulated system uses a Fizeau interferometer with an aperture of 100 mm for subaperture stitching interferometry. An advanced subaperture stitching method (called angle- and distance-assisted subaperture stitching, ADASS) is used to reconstruct the absolute topography of a virtual specimen. The topographies of the single subaperture measurements are positioned in space by using the angular position of the interferometer and the distance to the specimen. Thereafter the remaining height difference of the neighbouring subtopographies in their overlap region is corrected. The resulting surface form is fitted to the Zernike polynomials. In the simulation, a parabolic surface with a radius of curvature of 70 m was used as the virtual specimen. The reconstructed absolute surface form has a deviation of 418 nm (RMS) from the input surface over a diameter of 1.5 m.
Large optics with diameters of up to 1.5 m are being used more and more in industry and science. Flatness measurements of these optics are needed with uncertainties down to a few ten nanometres. For slightly curved specimens with radii of curvature down to 10 m uncertainties in the sub-micrometre range are required. We are currently building a new form measurement system which aims to fulfil these requirements. It will be set up in 2020 and the first measurements will be carried out in 2021. The setup can be operated with different sensor heads which use deflectometric- or interferometricbased methods. We plan, amongst other things, to use Fizeau interferometers with aperture sizes of 10 mm, 100 mm and 150 mm. The mechanical and optical setup of this new system is presented and simulation results of conventional subaperture stitching methods for this system with an aperture of 100 mm are shown. We also discuss the different measurement methods for the absolute form measurement of these optics.
The Physikalisch-Technische Bundesanstalt has developed a reference setup for measuring the modulation transfer function (MTF) of camera lenses with the goal of reaching an expanded measurement uncertainty of 0.01 (coverage probability of 95%) for various measurement configurations. We present optomechanical simulations of the setup behavior, which are used to investigate the influence of the combined mechanical misalignments on the MTF depending on the lens under test. The investigations are carried out as part of Monte Carlo studies for different sample lenses and field heights considering the correlations between the effects of different positioning errors on the MTF measured. The results of the sensitivity analyses have allowed appropriate alignment strategies to be identified that significantly reduce the uncertainty contribution of the positioning errors. By implementing these strategies, the target measurement uncertainty can be achieved for most of the desired measurement configurations. In addition, the comparison of three different sample lenses shows that the MTF sensitivity to misalignments strongly depends on the characteristics of the lens to be measured.
High-precision topography measurements of nearly flat specimens with nanometre uncertainty are still demanding. They are needed for synchrotron or other high-quality optical mirrors. At the Physikalisch-Technische Bundesanstalt (PTB), we operate an interferometric setup and a small-angle deflectometric setup for flatness measurement. We present both systems, show the measurement uncertainties and discuss the pros and cons of interferometric and gradient based measurement systems. To achieve the nanometre level in flatness metrology, the environmental conditions and the holder of the specimen are very important. We show measurements of topography changes due to delayed elasticity effects. We demonstrate current developments with the small-angle deflectometer, especially to improve the lateral resolution.
Due to the complexity of measurement systems for asphere and freeform surfaces, well-known artifacts are required to characterize the accuracy of the results of their form measurements. We present advancements in manufacturing and characterization of metrological multispherical freeform artifacts. The strong cooperation between the manufacturing and measurement units of Physikalisch–Technische Bundesanstalt (PTB) allows the form of the artifacts to be manufactured very accurately and the root-mean-square deviation from the desired design form to be improved by a factor of >20 compared to former results (e.g., from a value >500 nm to about 20 nm). Furthermore, a nickel/phosphorous coating is added to the copper surface, resulting in greater hardness and allowing the coated surface to be used as a reference for low-force tactile measurement systems. Four Gaussian peak fiducial marker structures are added to the design to improve the evaluation of comparison measurements. In addition to characterizing the radii in the spherical segments using PTB’s radius measurement bench, we also characterize the sphericity of the spherical segments using a Fizeau interferometer. We show form measurement results for a full-field measuring tilted-wave interferometer and compare the form measurement results in the spherical segments with measurement results obtained with the Fizeau interferometer.
Optical systems have increased in quality and capability and are today widely used in many fields of applications. An
important step forward was the introduction of aspheres and freeform surfaces. For manufacturing these surfaces in
high quality, the accurate measurement of them is highly important. A reliable measurement requires traceability.
The concept of traceability is presented, uncertainty sources are itemized and the steps towards traceability for an
asphere interferometer are discussed.
The measurement of optical flats, e. g. synchrotron or XFEL mirrors, with single nanometer topography uncertainty is still challenging. At PTB, we apply for this task small-angle deflectometry in which the angle between the direction of the beam sent to the surface and the beam detected is small. Conventional deflectometric systems measure the surface angle with autocollimators whose light beam also represents the straightness reference. An advanced flatness metrology system was recently implemented at PTB that separates the straightness reference task from the angle detection task. We call it ‘Exact Autocollimation Deflectometric Scanning’ because the specimen is slightly tilted in such a way that at every scanning position the specimen is ‘exactly’ perpendicular to the reference light beam directed by a pentaprism to the surface under test. The tilt angle of the surface is then measured with an additional autocollimator. The advantage of the EADS method is that the two tasks (straightness reference and measurement of surface slope) are separated and each of these can be optimized independently. The idea presented in this paper is to replace this additional autocollimator by one or more electro-mechanical tiltmeters, which are typically faster and have a higher resolution than highly accurate commercially available autocollimators. We investigate the point stability and the linearity of a highly accurate electronic tiltmeter. The pros and cons of using tiltmeters in flatness metrology are discussed.
The Physikalisch-Technische Bundesanstalt (PTB) has developed and set up a deflectometer with sub-millimetre lateral resolution. This device uses an enhanced deflectometric method called ‘Exact Autocollimation Deflectometric Scanning’ (EADS), which makes use of two angle sensors. One angle sensor has a small beam aperture and scans the specimen by using a pentaprism or a double mirror unit. It operates as a null angle sensor and controls the tilting of the specimen by means of a piezo actuator, so that the reflected beam of the specimen always propagates back in the same direction. The null-angle sensor allows sub-millimetre apertures with sensitivities of better than 0.01 arcsec to be achieved. The tilt of the specimen is measured with the second angle sensor – usually a commercially available autocollimator – with a large aperture at a fixed distance. The surface topography is obtained by the numerical integration of the measured tilt angles. The measurement uncertainty associated with the resulting topography typically scales with the scan length. Here, we present measurements with different lateral resolutions down to 0.3 mm.
There is an increasing demand for accurate form measurements of large optics with dimensions of up to one meter. For example, reference flats of Fizeau interferometers have to be known in the nanometer range. Synchrotrons or XFEL mirrors are often more curved, but nevertheless have to be manufactured and measured at a comparable or even better level of accuracy. Different approaches like deflectometry or interferometry in combination with stitching methods are typically used for these purposes, but obtaining the low uncertainty levels needed is still a considerable challenge. In this paper, measuring concepts and systems used at PTB for these purposes will be presented. For flatness measurements, a so-called Deflectometric Flatness Reference (DFR) system and a Fizeau interferometer were used. Slightly curved surfaces can be measured with the Traceable Multi Sensor (TMS) method, and CMMs with point or line sensors were also available. We will also present the current measurement capabilities and some measurement examples of form measurements of optical surfaces at PTB. The different setups and their pros and cons will be discussed. Future developments in the field of large-optics measurement will also be shown.
Precisely known artifacts are required to characterize the accuracy of asphere and freeform measuring instruments. To this end the best knowledge of the surface characteristics in conjunction with a low measurement uncertainty are necessary. Because this is a challenging task for typical freeform surfaces used in optical systems, the concept of “metrological” artifacts is introduced. We have developed a multispherical freeform artifact for performance tests of tactile touch probe and contact-free optical measuring systems. The measurement accuracy of the complete form and the deviation from calibrated spherical sections can thus be determined. The radius calibration of multiple spherical sections is performed with an extended radius measuring procedure by interferometry. Evaluated surface forms of different measuring methods and the radii determined can be compared to each other. In this study, a multispherical freeform specimen made of copper, with two differing radii, has been measured by two optical measuring methods, a full field measuring tilted-wave interferometer and a high accuracy cylinder coordinate measuring machine with an optical probe. The surface form measurements are evaluated and compared, and the radii determined are compared to the results of a radius measurement bench.
The concept of traceability is presented for the interferometric form measurement of optical surfaces. The calibration chain for interferometric flatness measurement is evaluated in detail, showing that only a few influence quantities are significant. For spherical surfaces, the complexity increases as the measurement separates into sphericity and radius measurement. Traceability in asphere metrology is much more complex, and some aspects are discussed in terms of the example of the Tilted-Wave Interferometer concept.
The tilted wave interferometer is a non-null test interferometer for the measurement of aspheres and freeform surfaces
without dedicated null-optics that uses an array of tilted waves to locally compensate the deviation of the surface from
the spherical form. The concept allows for short measurement times of only a few minutes and high lateral resolutions at
the same time. The calculation of the surface error is performed by perturbation of a polynomial representation of the
surface. Since we are also interested in higher frequency errors of the surface which cannot be described by a polynomial
of finite order these errors are evaluated in an additional step. Since every wavefront only covers a small area of the
surface the challenge here is to reconstruct the surface from the information that is distributed over the different patches.
We will present the method that was developed for the reconstruction of these high frequency errors as well as
measurement results of aspheres and freeform surfaces without rotational symmetry that were obtained by this method.
The tilted-wave interferometer (TWI) was recently developed by the University of Stuttgart for the high-accuracy measurement of aspheres and freeform surfaces. The system works in a non-null measurement fashion and si multaneously uses several test beams with different tilts. Reconstruction of the specimen under test from TWI measurements is challenging and in order to correctly separate the real surface topography from systematic aberrations, the employed interferometer needs to be characterized. This characterization, as well as the recon struction of the specimen from TWI measurements, requires sophisticated data analysis procedures including ray tracing and the solution of an inverse problem.
A simulation environment was developed at the Physikalisch-Technische Bundesanstalt (PTB) in order to inves tigate the accuracy and stability of TWI systems, and to explore possibilities and limitations of this promising measurement technique. Virtual experiments were carried out to quantify the sensitivity of the results with respect to the assumed linearity in the reconstruction procedure, positioning errors, and measurement noise. Our first results suggest that the mathematical TWI reconstruction technique basically allows highly accurate measurements with uncertainties down to a few nanometers, provided that calibration errors of the optical sys tems are kept small. The stability of the results and their accuracy can, however, depend significantly on the particular surface of the specimen and on the choice of experimental settings.
Interferometry is often used to measure the form of optical surfaces. While interferometry is generally expected to give
high accuracy results, a variety of error influences exist which have to be considered. Some typical error influences
which are often underestimated will be discussed in this paper. In flatness metrology, the main error influences are
imperfections of the reference surfaces, specimen support or cavity influences. For non-flat surfaces like aspheres or free
form surfaces, in particular the influence of errors from the determination of the lateral coordinates becomes very
important. Sub-aperture interferometry copes with stitching errors, which can be reduced by Traceable Multi Sensor subaperture
methods where the influence of the imaging system of the interferometer may dominate the error budget. This
can be similar for other types of interferometers.
At the Physikalisch-Technische Bundesanstalt (PTB) two new scanning deflectometric systems (Deflectometric Flatness
Reference: DFR) were installed for measuring topographies of nearly flat surfaces with sub-nanometre uncertainty. The
two systems can measure the form of horizontally and vertically orientated specimens with a diameter of up to 700 mm.
The systems are based on different deflectometric procedures, the so-called direct and the difference deflectometry. With
the flatness measuring systems, an uncertainty down to 0.1 nm (depending on the specimen and its peak-to-valley) is
being aimed at for the 95% coverage interval. Virtual experiments show that the optical and mechanical components
must be aligned in the arcsecond range in order to achieve errors for the topography in the sub-nanometre range. In this
paper we describe the setup of the new DFR system for horizontally orientated specimens in detail and show methods
and experimental results for an accurate alignment of the optical and mechanical components. We present an accurate
alignment strategy for ultraprecise deflectometric measurements and show a measurement of a section at a horizontally
orientated specimen.
In dimensional nano- and micrometrology, single sensors are often combined into an array of sensors to enable faster
measurements by utilizing parallel data acquisition. If combined with appropriate scanning techniques, the use of sensor
arrays additionally facilitates the estimation and correction of systematic sensor errors and, thus, enables more accurate
measurements. To exploit these options, the arrays have to be aligned carefully with respect to the scanning direction,
and, in addition, the lateral distances between the sensors have to be determined with sufficient accuracy.
This presentation describes a method to align an optical distance sensor array parallel to the direction of a linear translation
stage, which is used to scan the specimen under test, and it describes a method to evaluate and determine the sensor
distances with high accuracy.
Alignment is a multi step procedure: The first step is to orientate a step edge profile perpendicular to the scanning direction
of the sensor using an M-array and an auxiliary CCD camera. In a second step, the line sensor array is scanned
across the edge using different rotation angles of the sensor. The positions where the different sensors cross the edge are
evaluated to obtain the sensor orientation relative to the scanning direction, the distances between the sensors, and their
transversal displacements.
We will show experimental data obtained with an optical line sensor array of three single sensors. The measurements
will be compared to simulated data carried out with a virtual experiment programmed at PTB. Relevant error sources are
assessed and the limitation of the method is discussed.
The Nanometer Comparator is the PTB reference length measuring machine for high precision calibrations of line scales
and encoder systems. Up to now the Nanometer Comparator allows to measure the position of line structures in one
dimension only. For high precision characterisations of masks, scales and incremental encoders, the measurement of the
straightness of graduations is a requirement from emerging lithography techniques. Therefore the Nanometer
Comparator will be equipped with an additional short range measurement system in the Y-direction, realized as a single
path plane mirror interferometer and supposed to achieve sub-nm uncertainties.
To compensate the topography of the Y-mirror, the Traceable Multi Sensor (TMS) method will be implemented to
achieve a reference-free straightness measurement. Virtual experiments are used to estimate the lower accuracy limit and
to determine the sensitive parameters. The virtual experiments contain the influence of the positioning devices,
interferometer errors as well as non-perfect adjustment and fabrication of the machine geometry. The whole dynamic
measurement process of the Nanometer Comparator including its influence on the TMS analysis, e.g. non-equally spaced
measurement points, is simulated.
We will present the results of these virtual experiments as well as the most relevant error sources for straightness
measurement, incorporating the low uncertainties of the existing and planned measurement systems.
To minimize the measurement uncertainty of one dimensional length measurements on line scales, linear encoders and
interferometers the PTB in cooperation with the Dr. Johannes Heidenhain GmbH had built up a new length comparator.
The Nanometer Comparator [1,2] has already proven its performance during the measurements of incremental encoders
and line scales with an expanded measurement uncertainty of below 5 nm [3,4,5]. Due to the introduction of double and
multiple exposure in advanced lithography techniques the overlay and registration metrology requirements will
drastically increase so that reference metrology tools need to be developed further to be able to follow the resulting
decrease of the specifications. Therefore, the PTB further develops the new 1D vacuum comparator to add a
measurement possibility for straightness and to reach a measurement accuracy in the sub nanometer range [6]. One key
development will be the interferometric measurement of all six degrees of freedom of the measurement slide of the
comparator. A new multi axis heterodyne interferometer electronics and optical interferometer designs minimizing
nonlinearities by spatially separated beams are under development.
The so-called Nanometer Comparator is the PTB vacuum length comparator which has been developed for high precision
length metrology on measurement objects with micro- and nanostructured graduations, like e.g. line scales, incremental
encoders or photomasks. The Nanometer Comparator allows to achieve smallest measurement uncertainties in the
nm-range by use of vacuum laser interferometry for the displacement measurement. We will report on the achieved
measurement performance of this high precision vacuum length comparator and the already started developments to substantially
enhance its measurement capabilities by additional straightness measurement capabilities. The enhanced
Nanometer Comparator will provide traceability for photomask pattern placement measurements in industry, also facing
the challenges due to the increased requirements on registration metrology as set by the introduction of new lithography
techniques like double patterning methods.
The Traceable Multi Sensor (TMS) system is a scanning system for the measurement of the topography of large
optical surfaces. The system uses a compact interferometer with an aperture of some millimetres to realize
multiple distance sensors and an autocollimator for the angle measurement. In contrast to common stitching
techniques, the systematic sensor errors are calculated in addition to the entire topography by the TMS algorithm.
Additionally, piston and tilt at each position of the interferometer are determined by the algorithm. An essential
requirement for the algorithm is the exact lateral positioning of the sensor at given locations.
The goal of this paper is to investigate the influence of a class of error sources on the resulting topography
estimation using computer simulations. The errors of this class result in inexact measurement positions of the
distance sensors. Especially the lateral positioning errors of the scanning stage lead to increasing errors for short
wavelengths. For topography wavelengths below 3mm with an amplitude of 100nm the resulting topography
error increases to 3nm and more. For longer wavelengths the positioning errors are no longer the dominant
error source and the root mean square error of the resulting topography is approximately 1 nm for positioning
errors with a standard deviation of 5 &mgr;m. The pixel distance error and distortion of the interferometer strongly
influence the topography measurement of specimens with large deviations from a plane. The simulations show
that for a topography with a peak to valley of 50 &mgr;m the root mean square error of the reconstructed topography
is below 10 nm.
Furthermore, a possibility to compensate the lateral positioning error of the scanning stage is presented which
makes the TMS method nearly independent of positioning errors of the scanning stage. As a consequence, it
is possible to use systems of non equidistant distance sensors whose lateral distances are independent of the
positioning interval.
A new scanning system is presented for the high-accuracy form measurement of flat and slightly curved surface profiles. The system employs a small measuring head with multiple distance sensors that is scanned along the surface. Additionally, an autocollimator is utilized to account for angular scanning stage errors. For this setup, suitable design of experiment allows scanning stage errors as well as systematic distance sensor offset errors to be eliminated. As a consequence, high-accuracy form measurements of flat and slightly curved surfaces with high resolution can be achieved. A demonstrator setup has been realized and tested. The results confirm the high potential of this method.
Form measurements with scanning systems are well established. Many scanning systems are limited to almost flat surface forms. In parallel, there is a need for systems capable of measuring stronger curved specimens with a radius of curvature down to one meter or even below. The size of specimen is sometimes up to one meter and the form has to be known with nanometer uncertainty. Moreover, the desired lateral resolution is sometimes smaller than one millimeter. The realization of a form measuring system combining high lateral resolution, a large measurement range and low measurement uncertainty requires sophisticated measurement principles. A measurement principle suitable for this task is presented which uses a combination of multiple distance sensors and an angular sensor, and exemplary measurements for stronger curved surfaces are shown. As multiple distance sensor an interferometer is used. Exemplary measurements are shown and the limitations of the interferometer as multiple distance sensor are discussed.
The requirements for form or topography measurement become ever-challenging. Desired lateral resolutions are sometimes smaller than one millimeter, whereas the size of the specimen can exceed one meter. To meet these requirements, scanning measurement systems are increasingly applied with sensors being (much) smaller than the specimen. However, due to the presence of quadratic sensor and scanning stage errors, large errors of the reconstructed topographies can emerge. To overcome these problems, a novel scanning measurement system is proposed. The system includes a linear scanning stage, a compact interferometer used for multiple distance measurements and an autocollimator for additional angular measurements. The effect of quadratic sensor and first-order scanning stage errors which the system allows to eliminate is discussed. Topography reconstruction is outlined and first measurements with a demonstrator set-up are described which show that the proposed novel measurement principle works well under real measurement conditions.
A scanning system consisting of an angle sensor and coupled distance sensors is used for high-accuracy profile form measurements of optical surfaces. This system allows both scanning stage errors and systematic errors of the distance sensors to be estimated. Additionally, a high lateral resolution is achieved. The optical surface profile is reconstructed from the measurements made by the angle and distance sensors by application of least-squares. A test set-up is presented and the results of optical form measurements are shown.
For the production of aspheres and free-form surfaces, high-accuracy and flexible measurement techniques are necessary. The Large Area Curvature Scanning (LACS) method can be used to measure the surface form of arbitrary smooth surfaces with high accuracy. The curvature values of the surface elements along a scan line are captured by a curvature sensor and from these values the form is calculated. The curvature sensor typically is a small interferometer with an aperture of some millimeter. From a model fit to the measured interferogram, the local surface patch and the local curvature value are extracted.
Determining the curvature values from the captured interferograms with high accuracy is a challenging task and requires some kind of intelligent procedures for defect recognition and for the choice of matched surface models. Several aspects of these problems are discussed. Examples of measured surfaces are shown. Special emphasis is laid on the measuring speed of the LACS system, which is mainly determined by the speed of the curvature evaluation procedure, as this is important for the use as an in-situ measurement system integrated into production systems.
Surfaces as needed in optical systems, ranging from the visible even into the EUV region, become larger and often have a length or diameter of 500 mm and more. The form of these surfaces, describing the surface spatial frequency content with components below 1 mm-1, has to be characterized on the nanometer and sometimes even on the sub-nanometer scale. The extendibility of the measuring systems accuracy to large specimen dimensions basically depends on the method of the measurement and the scaling of different systematic uncertainty components with lateral coordinate values. This is analyzed for flatness and sphericity measuring systems, with a focus on the systems for Extended Shear Angle Difference (ESAD) and Large Area Curvature Scanning (LACS) used at PTB. Both are scanning methods working absolute and with a good natured scalability to large dimensions. For the measurement of optical flats the dominant uncertainty of topography is in the quadratic or spherical contribution of the surface in terms of a polynomial description. For calibration flats, as used for large interferometers, this often cannot be measured absolutely with sufficient accuracy. The potential of ESAD and other methods is analyzed with respect to this uncertainty component. Uncertainty considerations and measurement results for large flats are presented. For the form measuremetn of largely extended convex or concave surfaces, where classical interferometric set-ups are not possible due to the lack of a master surface or the extrme costs incurred for large optical components, the potential of LACS is presented.
The Large Area Curvature Scanning (LACS) method for measuring all types of surfaces including aspheres is based on scanning the surface at discrete lateral coordinate positions, but the amount of information captured during measurement is much greater. The Extended LACS (ELACS) method will be presented that uses this information to evaluate also the mid-spatial frequency components of the surface. An examples of measurement for a surface showing a relatively high mid-spatial frequency content is given and ELACS is compared to other methods. These are various methods subsumed under the term “stitching,” especially stitching interferometry. Different mathematical concepts for these methods are briefly discussed and a novel approach is presented and analyzed with special emphasis being laid on the uncertainty which can be achieved with stitching methods.
KEYWORDS: Sensors, Calibration, Interferometers, Time metrology, Prisms, Autocollimation, Telescopes, Environmental sensing, Distance measurement, Chemical elements
In the past few years two form or topography measuring methods have been developed at PTB that are based on measuring the slope (angle) difference or the curvature, which both are intrinsic properties of the surface. In a laboratory environment the methods offer a very low uncertainty of measurement down to, or in part even below, the nanometer range. One of them is the Extended Shear Angle Difference method (ESAD), in which the angle difference signal of two surface points separated by a rather large shear is the basic measurement signal obtained with an autocollimation telescope. The other is Large Area Curvature Scanning (LACS) where the curvature values of the surface elements along a scan line are captured by a small interferometer.
Error and uncertainty considerations are given for these methods which form the basis for determining the uncertainty which can be achieved under well optimized environmental conditions as necessary for tracing back and calibrating other measuring systems. Additionally, the uncertainty that can be reached in an industrial environment is discussed, and solutions are described to use these promising techniques also in such cases.
The measurement of the topography or the nanotopography of large wafers up to 450 mm in diameter with satisfactory lateral resolution and nanometer uncertainty is still an unsolved problem. The topography of wafers covers a relatively large measurement range as wafers have surfaces with a so-called "slightly unflat" topography which mostly exceeds the measurement capabilities of interferometers. For the ultraprecise and traceable measurement of the slope and topography of slightly unflat optical surfaces, a novel scanning deflectometry principle has been developed. An uncertainty of the topography in the nanometer range will be achieved, as this principle minimizes error influences and allows a highly precise calibration of the angle measuring device. The main goal is to use this principle for the ultraprecise measurement of the nanotopography of large wafers.
The measurement principle is based on the analysis of differences of reflection angles obtained at surface points which are separated by large lateral shears. It does not rely on external reference surfaces of matched topography and in first and second order is independent of any stage errors and the whole-body motion of the specimen. The measurands are directly traced back to the base units of angle and length. The specific idea of wafer measurement is to combine rotational and linear scanning with the measurement of slope difference vectors and to arrive at an unambiguous solution for the topography and nanotopography. The equations with the slope difference vectors are solved to reconstruct the slope vectors, as newly developed mathematical algorithms allow the surface slope to be reconstructed from slope differences for two different shears. This is reached by applying natural extensions and shearing transfer functions by a mathematically exact method over the whole surface area. Further the differential equations for the slope vectors are solved to unambiguously reconstruct the topography. With this method it is possible to achieve nanometer uncertainty and at the same time a high lateral resolution, short measurement times and the possibility of mastering the large measurement range necessary for slightly unflat wafer surfaces.
The paper is structured as follows: After some general statements, we will discuss types of aspheres, then typical applications for aspheres, general aspects of interferometric methods for measuring aspheres, special methods for weak aspheres, permutation methods for mathematically simple aspheres, compensation optics, traceable methods, and at the end other methods.
The use of interferometers is usually divided into two areas, single-dimensional length measurement where the distance along an optical axis is measured and interferometric surface or wavefront measurement where the laterally varying phase distribution of a largely extended wavefront is the property of interest. In this presentation a combination of the two types of use will be presented. One of the applications which requires simultaneous measurement of distance, slope and curvature is high-accuracy form measurement of aspheres and free form surfaces by the Large Area Curvature Scanning (LACS) method. For a high-accuracy LACS set-up it is most important to measure (and control) the absolute distance between the interferometer and the surface under test, as well as the slope of the wavefront. In the present set-up, a commercial general purpose interferometer is used. The simultaneous measurement of distance, slope, curvature and shape with this instrument is presented. Special emphasis is laid on the specific fringe evaluation method. It uses a model of the real interferogram image and the superimposed disturbing effects. A suitable global optimization method allows the desired quantities to be determined rapidly and reliably. For the future use of LACS, a new, specifically designed multi-purpose interferometer is introduced which can be used in LACS systems with nanometer form measurement accuracy.
Steep aspheres and general complex optical surfaces are of great importance for many optical technologies and necessary for many high-technology applications where nanometer and sub-nanometer accuracy is indispensable. An example of such an application is lithography in the DUV, VUV or EUV region of the electromagnetic spectrum. The problem of measuring the topography with this high accuracy has not been solved in general. Recently, a particular measurement principle has been developed and investigated. It is referred to as high- resolution large-area curvature scanning, and the topography is determined by mathematical calculations on the basis of the information available about the curvature. It focuses on the principles of the traceability and the avoidance of error influences and is intended for determining the figure of steep aspheres and complex surfaces with ultra-precision. An uncertainty budget will be presented for the method and the facility. Special emphasis will be put on the different principles of intrinsic two-dimensional methods, in contrast to scanning methods, external references and their influences, errors of scanning stages and their influences, whole-body movements of artifact and their influences, the properties of the measurement signal, lateral and vertical resolution of the detector, long-term stability of the facility, etc.
It is a promising method for measuring steep aspheres and complex surfaces with nanometer and sub-nanometer accuracy to measure the curvature and to calculate the topography from it, since unlike slope and distance, the curvature is an intrinsic property of a surface and less insensitive to error influences. For the development of a measuring instrument based on the physical property of curvature, various topics have been investigated. The method described does not rely on external form references, and the errors of the scanning stages and the whole-body movement of the artifact have only little influence on the accuracy. In comparison to other measuring techniques, it is an advantageous feature of the curvature measuring technique that distance and angle between sensor and surface element can be controlled and kept constant during scanning as it is the curvature and not the distance or the slope which is the measurand. This leads to the result that, apart from the calibration of the curvature sensor, the whole system no longer suffers from first-and second-order errors. The uncertainty budget shows that nanometer accuracy is achievable.
Very recently, in the context of measuring aspheres and complex surfaces with ultra-precision, a particular measurement principle was developed which determines the form (topography) of extended test samples by scanning measurements of curvature, being the reciprocal of the radius of curvature. The curvature sensor must be traceably calibrated with a low uncertainty. This back tracing can be done, first, by measuring radius of full spheres with a highly accurate sphere interferometer, second, by measuring roundness with highly accurate methods, and third, by measuring specially designed calibration aspheres. These procedures for traceably calibrating the curvature sensor will be described.
The problem of measuring the slope and figure of large, slightly non-flat or steep complex surfaces with nanometer and sub-nanometer accuracy has not been generally solved. Existing systems such as interferometers, three-coordinate measuring machines and slope measuring facilities do not fulfil the requirements for very high accuracy: related topics will be discussed: intrinsically two-dimensional methods vs. scanning methods, external references and their influences, the kind of measurements signal, errors of scanning stages and their influences, whole-body movements of the artifact and their influences, long-time stability of the facility. Very recently, particular measurement principles have been proposed, which focus on the principles of traceability and avoidance of error influences and are intended for determining slope and figure with ultra- precision. These methods are based on the direct measurement of slope difference and curvature, and in contrast to other methods, are directly traced back to the base units. They do not depend on external references and offer the advantage that the errors of the stages and whole- body movement of the artifact do not influence the accuracy of measurement. Application of these methods eliminates the influences of the first- and second-order errors of the facility. In addition, absolute measurements are traced back to relative measurements and long-time stability is achieved by constant recalibration of the sensor. The question which remains to be answered and which will be discussed is that of the basic or natural limits of systems intended for measuring slope and figure.
The problem of measuring steep aspheres with high accuracy has not yet been generally solved. In this presentation, a particular sensor type is presented which can be used to measure the form of optically smooth surfaces with no restrictions as regards lateral extent or complexity. It is a curvature sensor which can be guided along the surface. Curvature is an intrinsic property of the artifacts, which is independent of its position and angular orientation.
An apparatus for measuring the topographies of complex surfaces with high accuracy by curvature scanning has been setup. For this purpose, a new type of curvature sensor that processes information from a relatively large area of the surface under test is moved along the surface. The principles advantages and a technical realization of this method referred to as large-area curvature scanning will be presented.
The ultra-precise manufacturing of surfaces needs ultra- precise characterization. A novel optical scanning technique for testing flats, aspheres and complex surfaces is presented which offers ultra-precision for the measurement of slope and topography. The scanning technique needs no external references; it is traced back exclusively to the units of angle and length. The technique is based on a combination of two principles, namely to perform difference measurements for slopes with large shears and to use only a single pentagon prism for the difference measurements, keeping the angular position of the pentagon prism constant in space. The combination of these two principles eliminates the influences of all first- and second-order errors of the facility. Whole- body movement of the artefact under test and distortion of the facility itself do not result in errors.
For the testing of flats, spheres or aspheres, the optical scanning technique has become a competitor to interferometric techniques. The set-up implemented at PTB for the ultra- precise optical scanning of optical flats will be presented and the underlying principles explained, with a special view to the avoidance of error influences in the optical set-up and to the elimination of the remaining errors. The basic principle of the scanning set-up is that an autocollimation telescope combined with a movable pentagon prism is used to register the surface slope angles. To avoid higher-order errors originating in the pentagon prism, its angular position is kept constant in space by active stabilization. Another advantageous principle is that difference measurements for slopes are performed. This has become possible thanks to a new error-free difference method recently developed by PTB. The combination of these principles eliminates, theoretically, the influences of all possible errors of the scanning facility. An experiment using a high-quality flat ((lambda) /60) 140 mm in diameter showed a reproducibility of 0,2 nm over one week.
A novel optical scanning technique for testing flats, aspheres and complex surfaces is presented, which offers ultra-precision for the measurement of slope and topography. The scanning technique is based on a combination of two principles, namely to perform difference measurements for slopes with large shears and to use only a single pentagon prism for the difference measurements, keeping the angular position of the pentagon prism constant in space. The combination of these two principles eliminates the influences of all first and second-order errors of the scanning facility, as this scanning technique is traced back exclusively to the units of angle and length. Especially whole-body movement of the artifact under test and distortion of the facility itself do not result in errors. The evaluation of difference measurements with large shears has become possible very recently with high accuracy and high lateral resolution because an error-free algorithm was developed. It is further a special feature of this technique that it measures differences of slopes and, consequently, allows the measurement of aspheres and complex surfaces with large asphericities.
KEYWORDS: Lamps, Cameras, Glasses, Modulation transfer functions, Tungsten, Monte Carlo methods, Optical testing, Camera shutters, Photography, Digital photography
An illuminated table intended for illuminating transmission test charts used for the optical testing of different image capture devices has been developed. It provides a diffusely luminous table face, 300 mm in diameter, in a compact set- up. The spectral distribution of the emitted light is similar to Planck radiation of about 3050 K as required, for example, for testing, digital photographic cameras. The luminance level achievable is about 15000 cd/m2, the standard deviation of which is 1.8 percent within a circle of 300 mm in diameter. The illuminated table consists of an integrating hemisphere, with the large opening covered by an opal glass serving as the illuminated table face. Inside, a set of seven tungsten lamps is posited at locations such that the light output at the table face is as uniform as possible. Residual non-uniformities caused, for example, by slightly differing luminous intensities of the lamps are adjusted by controlling the currents of the seven lamps individually by means of a PC-controlled automatically running optimization process. This closed control loop uses a calibrated digital camera to capture the current spatial light distribution at the table face. The image is analyzed by an algorithm integrated in to the control loop.
The quantitative determination of wave aberrations for high- quality imaging systems for off-axis use is an important challenge to modern optical measurement techniques. Several proposals in the conjunction with Twyman-Green interferometry and shearing interferometry will be made for this purpose. Two of these methods allow a reproducible and definitely quantitative alignment of the Twyman-Green set-up to be achieve. One method uses the point image for the positioning and the other that the front face of the Twyman sphere can be the first surface of a so-called adjustment system and that the Twyman sphere is in its correct position when the adjustment system reports that correct adjustment has been achieved. Another proposal concerns a shearing interferometer that consists of a combination of two conventional shearing interferometers and is capable of capturing the complete information about a wavefront by recording a single interferogram. Moreover a new method for quantitatively determining the shape of a wavefront from this interferogram will be presented, which requires nearly no a priori information about the shape of a wavefront to be calculated, leads to accurate results with high lateral resolution for relatively large shears, reconstructs the information inside the whole aperture, and requires much less computational effort than procedures hitherto known.
A shearing interferometer unit is proposed, which is capable of gathering the complete information about a wavefront by recording a single interferogram and which consists of a combination of two conventional shearing interferometers. With its capabilities it is predestined to be used as a tool for accurate measurements because it can be built as a solid unit without moving parts. This aim will be supported by a new method for quantitatively determining the shape of a wavefront from the difference between the wavefront and a laterally sheared copy which allows experiments to be performed with shears up to 20 percent of the width of the aperture. The method is based on filtering of the wavefront difference in the frequency domain with a transfer function specifically determined for this purpose. A novel window function is used to deal with the influence of the edge of the pupil. This method hardly requires any a priori information about the shape of the wavefront to be calculated. Furthermore, it leads to accurate results with high lateral resolution, reconstructs the information inside the whole aperture, and requires a much lesser computational effort than procedures hitherto known.
Electronic still picture camera systems, mostly using CCD sensors, are ever more widely used both in the photographic world and in professional measuring technique. It is therefore necessary to develop adequate measuring facilities for the characterization and quality control of high-end measuring or imaging systems. In this presentation, a versatile measuring set-up for electronic still picture cameras is described. Several standards are being developed for the characterization of electronic still picture cameras that, among other things, define the methods for measuring the spatial frequency response (SFR) using a test chart and the modulation transfer function (MTF). Exemplary results are shown for the measurement of SFR and MTF of black-and- white and color CCD still cameras and the two properties are compared to one another.
The optical transfer function (OTF) of calibration lenses must be measured with a high accuracy, requiring complex setups. Thermal expansion effects of the measuring system are a limiting influence, resulting in changes of the imaging state (e.g., defocusing) during the period of measurement. A new concept is presented to overcome these disturbing effects. By taking a rapid tomography sequence of image planes, the three-dimensional irradiance distribution in the image space is captured and stored in a computer, analogous to a flash photo of a moving object in conventional photography. After this, there is time to find the desired imaging state in the data. This "flash" concept has been successfully implemented in a typical OTF measuring setup, improving the reproducibility by one order of magnitude.
KEYWORDS: Modulation transfer functions, Optical transfer functions, Sensors, Temperature metrology, Lenses, Optical testing, Spatial frequencies, Modulation, Data storage, Calibration
In measurements which are influenced by external parameters, computer control and automation can help to improve the accuracy. This is also true of basically optical measurement set-ups such as certain interferometers for measuring flatness or wavefront aberrations, or set-ups for measuring the optical transfer function of lenses, for example. For the latter, a critical external parameter is the temperature variation resulting in expansion and therefore defocusing effects. New developments and measurement principles leading to improved accuracy are discussed.
KEYWORDS: Modulation transfer functions, Optical transfer functions, Spatial frequencies, Modulation, Optical testing, Sensors, Signal detection, Data processing, Phase transfer function, Optical benches
The uncertainty of optical transfer function (OTF) measurements of lens systems -- as one example of high-precision optical measurement technique -- can be greatly reduced by computer-aided measurement control and analysis. Improvements resulting in a high repeatability are described and the problems of some analysis concepts are discussed. With such a high repeatability (10-3) and resolution possible, the influences of temperature and the resulting thermal expansion dynamics of the mechanical set-up and the measuring system itself become factors limiting the reproducibility. Realizations of temperature controlled or thermally insensitive designs are presented.
The measurement of wavefront aberrations for optical imaging systems plays a major role in the field of modern optics. Measurements with a Fizeau type interferometer and a shearing type interferometer are described. The interferometric set-up for the shearing interferometer and the algorithms, in particular for the evaluation of the wavefront aberration, are described in more detail. Various methods for the measurement of wavefront aberrations are compared.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.