The measurement and characterization of micro-optical elements, systems and materials related to a variety of aspects like shape, displacement, deformation, strain / stresses, material constants, chemical composition etc. is a huge area which requires extended, multidisciplinary knowledge and a wide range of expensive instrumentation. The NEMO Centre for Measurement and Instrumentation includes twenty four institutes, whose joint expertise and capabilities do not only cover nearly all important measurement and characterization methods; they also comprise a big pool of commercial and internally developed instrumentation. In this paper we present the main tasks of this Centre. These tasks consist in e.g. providing benchmarks of the instrumentation and analysis software, providing an integrated pool of equipment for measurement and characterization of materials and components, identifying the gaps in measurement capabilities of existing equipment and providing guidance for development of new measurement methods and instrumentation, the development of proof-of principle demonstrators of novel measurement systems, and the development of numerical-experimental methods in collaboration with the virtual Centre for Modeling to support the development of novel micro-optics. In this paper we will also review the expertise of the NEMO partners in micro-optics measurement and characterization with the emphasis on aspects such as: microlens array testing, waveguide and fibre optics characterization, measurement of static and active microelements, determination of 3D refractive index distribution, subwavelength structure characterization, and shape monitoring and strain distribution testing in micro-optic packages. Also interesting issues connected with the definition of optimised measurement chains, data analysis and standardization of measurement procedures will be touched upon. Finally we will discuss the future structure of the Centre and its potential to provide virtual or on-site access to an electronic library which already contains a considerable amount of information on the large pool of instrumentation, jointly provided by the NEMO's partners.

The standardization of a microlens brings the merit for both users and suppliers, but which has issues to be solved in the characterization and evaluation due to the small lens size, caused from difficulty to apply the definition of conventional optical systems. Japan has the leading position of the international standardization activities in ISO for characterization of microlens since the early stage. Additionally, the recent development activities of wavefront aberration testing technologies are described.

Today different technologies exist that allow the fabrication of individual high-quality micro-optical refractive components and more in particular spherical microlenses. In this paper we will focus on the characterization of the latter components obtained with Deep Lithography with Protons (DLP). In the past we first fabricated the DLP microlenses and secondly a full geometrical and optical characterization was performed. However, this working method is very time consuming due to the amount of experiments needed for a complete calibration of our fabrication process. Therefore, we developed an interferometer for a real-time in situ sag characterization of the microlenses. In a first step we built a Mach-Zehnder interferometer working in the visible wavelength range and demonstrated its proof-of-principle for the determination of the microlens sag. In a next step we then transferred the concept of this interferometer to the closed reactor in which the in-diffusion of monomer vapour in the irradiated zones takes place. This novel approach will allow us to continuously monitor the volume expansion of the desired areas until spherical microlenses with a specific lens sag are obtained.

Direct waveguide writing by femtosecond lasers is rapidly becoming a promising valid alternative to standard fabrication techniques. Significant research efforts are devoted to understanding the effects of interaction of the radiation with the material and determining the key parameters in the writing process. The assessment of a reliable fabrication process depends crucially also on the availability of high resolution inspection and measurement methods. In this paper we employ digital holography (DH) in a microscope configuration as the characterization tool for measuring the refractive index profile of the waveguides. The method offers the advantages of high spatial resolution, high sensitivity and it allows to determine an absolute value of refractive index change without the need of any calibration. We report on the optical characterization of optical waveguides operating at 1.5 micron wavelength in two commercial glasses written by a stretched-cavity femtosecond Ti:sapphire oscillator. Measurements made by DH have evidenced a strong dependence of the fabrication process on the type of glass substrate.

In this paper we report on the measurement of the refractive index profile of optical fibers exposed to the gamma radiation. The tool we used for determining the refractive index distribution is microinterferometric tomography. Nuclear radiation is known to affect the guiding properties of optical fibers and it is therefore essential to characterize these effects to assess the applicability of fiber-optic technology for communication and sensing in space applications and in nuclear industry. We show that the fibres exhibit a slight refractive index increase which confirms results reported earlier.

We report on a new high accuracy home-made sample holder for near field characterization of millimeter long wave guiding structures (Y junction, Multi Modes Interference coupler). The principle of near field characterization is based on an atomic force microscopy tip that is brought to the surface of the sample (in the near field zone) in order to coupled out a small amount of the light confined inside the wave guiding structures. Due to the size of the components, scans as long as a few millimeters are required to get an entire optical mapping of the structure [1]. With the commonly available equipments such a scan is performed by acquiring step by step more than 100 images for a 2 mm scan. The overlapping of the different images is time consuming and unsatisfactory unless a numerical stitching procedure based on topographical details is used. Effective refractive indexes are typically determined with a precision of 10^-3 which could be further improved by increasing the millimeter scan resolution. The reason why successive images do not overlap is mainly due to the mechanical system supporting the sample. Actually, the nonlinearity of the actuator and the thermal expansion of the mechanical part prevent us to reach nanometric scale of repeatability on the positioning for micrometric range of displacements. In order to enable long range scans with nanometric repeatability and accuracy, we develop a specific mechanical system controlled by a heterodyne interferometric apparatus and a home-made high frequency electronic board [2]. The position of the sample is measured in real time with a resolution of 0.3 nm. The servo-loop allows to control the position of the sample with a repeatability of 1 nm (1σ) for a displacement of 1 mm. Furthermore our method is insensitive to the nonlinearity of the actuator.

Scanning near-field optical microscopy (SNOM) in reflection is employed for high-resolution mapping of surface refractive-index distributions. Two different single-mode optical fibers with step-index profiles are characterized using a reflection SNOM setup, in which cross-polarized detection is employed to increase the contrast in optical images and, thereby, the method sensitivity. The SNOM images exhibit a clear ring-shaped structure associated with the fiber stepindex profile, indicating that surface refractive-index variations being smaller than 10^-2 can be detected. It is found that the quantitative interpretation of SNOM images requires accurate characterization of a fiber tip used, because the detected optical signal is a result of interference between the optical fields reflected by the sample surface and by the fiber tip itself. The possibilities and limitations of this experimental technique are discussed.

Within the framework of the General Support Technology Program of the European Space Agency (ESA), a compact dedicated confocal laser scanning microscope has been developed for 3D fluorescence imaging of biological samples. The microscope permits normal confocal mode operation with excitation at 488nm and fluorescence lifetime imaging (FLIM) with excitation at 630nm and a time resolution of 200ps. Each fluorescence signal is detected by a dedicated photomultipier tube. Proper optical signal separation and filtering is performed by a set of optical filters and dichroics. The software and hardware further include the specific imaging modes allowing for fluorescence recovery after photobleaching (FRAP) and fluorescence loss in photobleaching (FLIP). In addition to this dual wavelength fluorescence imaging mode, the microscope includes transmission imaging capabilities via differential imaging contrast (DIC). Both fluorescence and DIC imaging can be acquired simultaneously. The source for the DIC is a near infrared LED. This choice permits the decoupling of DIC and fluorescence signals by a dichroic cold mirror. The opto-mechanical assembly is constructed on the two sides of a rigid 16mm thick aluminum base plate of dimensions 389 mm by 575.5 mm. The total volume under light and dust shielding removable covers is just 53 dm³, excluding PC and control electronics. In this paper, the design, performance and limitations of this compact confocal microscope are discussed. Illustrative examples of applications on biological samples are shown.

High resolution optical microscopy is still an important instrument for dimensional characterisation of micro- und nanostructures. For precise measurements of dimensional quantities a highly accurate modelling of the optical imaging on the basis of rigorous diffraction calculation is essential, which accounts for both the polarisation effects and the 2D or 3D geometry of the structures. Some applications like for example the measurements of linewidths on photomasks demands for measurement uncertainties of few nm or less. For these requirements the numerical and the model induced uncertainty, respectively, may be limiting factors even for sophisticated modelling software. At PTB we use two different rigorous grating diffraction models for modelling of the intensity distribution in the image plane, the rigorous coupled wave analysis (RCWA) method and the finite elements (FEM) method. In order to evaluate the performance of both methods we performed comparative calculations on the basis of a test suite of binary chrome on glass gratings with different line widths reaching from 100nm to 10μm, and with different line/space ratios between 0.01 and 100. We present results of this comparison for TE, TM and unpolarised Koehler illumination of the grating. Residual deviations between both methods and the resulting measurement uncertainty and related to the corresponding time consumptions are considered.

A procedure and software have been developed to transform the area distribution of the residual surface heights available from the measurement with the Micromap interferometric microscope into a two-dimensional (2D) power spectral density (PSD) distribution of the surface height. The procedure incorporates correction of one of the spectral distortions of the PSD measurement. The distortion appears as a shape difference between the tangential and sagittal PSD spectra deduced from the 2D PSD distribution for an isotropic surface. A detailed investigation of the origin of the anisotropy was performed, and a mathematical model was developed and used to correct the distortion. The correction employs a modulation transfer function (MTF) of the detector deduced analytically based on an experimentally confirmed assumption about the origin of the anisotropy due to the asymmetry of the read-out process of the instrument's CCD camera. The correction function has only one free parameter, the effective width of the gate-shaped apparatus function which is the same for both directions. The value of the parameter, equal to 1.35 pixels, was found while measuring the 2D PSD distribution of the instrument self-noise, independent of spatial frequency. The effectiveness of the developed procedure is demonstrated with a number of PSD measurements with different X-ray optics including mirrors and a grating.

Fringe pattern demodulation by the Fourier transform method associated with fringe extrapolation by the Gerchberg algorithm was investigated in details for its application to fast profiling of surfaces with small patterns and/or cuttings. For such surfaces, spectral leakage is a major concern as it corrupts data in a large part of the areas of interest if extrapolation is not carried out. Simulated fringe patterns or real interferograms recorded on micromechanical devices by interference microscopy were used for this evaluation. Different filter shapes in the Fourier space were tested for the determination of valid areas in the interferogram, for the fringe extrapolation stage and for the final phase demodulation. It is demonstrated that filters with a shape adapted to the modulation sidelobe in the Fourier space allow automated measurements without user expertise while maintaining a high accuracy. Some practical rules for the choice of extrapolation margins, fringe density, fringe extrapolation iteration number and other parameters are clarified to reach accurate and fast automated measurements.

During the last five years scatterometry measurement using ellispometry and reflectometry has met a great interest in nano and microelectronics fab. Today, this technology of measurement is used to control lot production and has become mature for 1D-grating measurements. Nevertheless, some aspects of this method of measurement are always under research studies. This paper focuses on one of these aspects: the evaluation of the influence of the "real-life 1D-structure" (linewidth variations along the lines and line to line, roughness, defect inside the grating) on spectroscopic signatures and on scatterometry measurement methods. The measurements have been carried out on KLA-TENCOR ellispometer and on Nanometrics reflectometer in order to compare the two methods of measurement. The simulations have been done with MMFE (Modal Method of Fourier Expansion) software developed by LETI labs. To control defect characteristics and defect distributions, one wafer was printed using electron beam lithography. The aim is the evaluation of the impact of defects in the grating on the spectroscopic signatures and its influence on extracted geometrical parameters by fitting the experimental curves. Different deviations to real-life structures have been studied. First we focus on the influence of typical defects of lithography processes such as bridging and partial destruction of lines and on the influence of CD distribution values inside the grating. Then, we study the influence and the possibilities of measuring Line Edge Roughness (LER). For LER measurements different targets have been also exposed on e-beam tool. Simulations and experimental measurements have been carried out. All the results obtained have been compared with imaging standard tool: top down SEM measurement.

The idea of simultaneously using spectroscopic ellipsometry and interferometric microscopy for optical properties determination of thin absorbing layer is presented. Spectroscopic ellipsometry is a powerful method for thicknesses and refraction index determination, but in the case of thin absorbing layer a verification method is very useful. Such verification could be done with interferometric method. With interferometry a phase shift can be observed between two adjacent areas with different heights and/or optical properties. This phase shift can be accurately calculated with same model which is used for ellipsometry data analysis. Proposed method is most effective for metal-insulator-semiconductor structure with thin semitransparent metal layer. For this structure phase shift and sensitivity to metal thickness and optical indexes can be adjusted by preparing substrate with suitable dielectric layer thickness. Method is illustrated with spectroscopic ellipsometry and interferometry measurements on Al-SiO₂-Si structure.

A simple polarimeter based upon a quarter wave plate and a polarizer is presented. Rotating the quarter wave plate all Stokes vector components are calculated by processing the intensity variation. The effect of misalignment in the linear polarizer azimuth and the deviation from the π/2 retardance of the quarter wave plate is analytical and experimental investigated. The technique is consolidated by using two algorithms for Stokes vector calculation.

We report the investigation of a Fizeau interferometer-based OCT system. A secondary processing interferometer is necessary in this configuration, to compensate the optical path difference formed in the Fizeau interferometer between the end of the fibre and the sample. The Fizeau configuration has the advantage of 'downlead insensitivity', which eliminates polarisation fading. An optical circulator is used in our system to route light efficiently from the source to the sample, and backscattered light from the sample and the fibre end through to the Mach-Zehnder processing interferometer. The choice of a Mach-Zehnder processing interferometer, from which both antiphase outputs are available, facilitates the incorporation of balanced detection, which often results in a large improvement in the Signal-to-Noise ratio (SNR) compared with the use of a single detector. Balanced detection comprises subtraction of the two antiphase interferometer outputs, implying that the signal amplitude is doubled and the noise is well reduced. It has been discerned that the SNR drops when the refractive index variation at a boundary is small. Several OCT images of samples (resin, resin + crystals, fibre composite) are presented.

To eliminate mechanical scanning in the probe head of an endoscopic OCT system, we propose the use of an imaging fibre bundle for probe beam delivery. Each fibre in the bundle addresses a Fizeau interferometer formed between the bundle end and the sample, allowing acquisition of information across a plane with a single measurement. Depth scanning components are now contained within a processing interferometer external to a completely passive endoscope probe. The technique has been evaluated in our laboratory for non-biological samples, including glass/air and mirrored/air interfaces. Images resulting from these experiments are presented. The potential of the system is assessed, with reference to SNR performance and acquisition speed.

In this paper the optical parameters of Low Coherence Speckle Interferometry (LCSI) are analysed in order to optimise the technique and increase the probing depth. The contrast of the interference signal depends on the configuration of the optical setup and the optical properties of the material. Theoretical investigations and measurements for optimising the beam ratio, the coherence function and imaging parameters are presented. By optimising the interference signal the probing depth of the technique can be extended. This is demonstrated for a semi-transparent polymer material. Finally, measurements of the deformation of interfaces in a multi-layered material are presented.

The relationship between statistical structure parameters of rough surface and associated correlation parameters of scattered field is used to develop a method for rough surface diagnostics. The treatment is based on the model of random phase object with inhomogeneity phase variation less than unit. The proposed diagnostic method is applicable to arbitrarily shaped surfaces. The sensitivity limit of the method in measuring the standard deviation of surface profile from base line is about 0.001 μm.

In this paper, we will describe a technique that combines a common path scanning optical interferometer with artificial neural networks (ANN), to perform track width measurements that are significantly beyond the capability of conventional optical systems. Artificial neural networks have been used for many different applications. In the present case, ANNs are trained using profiles of known samples obtained from the scanning interferometer. They are then applied to tracks that have not previously been exposed to the networks. This paper will discuss the impacts of various ANN configurations, and the processing of the input signal on the training of the network. The profiles of the samples, which are used as the inputs to the ANNs, are obtained with a common path scanning optical interferometer. It provides extremely repeatable measurements, with very high signal to noise ratio, both are essential for the working of the ANNs. The characteristics of the system will be described. A number of samples with line widths ranging from 60nm-3μm have been measured to test the system. The system can measure line widths down to 60nm with a standard deviation of 3nm using optical wavelength of 633nm and a system numerical aperture of 0.3. These results will be presented in detail along with a discussion of the potential of this technique.

Possibilities of using recently-developed femtosecond pulse lasers for advanced precision length metrology are investigated. Special emphasis is placed on the use of femtosecond lasers particularly for absolute distance measurements with sub-micrometer accuracy over extensive ranges. This investigation reveals that femtosecond lasers are capable of providing a suitable means of nanometrology by implementing dispersive comb interferometry in combination with synthetic wavelength interferometry and heterodyne interferometry.

We present one of the applications of the Optical Vortex Interferometer (OVI). OVI is based on the regular net of optical vortices which are generated by the interference of three plane waves. Disturbing one of the interfering waves causes a change in the position of the vortex points in the vortex net. The measurement is based on tracking the vortex position change. This method can be used to determine small-angle rotation. OVI distinguishes two axis of rotation and the corresponding two rotation angles can be measured with sub-second resolution. The linear vibrations of the measured element are automatically subtracted. The single measurement provides hundreds of measurements points, so the statistical methods for data analysis and corrections can be effectively applied. In the paper we present the experimental testing of the method. To get the precise rotation of one of the interfering wave's the optical wedge is put into one of the interferometer's arm. The analysis shows that the amplitude`s decrease does not influence the measurement accuracy. From the vortex net shifting the rotation angle of one of the interfering waves is calculated and this rotation is also used to calculate the refracting angle of the applied optical wedge.

Micro-refractive lenses are an important example of components and subsystems that are being used increasingly in optical sensors, communications, data storage, and other diverse applications. These lenses have a continuous relief surface such that details of their dimensional shape, refractive index, and homogeneity all influence performance. Measurement capabilities for micro-refractives fall short of current and future needs and are complicated by the need to fabricate non-spherical refractive surfaces. To control the fabrication process, the target measurement uncertainties are approximately 3 parts in 10^-4 for radius and on the nanometer scale for figure measurements. Carrying out metrology at this level is very challenging and especially so for micro-scale components. Micro-interferometry is the most promising tool and can be used to measure radius of curvature, focal length, dimensional surface errors, and transmitted wavefront. Common practice is to calibrate with a single high quality artifact for measurements of a range of different radii, and we see that this is only approximately valid. Figure measurement calibration, for example, will be improved if the radius of the calibration artifact closely matches the radii of the test lenses, but acquiring such a range of artifacts is not practical. We have demonstrated the application of a self-calibration procedure for figure measurement and transmitted wavefront measurement calibration, called the random ball test. Radius measurements on the micro-scale are also challenging. Our research focuses on measurement advances for refractive components and new data analysis strategies to optimize the impact of measurement results.

Different measurement methods exist for the extraction of the coefficient of thermal expansion (CTE). Among them the observation of the sample length as a function of its temperature is the direct way. In the last decade, the use of phase shifting interferometry in combination with computer-based analysis of interference phase maps drastically improved interferometrical length measurements. In addition to the observation of the length itself, such measurements allow the extraction of a length topography L(x,y) of the sample as shown in this paper. From the behaviour of the length topography at different temperatures an upper limit of CTE-homogeneity can be obtained. It is shown in which way disturbing influences can be removed so that uncertainties of L(x,y) in the sub-nm range can be reached for different shaped samples

In the last years lithium niobate (LN) has become one of the most important optical material in optoelectronics and nonlinear optics for its large electro-optics and nonlinear optical coefficients. Ferroelectric materials are employed in several electrooptic, acousto-optic, and nonlinear optical devices, as modulator of light, beam deflector, optical frequency converters, or tuneable sources of coherent light for spectroscopic applications. Manipulation of ferroelectric domains into gratings, matrices, or other shapes is possible. Fabrication of new ordered microstructures in LN samples through domain engineering followed by differential etching has been developed recently for applications in the fields of optics and optoelectronics. These microstructures have a range of applications in optical ridge waveguides, alignment structures, V-grooves, micro-tips and micro-cantilever beams and precise control of the surface quality and topography is required of for photonic band-gap structures. Moreover engineering ferroelectric domains by an electrical poling technique represent a key process for the construction of a wide range of photonic devices. Therefore, a thorough understanding of material properties and of the poling process are crucial issues. We will show that interferometric approach based on Digital Holography can provide a very useful tool for investigation and characterization of materials and of the engineered structures.

A novel sensor for use in displacement metrology is proposed. The proposed displacement sensor is based on the grating imaging, which conventionally uses two amplitude gratings with rectangular apertures of fifty percent width of the period. In the conventional way, signal to noise ratio of displacement output is one of issues to be overcome for precise measurement because about seventy five percent of the illumination light is trapped by two amplitude gratings. On the other hand, in the proposed sensor a cylindrical lens array and a phase grating are applied as the first and the second grating, respectively. Therefore, the illumination light is trapped neither by the first grating nor the second grating except absorption. In our experiments, the cylindrical lens array with 200 μm period and the reflective sine phase grating with 100 μm period are used. Experimental results demonstrate that higher position resolution and higher accuracy of the displacement measurement is feasible by our proposal.

Refraction index evaluation by means of a quasi-normal incidence technique is presented. The theoretical procedure involves Fresnel equations as well as a complete statistical algorithm developed for experimental values treatment. The characteristics of the experimental technique are analyzed in depth and rules for high precision measurements are given. Refraction indices of soda lime and BK7 substrates were evaluated as function of wavelength. Accuracies of the order of 10^-3 in refraction indices determination were obtained. Finally, and making use of high precision polishing techniques, the authors are adapting this method for the reproduction of step and graded index profile functions and diffusion depths of integrated optical waveguides.

A regular net of optical vortices generated by three plane waves interference allows for a new kind of interferometer - Optical Vortex Interferometer. The precision of that kind of interferometer depends on a localization accuracy and phase reconstruction. One problem is the unique phase reconstruction. Interference of three waves can generate two identical interferograms with opposite topological charge of vortices, so information from three waves interferogram is not enough to the unique phase reconstruction. First method of topological charge determination requires one interferogram of two waves analyzed in an experiment and knowledge about a direction of laser beams. Second method is based on analysing interferogram with the fourth wave added.

We describe the optimum design of the near-field scanning optical microscope (NSOM) based on a short probe tapping mode tuning-fork (TMTF) configuration and its applications in optoelectronic characterization and optical measurements. The short probe TMTF-NSOM is constructed to operate both in collection and excitation modes, in which a cleaved short fiber probe attached to one tine of the tuning fork is used as the light collector/emitter as well as the force sensing element. Interference fringes due to standing evanescent waves generated by total internal reflection are imaged by collection mode. On the other hand, excitation mode of short probe TMTF-NSOM is applied to perform near-field surface photovoltage measurements on AlGaInP light emitting diode structures.

A data acquisition technique and relevant program for suppression of one of the systematic effects, namely the 'ghost' effect, of a second generation long trace profiler (LTP) is described. The 'ghost' effect arises when there is an unavoidable cross-contamination of the LTP sample and reference signals into one another, leading to a systematic perturbation in the recorded interference patterns and, therefore, a systematic variation of the measured slope trace. Perturbations of about 1-2 μrad have been observed with a cylindrically shaped X-ray mirror. Even stronger 'ghost' effects show up in an LTP measurement with a mirror having a toroidal surface figure. The developed technique employs separate measurement of the 'ghost'-effect-related interference patterns in the sample and the reference arms and then subtraction of the 'ghost' patterns from the sample and the reference interference patterns. The procedure preserves the advantage of simultaneously measuring the sample and reference signals. The effectiveness of the technique is illustrated with LTP metrology of a variety of X-ray mirrors.

The rigorous coupled wave analysis (RCWA) implemented as the Airy-like internal reflection series (AIRS) is applied in a theoretical analysis of the optical response of diffraction gratings. Detailed theoretical description of the RCWA with respect to the AIRS implementation is provided, including the application of Li's Fourier factorization rules and the recursive algorithm for sliced relief gratings. Numerical analysis of convergence properties including computation time is demonstrated for structures made of transparent, semiconductor, or metallic materials.

For precision traceable measurements with interference microscopes it is necessary to know the aperture correction factor. The source of the correction is the finite diameter of the illumination light source. If there is no further knowledge about it, the usual way is the calibration of the instrument with a reference standard. In this contribution a practical way is described to determine the aperture correction factor of an interference microscope by direct measuring the diameter of the illumination aperture. A diffraction pattern of a line scale is directed into the microscope objective. In the back focal plane the set of diffraction orders acts a scale with known dimension. The scale of the diffraction orders can be observed through the eye piece tube together with the image of the aperture. Under the assumption of an accepted model the correction factor can be determined. Some considerations for the practical use and results are presented.

Dynamic speckle or biospeckle is observed in biological samples illuminated by laser light. The properties and applications of this phenomenon have been treated in the literature. In this paper, we present a method of dynamic speckle analysis based on the filtering in frequency bands of the temporary history of each pixel. Butterworth filters are applied to the temporary evolution and different images are constructed showing the energy in each frequency band. Applications on vegetable specimens examples are shown.

With decreasing critical dimensions (CD) on lithography masks, increasing demands on CD metrology techniques come along. Already today the results of the three standard methods for CD measurements currently used, atomic force microscopy (AFM), scanning electron microscopy (SEM), and optical microscopy, typically do not yield the same results. This is because of, e.g. incomplete knowledge of the material parameters, insufficient modelling accuracy or- especially in the case of optical microscopy-insufficient resolution. With decreasing CDs these systematic differences increase. The need on new cross-calibration strategies arises. Non-imaging metrology methods like scatterometry as non destructive, non diffraction limited, fast optical methods offer access to the geometrical parameters of periodic structures like e.g. top and bottom CD, pitch, side-wall angle, line height, or roughness. Therefore, these methods provide independently achieved additional information that can be used for cross-calibration. At the PTB two scatterometers are in use (VIS @ λ=633 nm and EUV @ λ=13.55 nm). A third device (DUV @ λ=193 nm) is under construction. It will offer a wide spectrum of measuring principles like scatterometry, ellipsometric scatterometry, reflectometry, polarisation reflectometry, and using a broadband light source spectroscopic ellipsometry and spectroscopic reflectometry. For simulation and modelling of the intensity distribution of the diffraction pattern two programs based on the rigorous coupled-wave analysis (RCWA) method and a finite element method (FEM), respectively are used. The program features will be illustrated. A comparison of the simulations with the results achieved with the scatterometers on different types of lithography masks (chrome on glass, EUV-masks) will be presented as well.

Grating pseudo images are formed in cascade grating systems and are useful in different fields and in many different applications, such as interferometry, optical encoding of position, etc. There are several processes for creating images of a grating by using only gratings as imaging elements, being the most known the Talbot effect. In other configurations one grating acts as an imaging element for another grating, such as the Lau effect where two identical gratings are illuminated with an extended monochromatic light source, and a pseudoimage of the first one is formed at infinity. In Generalized Grating Imaging, pseudoimages form, without the need for lenses, at finite distances of the gratings. One disadvantage that a grating pseudoimaging system presents for most applications is the fact that the contrast of self- and pseudo-images strongly depends on the distance between gratings. This makes the optical devices less tolerant to positioning and/or manufacturing. It has been shown that, for Talbot effect, the use of polychromatic light can eliminate the dependence of the contrast on the grating position. A similar result has been numerically demonstrated for some pseudoimages in a double grating system with spatial and temporal partially coherent light. In this work we present an analytical model for Ronchi gratings that justifies and explains the numerical results previously obtained.

A digital holographic microscope (DHM) is employed as non-invasive metrological tool for inspection and characterization of a micromechanical shunt switches in coplanar waveguide configuration (CPW) for microwave applications. The switch is based on a bridge that can be actuated by using electrodes positioned laterally with respect to the central conductor of the CPW. The DHM features, such as speed, contact-less and non-destructivity, have allowed a full characterization of an electrical actuated shunt switches. In particular, the out-of-plane deformation of the bridge due to the applied voltage has been investigated with high accuracy. DHM inspection allows to investigate the shape of the bridge during the actuation, the total warpage due to the actuation, possible residual gap, possible hysteresis, and so on. These characterizations have been carried out both in static and in dynamic condition. In full paper the complete characterization will be reported together with an accurate description of the optical system employed for the investigation.

Scaling down the critical dimension (CD) also requires a reduction of systematic and stochastic errors in the measurement of phase shifts introduced by phase shift masks (PSM). Furthermore, in current mask design all degrees of freedom are used to satisfy the requirements and consequently different types like attenuated or alternate PSM are in use. A direct measurement of the introduced phase shift - the wave field propagating from the mask - is mandatory to further reduce the CD error budget. Although different interferometer set-ups might be used, a common path design enables a high repeatability of the measured phase values by being less sensitive to vibrations and air turbulence. Furthermore systematic errors due to different optical aberrations present in different arms of an interferometer are eliminated. A lateral shearing interferometer was realized by adding two phase gratings and one amplitude grating to a microscope working in transmission. Thus a very stable tool for the measurement of phase shifting structures at λ = 193nm was obtained. The repeatability of the phase measurement is less than 2π/1300. A simple evaluation can be used if the lateral restricted phase shifting region is doubled, i.e. the phase ridge is compared with a neutral surrounding area. This situation of a totally sheared object is obtained if the amount of the shear s is larger than the object's lateral distance d. Moreover, this method provides a larger number of measured phase values compared to the case of differential phase contrast. To reduce the contribution of the visibility to the uncertainty budget the spatial degree of coherence was tailored to obtain a visibility V > 0,8 even at large shear distances. The 5-phase algorithm was used to introduce the reference phase steps Δφ_i. A model based description of the measurement was made. The uncertainty of the measured phase value of the introduced interferometer is |Δφmax| = 2π/330 if an optimal alignment of the system is achieved. This value can be reduced by a factor of 2 if the dependence of the phase shifting algorithm on the measured phase value is taken into account. A further reduction can be obtained if an intensity monitoring is implemented. Measurements performed at different adjustment states are given to discuss e.g. the influence of the adjustment of the two phase gratings on the uncertainty budget. The estimation of the phase values and the dependence of this values on the layout of the object, respectively the surrounding area will be discussed. Although the results obtained - due to the prototype status - are not yet completely satisfactory, the described analysis proves the potential of the newly developed interferometer for further reduction of uncertainty and possible use as a reference phase measurement instrumentation.

The most common technique for diameter measurement of thin metallic cylinders is optical diffractometry. It consists in illuminating the cylinder with a collimated monochromatic light beam, determining the diameter from the location of the minima of the far field diffraction pattern. Babinet principle is normally assumed, being the diffraction pattern of the cylinder equivalent to that of a strip whose width is equal to the cylinder diameter. Due to the three dimensional nature of the cylinder, this model is not valid for accurate measurements. It has been experimentally shown that, when compared to interferometry, Fraunhoffer model presents a systematic overestimation in the cylinder diameter. Rigorous models which assume that the wire presents an infinite conductance have been developed. However, the refraction index of the material has also appeared important for an accurate estimation since it produces a phase shift of the reflected wave by the wire surface, modifying the state of polarization of the incident light beam and, as a consequence, the location of the diffraction minima. In this work we propose a model based on the Geometrical Theory of Diffraction that assumes both the three-dimensional nature and a finite conductance of the wire. Results for several materials are presented, showing that the overestimation of the wire diameter depends on the state of polarization and wavelength of the incident light beam, as well as the diameter and refraction index of the metallic wire.

Several microscope techniques namely confocal microscopy have a huge number of applications in metrology reported in literature. Although they are essentially imaging techniques metrology and particularly profilometry is a very attractive field of application owing to their ability to obtain depth discrimination. In recent years there has been a growth of three dimensional microscopy methods namely those based on structured incoherent illumination. In spite of different implementation approaches they are based on the fact that the optical transfer function (OTF) of the imaging system attenuates with defocus for every spatial frequency with the exception of zero-frequency. In this way it is able to get depth information through the projection of a grid structure so it is suitable for application in profilometry. The purpose of this work is to develop and test a profilometer based on a low-cost bench microscope. The optical layout is an epi-illuminated configuration of scanning-stage type for reflection microscopy. Tests with structured incoherent illumination in this case line-illumination are being carried out in order to achieve depth discrimination using a linear image sensor of CMOS type as detector. This paper presents a description of the optical arrangement as well as the acquisition and control system. Preliminary results are shown that were obtained using a plane mirror to measure its axial resolution and a micromachined component to test the application of this bench microscope in profilometry.