WO2016015734A1

WO2016015734A1 - A scatterometer apparatus

Info

Publication number: WO2016015734A1
Application number: PCT/DK2015/050228
Authority: WO
Inventors: Morten Hannibal MADSEN; Poul-Erik Hansen
Original assignee: Dfm A/S
Priority date: 2014-08-01
Filing date: 2015-07-27
Publication date: 2016-02-04

Abstract

The invention comprises a scatterometer apparatus suitable for determining dimensional information of a sample. The scatterometer apparatus comprises a light source operable for emitting illuminating light; a low NA objective system; an image recorder comprising a 2D array of detectors, and an image analyzer operatively connected to the image recorder for analyzing recorded images. The light source is arranged for illuminating a sample by an illuminating spot size with a dimension larger than μm. T low NA objective system has an effective numerical aperture (NA) of less than 0.3 and is arranged for collecting scattered light from the sample upon illuminating the sample by the light source and the image recorder is arranged for recording at least a portion of the collected scattered light comprising at least one wavelength in form of at least one image.

Description

A SCATTEROMETER APPARATUS

TECHNICAL FIELD

The invention relates an optical metrology system suitable for determining dimensional information and preferably for characterization of topography of micro/nano-structures on a surface or embedded in a semi-transparent material. The invention comprises a scatterometer apparatus which is operable to have a large imaging area, preferably in millimeter scale and, simultaneously, a high resolution of structural features, such as in nanometer scale. The scatterometer apparatus is advantageously configured for providing high resolution information for multiple segments within an imaging area e.g. in micrometer scale.

BACKGROUND ART

In recent years, there has been an interest in fast and reliable measurements of micro/nano-textured surfaces. This includes measurements of structural features on surfaces and/or embedded in materials. The interest is mainly driven by the semi-conductor industry for quality control of products, accuracy in production steps and of production equipment. Furthermore, for

increasingly smaller features sizes and higher component count in the semiconductor industry, there is a tendency to demand even higher resolution and faster operation of the measurement systems. Other industries and technology areas also show needs for such measurement systems, including anti-reflective coatings, hydrophobic surfaces, and holography.

There exist many measurement techniques for structural measurements, for example optical microscopy, scanning electron microscopy (SEM), atomic force microscopy (AFM), and imaging ellipsometers. However, these techniques all have one or more disadvantages;

Conventional optical microscopes are fast, but are struggling to resolve feature sizes with lateral dimensions less than 1 μιη. Advanced optical microscopes, such as interference and confocal microscopes, can improve resolution, but are still limited in their lateral resolution due to the wavelength of the light and are not useful in accurate characterization of nano- textures/nano-structures.

Scanning electron microscopes give high resolution (nanometer scale).

Furthermore, they are able to provide a large field-of-view (micron to millimeter scale). However, scanning electron microscopes are typically cumbersome to work with and impractical in many settings. Also, they require low ambient pressure around the measured sample, whereby the dimensions of the sample are limited by the evacuation chamber.

Atomic force microscopy gives high resolution (nanometer scale), but are hindered by a small field-of-view, i.e. only on a relatively small sample or portion of a samples is measured. Typically, field-of-view is less than 100x100 μιη². Furthermore, atomic force microscopes are not able to resolve

embedded structures. Furthermore, measurements with atomic force microscopy are laborious and time consuming. Imaging ellipsometer systems can have a large field-of-view (up to several hundred square microns). However, they have a limited resolution of around 1 μιη. The resolution is limited by the detector (typically a CCD camera). Further disadvantages of such systems are noise, drift and cost due to high optical component complexity. These instruments are often based on the principles of null-ellipsometry, where the polarization of the incident light is controlled, such that the reflected light is purely linear polarized. It is a disadvantage that this requires several expensive optical components for controlling and analyzing the polarization of the light. Also, the angle of the incoming light and the detector can be varied, typically from around 35 degrees to grazing incidence, thus requiring special built equipment. Furthermore, the angle of detection tends to bring part of the image out of focus. It is a further disadvantage that the system manufacturers have to correct for this by dynamic focusing and/or advanced optics components - adding a further cost to such systems. For another type of imaging ellipsometer, the ellipsometer is built into a

conventional microscope. For these, a high numerical aperture objective, 0.70 or higher is required to provide the illuminating light at a large angle of incidence (typically 60 to 80 degrees) and to collect light from many angles. However, only a small fraction of light is collected within the angular ranges and the configuration for this collection of light is complicated, in terms of component count and cost.

In recent years, scatterometry has become a popular technology for accurate measuring of structural dimensions on a surface or embedded in a semi- transparent layer, as well as of film thicknesses. The reason for this popularity is that scatterometry overcomes a number of the disadvantages from the other instruments mentioned above. Scatterometry provides high resolution (nanometer scale) similar to SEM and AFM, but at lower cost and more practical handling of samples. However, as prior art scatterometry systems have no information about the scattering from different segments of a sample and are therefore not able to quantify multi-segmented structures. By multi- segmented structure is meant a sample comprising segments of different structures. Furthermore, prior art scatterometry is not able to provide the imaging of a large part (or all) of a sample and simultaneously measure structural dimensions of a smaller segment of the sample.

Prior art scattero meters are only able to perform averaged measurements of a specific area. Thus, prior art scattero meters require that an operator selects an area-of-interest prior to performing dimensional measurements. Typically, the area-of-interest is identical to the spot size on the sample.

Prior art scattero meters are also challenged by locating a small area-of- interest on a large sample. This is typically done either imaging a large uniform area, such as for thin film layer characterization (where accurate positioning is not required) or by another externally imaging system. With the external imaging system an operator first needs to find the overall area, then selecting an area-of-interest within this, and thereafter adjusting the scatterometry system for performing dimensional measurements.

Misalignment and drift are detrimental to such measurements.

For other scatterometers an area-of-interests smaller than the field of view can be obtained by introducing an aperture in front of a spectrometer. As an example, a high magnification lens, for example 100x objective, combined with a 100 μιη aperture can resolve an area-of-interest down to for example 1 μιη. This again requires the operator to select the area-of-interest prior to performing the measurement. Also, these scatterometers are very sensitive to drift of the sample and misalignment of the detectors. Therefore, such scatterometers are often impractical for nano/micro-textured surfaces, and are therefore mainly used for characterization of thin film layers.

DISCLOSURE OF INVENTION

An object of the present invention is to provide an alternative or an improved metrology apparatus that facilitates an effective way of measuring samples structural features with nanoscale resolution.

In an embodiment of the invention it is further an object of the present invention is to provide a metrology apparatus and a method suitable for determining dimensional information of a sample with a large freedom with respect to sample dimensions and segmentation of the samples. These and other objects have been solved by the invention as defined in the claims and as described herein below.

It has been found that the invention and/or embodiments thereof have a number of additional advantages which will be clear to the skilled person from the following description. It should be emphasized that the term "comprises/comprising" when used herein is to be interpreted as an open term, i.e. it should be taken to specify the presence of specifically stated feature(s), such as element(s), unit(s), integer(s), step(s) component(s) and combination(s) thereof, but does not preclude the presence or addition of one or more other stated features. The term "substantially" should herein be taken to mean that ordinary product variances and tolerances are comprised.

The scatterometer apparatus of the invention has been found to be highly suitable for determining dimensional information of a sample even where the sample comprises very small structures and where a relative large area of the sample is to be examined. The scatterometer apparatus comprises a light source operable for emitting illuminating light , preferably by a light beam, such as in a single beam path, the light source is preferably a broad band light source, a low NA objective system an image recorder comprising a 2D array of detectors, and an image analyzer operatively connected to the image recorder for analyzing recorded images, wherein the light source is arranged for illuminating a sample by an

illuminating spot size with a dimension larger than 10 μιη, such as at least about 0.1 mm, such as at least about 0.5 mm , the low NA objective system has an effective numerical aperture (NA) of less than 0.3 and is arranged for collecting light scattered light from the sample upon illuminating the sample by the light source and the image recorder is arranged for recording at least a portion of the collected scattered light comprising at least one wavelength in form of at least one image.

The term "dimension" when referring to the spot size is to be interpreted to mean a cross-sectional dimension, such as a diameter of the spot size. The term spot size means the spot size where the illuminating light is impinging onto the sample. Where the illuminating light is focused onto the sample the spot size is determines at the waist of the illuminating light and where the illumination light is not focused onto the sample but for example illuminated directly from the laser source and onto the sample, the spot size is determined as the size of the beam as it is emitted from the laser sight source. In the latter situation the emitted laser light is advantageously spatially coherent.

It is envisaged that the term "dimensional information" may comprise any technique to obtain qualitative or quantitative numbers of parameters describing a structure. This information might be used for, but not limited to, quality control or absolute determination of surface parameters. In an embodiment the dimensional information comprises one or more parameters describing, height, width and/or sidewall angle of a 1 D grating, diameter and/or roundness of pillars or holes and/or pitch of a 2D grating. The term "grating" is herein used to designate structural shapes in particular of a sample to be analyzed by the scatterometer. It is envisaged that the term "grating" should be interpreted to comprise samples with shapes comprising regularly spaced identical, or almost identical, structures or set of structures. The structures can be repeated from two to infinite times. A I D grating' refers to a grating with repeated structures in one direction, this includes, but is not limited to, line gratings. Two or more 1 D gratings on top of each other may still be referred to as a 1 D grating if the repetition direction is the same for all the gratings. Two or more 1 D gratings on top of each other can be referred to as a 2D grating if the repetition direction is not the same for all the gratings. A '2D grating' refers to a grating with repeated structures in two directions. This includes, but is not limited to, meshes, grids and hexagonal arrays of e.g. holes or pillars. Two or more 2D gratings on top of each other can still be referred to as a 2D grating. 1 D gratings and 2D gratings can also be stacked.

It is envisaged that the term "quality control" may comprise any technique to validate a sample or device with respect to an existing sample or device.

Advantageously one or more measuring parameters are compared in the quality control. The result may advantageously be a pass/fail result from the processed measurements. In an embodiment the result of a quality control is in form of a percentage of how close the sample or device is to either pass or fail. In an embodiment a reference sample and the sample under test is measured simultaneously. In an embodiment measurements of the sample and a reference do not need to be obtained at the same time, same place or with the same equipment. In other words, a reference spectrum can be obtained in another lab, using a different setup operated by another user, at another day. In an embodiment the reference spectrum is found using computer modelling as described below. The scatterometer apparatus of the invention has been found to be suitable for characterization of micro- and nano-scale structures, such as structures with dimensions of approximately 1 μιη or smaller. The topography of such structures may be characterized using the scatterometer. Scatterometry is defined as the measurement and analysis of light diffracted by structures. The scattered or diffracted light is a fingerprint or 'signature' that reflects the details of the structure itself. The fraction of the incident power diffracted into any orders is sensitive to the shape and dimensional

parameters of the diffracting structure, and therefore may be used to characterize the structure itself with lateral and vertical resolutions of a few nanometers. Scatterometry relies on a priori knowledge of the approximately dimensions and shape of the imaged structure. Thereby, it is possible to deduce structural parameters with an accuracy of a few nanometers. The deduction is done by finding a best match of modelled and measured diffraction efficiencies.

The present invention provides a scatterometer apparatus or system (in short a scatterometer) that preferably is configured for providing an image of local diffraction efficiencies and deduce measurement data for one or more segments of a sample. The scatterometer advantageously comprises an image analyzer with database information and it is capable of delivering parametric data of micro- and nano-scaled structures of a relatively large area in a rapid, reliable and cost-effective manner. The scatterometer is

advantageously configured for integration with conventional optical

microscopes. The scatterometer apparatus of the invention is in particular advantageously for use for characterization of micro- and nano-textures of a surface over a relatively large area (typical minimum dimension of 10 μιη or more). The present invention also provides a method for this characterization by providing a light source that illuminates the surface with light of a single polarization. The illumination is in an embodiment performed using a low NA objective lens that is capable of both illuminating and collecting scattered light from a large surface area. The collected scattered light is subsequently recorded „ advantageously as a scatter intensity image using an image recorder with a 2D array of detectors, herein also referred to as a detector unit. The detection unit is typically a polarization-insensitive 2D camera, such as a CCD camera, or a CMOS camera, or a hyper-spectral camera, or another type of camera as described further below.

In an embodiment the image recorder obtains data for several wavelengths simultaneously using a single image recorder. One or more pixel sensors for example arranged at a first segment of the image recorder measures light at one wavelength and one or more pixel sensors for example arranged at a second segment of the image recorder, simultaneously, measures light at another wavelength. Such system can be realized using e.g. gradient optical filters or multiple segment optical filters. The detector can be, but is not limited to, a 1 D array of pixels or a 2D array of pixels. The advantage of measuring several wavelengths on the same recorder is a fast and cost effective system.

The illumination and recording is advantageously performed at multiple spectral ranges, i.e. at multiple wavelengths, to provide multiple scatter intensity images. The scatter intensity images are analyzed using the analyzer. The analyzer advantageously comprises information of one or more diffraction efficiencies of known morphology. The information is advantageously comprised in a database and/or in a theoretical/numerical model describing diffraction efficiencies. The analyzer preferably operates to compare a recorded scatter intensity image and deduced diffraction efficiencies with the database, for example in an iterative approach. Via an optimization process, the analyzer may provide information of one or more structural parameters (also referred to as structural features) of the surface. Preferably, the iterative approach is done in combination with theoretical/numerical model information.

The present inventors have realized that it is advantageous to use low NA objectives to enable simultaneous illumination and collection of scattered light for large area 2D surface imaging and multispectral data collection to obtain fast and reliable high resolution scatterometry. Hence, the present invention preferably provide a light intensity at a given position of a surface which depends on the specific local morphology of the surface and by recording images at multiple wavelengths of a large surface area via a low NA objective, a rapid and cost-effective method of characterization of micro- and nano- textured surfaces is obtained.

The scatterometer apparatus of the invention has been found to be very beneficial in that it is suitable for measuring one or more segments of a structured sample and determined structural parameter of such segments with high accuracy such as nanometer accuracy. Simultaneously, the system is capable of imaging a large region (or all) of the sample.

When referring to a sample or a region of a sample it should be interpreted to mean a surface region a surface of the sample and/or an embedded structure of the sample or a region thereof where the sample is at least semi- transparent.

It is envisaged that the term "semi-transparent" throughout this description may comprise any material with absorption different from vacuum and with a thickness such that at least one photon can penetrate.

The terms "transparent" and "semi-transparent" should be interpreted to mean transparent respectively semi-transparent to at least one wavelength within the bandwidth of the light source.

It is envisaged that the term "light" throughout this description may comprise any electromagnetic radiation, including, but not limited to, any

electromagnetic radiation in the wavelength range from infrared wavelength radiation to ultraviolet wavelength radiation, such as any electromagnetic radiation in the visible wavelength range.

It is envisaged that the term "wavelength" throughout this description may comprise a range of wavelengths. Unless otherwise specified, a specification of a given wavelength should be interpreted to mean the center wavelength of a distribution of wavelengths within a preferably narrow band, such as a bandwidth of 1 nm, such as a bandwidth of 3 nm, such as a bandwidth of 7 nm, such as a bandwidth of 15 nm. Advantageously the scatterometer apparatus is configured for obtaining dimensional information about one or more structured segments on or within (embedded) in a sample. A structured segment may for example be present on the surface of the sample.

Alternatively, if the sample is transparent or semi-transparent, a structured segment can be present within the sample. The term "wavelength profile" is used to mean a spectral wavelength profile.

The scatterometer apparatus advantageously comprises a polarizer arranged for providing a single polarization state of light illuminating the sample or for polarizing scattered light collected by the low NA objective system. The polarizer may in principle be positioned anywhere in the scatterometer apparatus for ensuring that the image recorder is capable of recording a single polarization state of collected scattered light. In an embodiment the polarizer is incorporated into the light source. In an embodiment the polarizer is arranged between the light source and the sample. In an embodiment the polarizer is arranged between the low NA objective system and the image recorder.

It is envisaged that the term "low NA objective system" may comprise any objective or light collection system with low effective numerical aperture. The low NA objective system is for example a lens or multiple lenses which collects light with a numerical aperture less than 0.3, a lens or multiples lenses combined with a substrate with an aperture, or another substrate able of partially blocking light, in front of it thus effectively reducing the numerical aperture to below 0.3, a lens or multiple lenses combined with a substrate with an aperture, or another substrate arranged for partially blocking light, on the input side. The substrate with an aperture to partially block the light may in principle have any shape. In a preferred embodiment it is circular disk with two holes, or two lines, or two semi-circles, positioned in equal distance from the center. The substrate with aperture can also comprise multiple holes, lines, semi-circles or other.

It is envisaged that the term "effective NA" should be interpreted to be a resulting NA of the NA objective system which may comprise any objective or light collection system, such as described above including a lens, a lens system and/or where NA has been changed from its native NA of individually elements of the NA objective system. The term "effective NA" also include NA of a collection area for collecting scattered light of a cone of light non-normal to the substrate. This deviates from the strict definition of NA that is defined for normal incidence. As an example, the phrase "an effective aperture of 0.3" should be interpreted as a collection area equal to the collection area of a normal beam direction with an NA of 0.3, independent of the incoming angle of the light.

In an embodiment the low NA objective system comprises or consists of a low NA objective lens. In an embodiment the low NA objective system comprises a lens assembly comprising multiple lenses coordinated to obtain an effective numerical aperture less than 0.3.

In an embodiment the low NA objective system comprises a lens-light blocking substrate assembly comprising at least one lens and at least one light blocking substrate coordinated to obtain an effective numerical aperture less than 0.3.

In an embodiment the low NA objective system comprises a high NA lens or lens system and a lens-light blocking substrate assembly arranged to reduce the NA to obtain an effective numerical aperture less than 0.3. The high NA lens or lens system has an NA higher than 0.3, such as an NA higher than 0.5, such as an NA higher than 0.75, such as an NA higher than 0.90, such as an NA higher than 0.95.

In an embodiment the low NA objective system, such as in form of a lens has an NA up to about 0.20, such as up to about 0.15, such as up to about 0.1 . In an embodiment the low NA objective system has an NA of smaller than 0.20, such smaller than 0.15, or preferably in an interval from 0.01 to 0.30, such as an interval from 0.05 to 0.15. It has been found that these preferred values and intervals add to improve the cost-effectiveness, accuracy, performance or use of the scattero meter. In an embodiment the scatterometer apparatus comprises a focusing lens arranged for focusing the illuminating light onto the sample in a spot size with a dimension larger than 10 μιη. The focusing lens is advantageously in form of a low NA objective system arranged for focusing the illuminating light onto the sample in a spot size with a dimension larger than 10 μιη, and

advantageously, the low NA objective system has a numerical aperture (NA) of less than 0.3.

In an embodiment the low NA objective system arranged for focusing the illuminating light onto the sample and the low NA objective system arranged for collecting light scattered from the sample is a common low NA objective system, the apparatus further comprises a splitter arranged for enabling a common beam path through the low NA objective system of light illuminating the sample and of light scattered from the sample. The splitter is for example a beamcube. The phrase "arranged for enabling a common beam path through the low NA objective system of light illuminating the sample and of light scattered from the sample" should be interpreted to mean that the splitter provides that a portion of the collected scattered light is directed away from the light source and towards the image recorder optionally using one or more intermediate lens or mirrors.

In an embodiment the scatterometer apparatus comprises a water immersion objective or oil immersion objective. The scatterometer is not limited to be used in air or vacuum. The scatterometer can also be used, but is not limited to, in water, in oil solutions, in gaseous atmospheres, through crystals or prisms.

In an embodiment the scatterometer apparatus comprises a broadband light source for illuminating the sample with radiation in a single beam path, a polarizer for providing a single polarization state of light illuminating the sample, a low NA objective lens for focusing illuminating light onto the sample and for collecting scattered light from the sample, a beamcube being positioned to enable a common beam path through said low objective lens of light illuminating the sample and of light scattered from the sample, an image recorder with a 2D array of detectors, an image analyzer preferably with information of diffraction efficiencies of one or more reference textured samples. The low NA objective lens has an NA of less than 0.3 and focus illuminating light onto said sample in a spot size with a diameter larger than 10 μιη. The image recorder records images at at least one wavelength, such as at at least three different wavelengths in order to accurately obtain scattering information from the sample.

An advantage of the scatterometer apparatus is that it is capable of providing nanoscale resolution and large-field-of-view. Hence, a sample with

dimensions on the micrometer or even millimeter scale can be investigated and nanometer resolution of structural features can be analyzed. A further advantage is that the scatterometer apparatus is capable of analyzing samples with two or more segments with different structures close to each other.

In an embodiment the image recorder is configured for recording one or more images of the collected scattered light comprising recording images comprising at least 3 different wavelengths, such as at least 5 different wavelengths, such as at least 10 different wavelengths, such as at least 20 different wavelengths.

The phrase "that the image recorder is configured for recording one or more images of the collected scattered light" should be interpreted to mean that the image recorder is configured for recording one or more images of at least a portion of the collected scattered light. In an embodiment the image recorder is configured for recording a plurality of images of the collected scattered light comprising recording images of different wavelength or wavelength profile.

In an embodiment the scatterometer apparatus comprises 2 or more image recorders. The scatterometer apparatus may in principle comprise as many image recorders as desired and the number and type of image recorder(s) is advantageously selected in dependence of the intended use of the

scatterometer apparatus. In an embodiment the scatterometer apparatus comprises at least 3 separate image recorders, such as at least 5 separate image recorders, such as at least 10 separate image recorders, such as at least 20 separate image recorders. Advantageously each separate image recorder is arranged for recording images at separate wavelengths/and or at separate polarization state of the collected scattered light. The apparatus optionally comprises one or more splitters and/or one or more filters and/or one or more polarizer (polarization splitters) arranged between the low NA objective system and the respective image recorders in order to guide the collected scattered light or portion(s) thereof towards the respective image recorders.

In an embodiment the apparatus comprises a Wollaston prism arranged between the low NA objective system and the respective image recorders, wherein the Wollaston prism is configured for directing one polarization state of the collected scattered light towards a first of the image recorders and for directing a second polarization state of the collected scattered light towards a second of the image recorders.

Where the scatterometer apparatus comprises two or more image recorders it is preferred that the image analyzer is a common image analyzer operatively connected to each of the image recorders for analyzing recorded images.

The image recorder can in principle be any type of recorder suitable of recording an image. Advantageously the image recorder is a camera, such as a CCD camera or a CMOS camera. In an embodiment the image recorder is a 2D recorder comprising a 2D array of detectors in form of pixel detectors. Preferably the image recorder comprises at least 1000 pixels, such as at least 10.000 pixels. In an

embodiment the image recorder is a mega pixel recorder comprising at least 1 mega pixels. In an embodiment the image recorder is a 1 D recorder comprising a 1 D array of detectors in form of pixel detectors. Preferably the image recorder comprises at least 50 pixels, such as at least 300 pixels, such as at least 1000 pixels. In an embodiment the image recorder is a monochrome image recorder.

In an embodiment the image recorder is a color image recorder configured for imaging wavelengths within the visible range.

In an embodiment the image recorder is a camera, such as a CCD camera or a CMOS detector including both color and monochrome detectors. The camera comprises an array of detectors. The array of detectors is

advantageously configured for providing information within the field-of-view and allows different segments within the field-of-view to be analyzed.

In an embodiment a larger field-of-view can be achieved by stitching of images. Stitching, also referred to as merging, is where two or more images are put together to expand the imaged area. Stitching can be performed using either a stage for precise movement of the sample or using software algorithms. As an example, the commercial available software package Adobe PhotoShop can be used for stitching of images. The area-of-interest can be selected either before or after the stitching has been made. Advantageously the image recorder is a multispectral or hyperspectral imaging camera. Such camera is operating relatively fast which may add a further advantage to the scatterometer.

.In an embodiment the filtering of wavelengths may take place in the camera unit. This can be from narrow band sensitive pixels in a pixel detector. As mentioned it is desired that the scatterometer apparatus is configured for recording images at multiple wavelengths for thereby obtaining increased accuracy of the scatterometer determination. The configuration for recording images at multiple wavelengths may for example comprise one or more optical filters as described below. The terms "filter" and "optical filter" are used interchangeable.

In an embodiment the scatterometer apparatus comprise an optical filter. In principle the optical filter may be any type of filter capable of optically filter the light e.g. the illuminating light or the scattered light. In an embodiment the scatterometer apparatus comprises at least one of a monochromator, a passive filter for a fixed range of wavelength, or a configuration of

interchangeable passive filters, or a liquid crystal tunable filter, or a variable optical filter, or a multiple segment optical filter. Examples of suitable optical filters are "Ultra-Narrow LV VIS Bandpass Filter" marketed by Delta Optical Thin Film A/S, Denmark, and "VariSpec LC

Tunable Filters", marketed by PerkinElmer Inc, USA.

In an embodiment the optical filter is in form of a filter unit.

Advantageously the optical filter is arranged to filter at least a portion of the collected scattered light in front of the image recorder.

It should be understood that any position or arrangement of elements of the scatterometer apparatus in relation to it other should be determined in relation to the propagation of the illuminating light or the scattered light. The phrase "that the optical filter is arranged to filter at least a portion of the collected scattered light in front of the image recorder" should therefore be interpreted to mean that the filter is arranged such that the collected scattered light will be filtered prior to reaching the image recorder.

In an embodiment where the apparatus comprises two or more image recorder it is desired that separate optical filters are positioned in front of the respective separate image recorders, in order to record images comprising different wavelength(s) for the respective separate image recorders.

In an embodiment the optical filter or the separate optical filters is/are configured for filtering at least 3 different wavelengths from the broadband light source, such as at least 10 different wavelengths, such as at least 20 different wavelengths

In an embodiment, the scatterometer apparatus comprises a filter unit, such as a monochromator or a configuration of interchangeable passive filters. This may be advantages for cost effective reasons. Preferably, the filter unit is capable of filtering at least 3 different wavelengths from the broadband light source, such as at least 10 different wavelengths, such as at least 20 different wavelengths.

In an embodiment the optical filter or the separate optical filters is/are synchronized with the image analyzer(s) with respect to at least wavelength(s), preferably the analyzer are synchronized with the optical filter(s) and/or the image analyzer(s) at least with respect to wavelength(s).

In an embodiment, the image recorder and the image analyzer are

synchronized with respect to wavelength. This provides increased accuracy of the scatterometer apparatus.

In an embodiment the one or more light sources are synchronized with the image analyzer(s) with respect to at least wavelength(s) and/or light intensities. Preferably the analyzer is synchronized with said light source(s) at least with respect to wavelength(s) and/or light intensities. In an embodiment the intensity (intensities) of the light source(s) is/are modulated. This has the advantage of reducing noise on the measurements and/or obtaining the dark images faster.

The image analyzer is advantageously configured for providing information of at least two segments having different structures of the sample. The segments may for example comprise one or more structural features which may be equal or different from segment to segment.

In an embodiment the image analyzer is programmed for determining dimensional information of at least two distinguished segments having different structures and/or structural features of the sample. The image analyzer advantageously comprises a database comprising reference data, preferably in form of diffraction efficiency data of known references. Advantageously the diffraction efficiency data comprises information of diffraction efficiencies of one or more reference features. In an embodiment the image analyzer is programmed for analyzing the recorded images, the image analyzer is preferably programmed for

determining diffraction efficiency or diffraction efficiency pattern for a pixel sensor and/or a group of pixel sensors and/or at least one wavelength for one or more of the recorded images. In an embodiment the image analyzer is programmed for comparing the determined diffraction efficiency or diffraction efficiency pattern and for determining dimensional information of the sample based on the comparison.

The light source is advantageously a laser light source. In an embodiment the light source is a broad band light source spanning at least 100 nm, such as at least 250 nm, such as at least 500 nm, such as at least 800 nm.

Advantageously the light source comprises a broad band light source selected from a broad area laser diode, a super luminescent diode laser, a Xenon lamp, a laser-driven light source, or a supercontinuum laser.

In an embodiment the broadband light source is a swept laser source, a tunable laser source, a broad area laser diode or a super luminescent diode laser, or a supercontinuum laser, or several individual lasers, or an array of laser diodes. This may be advantageous as such lasers may have better light performance than alternative sources, such as lamps or light emitting diodes. Also, lasers are available at many different wavelengths. Lasers may also be preferred due to high brightness. This may for example reduce noise in a scatterometer apparatus. However, for cost-effective reasons also lamps or light emitting diodes may be preferred for the broadband light source.

Preferably, the illuminating light is from a broad band light source and is filtered by a monochromator. In an embodiment the scatterometer apparatus comprises a monochromator arranged to filter light from the broadband light source before the light is incident on the sample.

In an embodiment the light source comprises a narrow band light source having a bandwidth of up to 100 nm, such as up to 50 nm, optionally the narrow band light source comprises a diode and/or a tunable laser.

In an embodiment the light source comprises a laser array, such as an array of narrow band light lasers, such as a diode array.

In an embodiment the light source is arranged for illuminating substantially collimated light, preferably the light source is arranged for illuminating substantially collimated light onto the sample with a spot size with a dimension larger than 10 μιη.

In an embodiment the light source is arranged for illuminating the illuminating light onto the sample without any intermediate lens. In an embodiment the apparatus is configured for operating in a stray through configuration for analyzing an at least semi-transparent sample where the light source is arranged on a rear side of the sample and the low NA objective system for collecting scattered light and the image recorder is arranged on a front side of the sample. In an embodiment the apparatus is configured for operating in a split configuration where the light source is arranged for impinging onto the sample in an impinging center direction and the low NA objective system for collecting scattered light is arranged for collecting scattered light in a collecting center direction, which is different from the impinging center direction. In an embodiment the angle between the incoming light and the detector is up to 175 degrees, such as from 5 to 160 degrees. In other words, both the incoming light and the detected light can be varied with an angle with respect to the normal of the sample from 0 to 87.5 degrees, such as from 2.5 to 80 degrees. The invention also comprises a method of determining dimensional information of a sample by use of the scatterometer apparatus as described above.

The method of the invention advantageously comprises the following steps. · illuminating the sample by impinging illuminating light from the light source onto the sample an illuminating spot size with a dimension larger than 10 μιη,

• collecting light scattered from the sample by the low NA objective

system,

· recording one or more images of at least a portion of the collected

scattered light comprising at least one wavelength by the image recorder and

• analyzing the recorded images by the image analyzer for thereby

determining the dimensional information.

The individual steps can be performed simultaneously in an overlapping configuration or at least the step of analyzing can be performed in a separate step.

Advantageously the sample comprises one or more structured segments and the method comprises determining dimensional information of the one or more structured segments. In an embodiment the method comprises recording images of one or more structured segments of the samples at at least three different wavelengths.

In an embodiment the analyzing of the recorded images comprises

• providing a priori information of the micro/nano-textured surface, · measuring images of scattering intensities, preferably in form of

diffraction efficiencies or diffraction efficiency patterns, and

• determining a best set of dimensional parameters from database

lookup and/or from electromagnetic calculations. Advantageously the electromagnetic calculations comprise using a numerical solver.

Advantageously the method further comprises at least one of:

• generation of a database of reference data comprising computer

modeled scattering intensities, preferably in form of diffraction efficiencies or diffraction efficiency patters,

• optimization of parameter fit and check for convergence,

• theoretical profile matching of the profile of the structured sample.

The method may for example comprise performing of two or more steps in an iterative process and/or estimating a measurement uncertainty.

The invention also comprises a quality control method for determining at least one structural quality parameter of a sample by use of the scatterometer apparatus described above. The structural quality parameter is

advantageously a quality parameter of structural features of a surface of a sample or embedded in a semi-transparent sample. The quality control method comprises

• providing reference data comprising sets of dimensional parameters describing said structural quality for one or more reference samples; · determining a set of dimensional parameters of the sample according to the method described above;

• comparing said determined set of dimensional parameters of the

sample with said reference data and determining said structural quality parameter of the sample.

The reference data may be as described above or in the examples. In an embodiment the reference data comprises experimentally measured data using the same or any other instrument capable of providing dimensional parameters e.g. in form of scattering intensities. In an embodiment the reference data comprises sets of dimensional parameters describing the structural quality for at least two reference samples.

In an embodiment the quality parameter is in form of a pass-or-fail parameter or in form of percentage of how close the sample or device is to either pass or fail.

All features of the inventions and embodiments of the invention as described above including ranges and preferred ranges can be combined in various ways within the scope of the invention, unless there are specific reasons not to combine such features.

Further it should be understood that a term when used in singular is to be interpreted to also include the plural meaning of the term unless otherwise stated or made clear from the context.

As it will be further made clear to the skilled person the scatterometer apparatus and the method of the invention provides a may offer a cost and space effective replacement for complex measurement systems and methods. An additional advantage is that in an embodiment the scatterometer

apparatus may be connected to an existing microscope.

DESCRIPTION OF EXAMPLES WITH REFERENCE TO THE DRAWINGS The above and other features and advantages of the present invention will become readily apparent to those skilled in the art by the following detailed description of exemplary embodiments thereof with reference to the attached drawings.

Fig. 1 schematically shows an exemplary imaging scatterometer, Fig. 2 schematically shows an exemplary imaging scatterometer,

Fig. 3 schematically shows exemplary 1 D and 2D periodic structures, Fig. 4 schematically shows exemplary single structures, Fig. 5 schematically shows an exemplary buried structure,

Fig. 6 schematically shows exemplary sample, field-of-view, segment/area-or- interest, micro/nano-structures, and dimensions of the micro-nano-structures.

Fig. 7 schematically shows an exemplary dataset of diffraction efficiencies for a scatterometer,

Fig 8 shows a flow diagram of an exemplary method for measuring or determining a profile and dimensional parameters of a structured sample.

Fig. 9 shows schematically field of view of an exemplary imaging

scatterometer that provides imaging of a sample with multiple structured segments.

Fig. 10 shows schematically how diffraction efficiency varies as function of wavelength for a segment of a sample with multiple structured segments.

Fig. 1 1 schematically shows an exemplary imaging scatterometer in a stray through configuration. Fig. 12 schematically shows an exemplary imaging scatterometer in a split configuration.

Fig. 13 schematically shows another exemplary imaging scatterometer in a split configuration.

Fig. 14 shows examples of experimental data obtained with a filter with a maximum transmission at 440 nm.

Fig. 15 shows example of experimental data of 4 pixels (binned 2x2) in the range 440 nm to 690 nm.

Fig. 16 schematically shows an exemplary imaging scatterometer with double imaging. Fig. 17 schematically illustrates examples of objectives with different NAs and a high NA lens system with an effective low NA. Fig. 18 schematically illustrates optionally parts of an effectively low NA objective system.

Fig. 19 shows experimental data at 488 nm obtained using filters at the output side with the scatterometer in a beam cube configuration. The figures are schematic and may be simplified for clarity. Throughout, the same reference numerals are used for identical or corresponding parts.

Fig. 1 schematically shows an example of a scatterometer apparatus according to the present invention. The apparatus provides dimensional information of a structure formed on a sample 6. The scatterometer apparatus comprises a broadband light source 1 for illuminating the sample with radiation in a single beam path, the sample is held on a xyz stage (typically in a microscope), a polarizer 3 for providing a single polarization state of light illuminating the sample, a low NA objective lens 5 for focusing illuminating light onto the sample and for collecting scattered light from the sample, an optical filter 2 is used to select a certain wavelength or wavelength range from the white light source, a beamcube 4 is positioned to enable a common beam path through the low NA objective lens of light illuminating the sample and of light scattered from the sample. The apparatus also comprises an image recorder with a 2D array of detectors 9. The camera unit may comprise an integrated image analyzer with information of diffraction efficiencies of one or more reference textured samples. The image analyzer may also be a separate unit. The apparatus use a low NA objective lens with an NA of less than 0.3 that focus illuminating light onto the sample in a spot size with a diameter larger than 10 pm, and the image recorder records images at at least three different wavelengths.

Fig. 2 schematically shows another example of a scatterometer apparatus according to the present invention. In this example the optical filter is located in the beam path after the light is incident on the sample. Thus the filter is used to filter the collected scattered lights. In the example of Fig. 1 , the filter is used to filter the incoming light (the illuminating light). The present invention can be used on samples comprising a broad range of structures, including periodic structures (all pitches are possible, for example from 50 nm to mm range), single structures (i.e. microfluidic channel). The structures can be buried in a semi-transparent material. Fig. 3, 4 and 5 give examples of structures of relevance for the present invention. As indicated by the structures shown in Fig. 3 and/or Fig. 4 the scatterometer apparatus and the method can be applied for determining dimensional information from both grooves and ridges. As shown in Fig. 5, structures can also be measured when they are buried in a transparent or semi-transparent material. This works for both 1 D, 2D and individual structures.

In Fig. 3 are examples of 1 D periodic structures, also referred to as 1 D gratings, and 2D periodic structures, also referred to as 2D gratings, sketched. Typical parameters to describe a 1 D grating are indicated in Fig. 3. These parameters are the pitch, also referred to as the period of the grating, Γ, the height, h, the width, w, and the angle of the sidewalls, a. Typically the width is defined as the full width at half maxima (FWHM). For a 1 D grating with a sidewall angle of 90 degrees, the width is the same at all heights. More parameters not sketched in the figure can also be used to describe the shape of the structures. Examples of more parameters are: roundness of top corners, roundness of bottom corners, sidewall roughness, line edge roughness, and skewness. For more detailed description of a grating, one may use several structures on top of each other. As an example, the sidewall angle can change as a function of the height of the structures. Typical parameters to describe 2D periodic structures, also referred to as 2D gratings, are the pitches in two dimensions, Γ and Γ₂, the height, h, of the structure and the radii at the bottom and top of a circular pillar. Examples of more parameters for 2D gratings are: roundness of shape, roundness at the structure/sample interface, skewness and angle between the repetition directions.

In Fig. 4 are individual structures sketched. An infinite line can be described with the same parameters, except the pitch, as a 1 D grating. A single structure confined in both lateral dimensions can be described with the same parameters, except the pitches, as a 2D grating. The scatterometer apparatus and method described in this invention is capable of imaging areas smaller than 1 μιη² (area-of-interest) with nanometer resolution over a total area of several thousand μιη² (field-of-view). Also, the scatterometer and method provided by this invention provides total image areas (field-of-view) over 1 mm² with a local area resolution (segment/area-of- interest) of the order 10 μιη². In a preferred embodiment, the scatterometer and method provided by this invention provide total image areas (field-of-view) larger than 1 cm². Hence, providing a performance advantage by an increase in local area resolution (segment/area-of-interest). The scatterometer apparatus of the present invention may accordingly provide an imaging scatterometer and a method that present a much faster and cheaper solution for finding the area-of-interest of a given sample and simultaneously allows scatterometry to be performed on submicron scale. Fig. 6 shows

schematically the use of the terms field-of-view and segment/area-of-interest for the present invention. The sample is imaged to define the field-of-view and the area-of-interest is the area from where scattered light is collected. As illustrated the area-of-interest can have several micro or nanostructures which can be analyzed using a scatterometer of the invention.

Samples to be measured of the present invention can be of any size. Regarding field-of-view: It is generally of interest to have as large field of view as possible. In a microscope setup with a 5x objective, a field-of-view is around 1 mm x 1 mm. Larger field-of-views are also possible by using a setup with less magnification.

Regarding segment or area-of-interest: Typically, it is challenging to study small segments, where the segments size is comparable to a period or feature dimension of the structure under observation. Hence, typically the segment is larger than the period to be measured. However, it is possible to have small segments sizes, for example with high magnification (such as 50x) a segment dimension of <1 μιη is achievable. Regarding resolution on structures: Within each segment it is possible to obtain resolution in the nanometer range (~1 nm) for the structures. Resolution depends on model and parameters, i.e. height, width, sidewall angle. Various structural parameters are indicated in Fig. 3, 4 and 5.

A typically use of a scatterometer apparatus according to a preferred embodiment of the present invention is in a microscope setup. For such a setup, sample size is typically limited by ~10 cm x 10 cm, but larger samples and roll-to-roll imaging is also possible.

A scatterometer apparatus according to the present invention may have several advantages compared to other measurement instruments on the market e.g. as explained above, for example it can be built into a conventional optical microscope, and it can be priced low compared to other products available on the market.

It is preferred to analyze only a small part of the optical spectrum at a time by using almost monochromatic light. Almost monochromatic light can be generated either at the input side of the microscope, that is, before the light interacts with the sample, or on the output side, that is, after the light has interacted by the sample. One can for instance use narrowband filters to select the wavelengths. This requires one filter for each measurement point. To obtain 20 data points, one would need 20 different filters. The filters can be changed by mounting them on a filter wheel, which can be controlled by a computer. Filters can be placed on either the input or output side of the microscope. Another option is to use a monochromator to select the

wavelengths. By selection of grating and slits in the monochromator, specific wavelength ranges can be chosen. A monochromator is used on the input side of the microscope. By using almost monochromatic light with a bandwidth of e.g. 3 nm instead of pure monochromatic light (e.g. laser sources), coherence effects are suppressed.

It is possible that the data acquisition time do not have to be the same for every image. It is possible to increase the signal to noise ratio by having an individual acquisition time for different parts of the spectrum. It is possible that different areas within the same sample do not have the same integration time. The reference and dark spectra should be acquired with the same integration time. The acquisition time can often be found by using a proper reference sample in first place.

In a preferred embodiment of the invention, a hyperspectral CCD imaging system is used. For hyperspectral CCD imaging a spectrum is recorded for every single pixel. There are typically three methods to acquire such an image: (1 ) One can acquire the spectrum simultaneously and spatially scan the sample, (2) one can acquire a full image and scan the spectrum in steps, and (3) interferometer based acquisition where image and spectrum is recorded simultaneously. The incoming beam is linearly polarized by passing through a polarizer. The polarizer can be either a filter or a crystal e.g. Wollaston Glenn-Taylor,

Rochon and/or Glan-Laser. The orientation of the polarization should be known as the sample has been aligned with respect to this. Besides this, the polarizer can be treated as a passive component, that once installed, does not have to be adjusted during imaging. This a major difference from null- ellipsometry setups (e.g. imaging ellipsometers) where the optical

components are adjusted for each image. The passive setup for this invention makes the instrument more suited for use outside a research environment, such as a production floor. The light is collimated at the entrance port of the microscope.

The light is coupled in using a non-polarizing fifty/fifty beam splitter. With a beam cube the light can travel normal to the substrate both before and after interaction with the sample.

The light is focused onto the sample with an objective. By mounting a revolver on the microscope, one can easily change objectives, and thus magnification. With a low magnification objective, say 5x, it easy to find the overall area of interest. One can then change to a higher magnification objective, say 50x, for the measurements. However, to avoid depolarization effects and collecting higher order scattered light from periodic structures, it is favorable to use an objective with small numerical aperture. This is in contrast to conventional microscopy where a high numerical objective gives the best images. Objectives up to 0.95 for air and 1 .6 for immersion oil based objectives are widely used in microscopy. These advanced objectives can be very expensive and is a major cost of the total microscope price. By avoiding these objectives the price of the microscope can be kept low and at the same time one obtain better scatterometry data.

In a preferred embodiment the image can be further magnified after the objective using one lens or a set of lenses. The magnification can take place after the beam splitter. The magnification can take place before the beam splitter. The magnification can take place both before and after the beam splitter or the magnification can take place with the beam splitter in its beam path. It can be advantageous with this expansion to fill the camera unit with light and to increase the magnification without having to increase the numerical aperture of the objective.

A digital camera, e.g. CCD or CMOS sensor, is placed in the image plane. A set of optical lenses is used to project the image plane on the detector. The lenses can be moved to adjust the projection of the image plane. Usually, this only has to be done once during initial alignment. Afterwards the image plane is brought into focus by adjusting the height of the objective lens, the microscope and/or the sample. To perform a measurement, for each wavelength, two images are acquired before measuring on the sample. First a reference image, Iref(A), on a surface with known reflection coefficients is acquired, e.g. a Si(100) substrate.

Acquisition time is preferably set to take advantage of the dynamic range of the camera. Secondly, a dark Image, Idark(A), is obtained by removing the sample or in some cases by turning the light off. Reflections from the optical elements give a constant signal which is corrected by this measurement. Then an image of the sample is acquired, l_Sam_Pie(A). It is possible to obtain several images of one or more samples with the same set of l_ref(A) and Idark(A). A segment/area-of-interest is selected from the three images l_ref(A), Idark(A), and Isampie(A) leading to local intensities l'_ref(A), l'dark(A), and l'_Sampie(A) to give a diffraction efficiency for each wavelength using: „ I'sample W ^_ I'darkW _.„

1 refW ^{_ 1} darkW

Where R(A) are the wavelength dependent reflection coefficients of the reference sample.

The measured diffraction efficiencies are fitted to a theoretical model using an iterative optimization. The speed of the iterative approach depends on the number of adjustable parameters and requires substantial computation for convergence if many adjustable parameters are used. Alternatively, when more parameters are used in the model that require more computation for convergence and it is possible to use a technique based on database lookup. Hence, in preferred embodiments of the present invention, database lookup is used.

The databases of diffraction efficiencies can be generated either by finite element method software, rigorous coupled-wave analysis (RCWA) or analytically with scalar diffraction theory for the simplest structures. Finite element based software, such as COMSOL or JCMwave can be used to calculate diffraction efficiencies for complex structures, but is a time

consuming method. RCWA is a much faster algorithm for modelling of diffraction efficiencies. It is based on calculating the diffraction individual slabs and then couple each slab with each other through boundary conditions. The RCWA algorithm is the preferred algorithm for fast and precise simulation of diffraction efficiencies from a structured surface. With periodic boundary conditions one can calculate structures such as 1 D gratings and 2D arrays. By applying non-periodic boundary conditions it is also possible to calculate single features such as a single groove or line. Fig. 7 shows an example of data used to obtained structural information of a sample. The sample is a Si(100) wafer with a structured surface of 1 D period structure (1 D grating, as schematically illustrated in Fig. 3, top), having dimensional parameters:

r=800nm (period), h=190nm (height), w=380nm (width) and a=89 degrees (side-wall angle). Models with dimensional parameters a, i.e. variations in height, width and sidewall angle are modelled for each pitch, that is a=(aheight,a_Width,a_Sw)- ^A" modelled diffraction efficiencies are stored in a database and then compared to the experimental data using a least-square optimization as a measure for the goodness of the fit:

Where σί are the uncertainties on the experimental data e.g. as described in Northern Optics, 2006, Issue Date: 14-16 June 2006, Written by Hansen, P.- E.; Agersnap, N.; Kuhle, A.; Garnaes, J.; Petersen, J.C. and f, are the modelled diffraction efficiencies for the i'th element with the shape a. The database element with the lowest χ² value is the best match to the model. It can be seen that experimental data points with an associated large

uncertainty gives a smaller contribution to the sum than data points with a relative small uncertainty. Thus, the best model is found with most weight on the data points with smallest uncertainty. It is further preferred to combine the two methods, that is, first compare the measured reflection coefficients to a pre-generated database, and then perform an optimization from this point. This approach can reduce the computation time significantly without compromising the resolution.

Fig. 7 sketch data points, (λ), and the best fit fj(a) for a nano-textured sample, e.g. a 1 D grating in Si(100). The data is obtained for multiple wavelength, 15 different wavelengths in this example, but fewer as well as higher numbers may also be used.

Fig. 8 show a diagram of a suitable iterative process which may

advantageously be applied in the method of the invention. The scatterometer apparatus may advantageously be programmed to perform such iterative process.

Fig. 9 shows schematically an example of a field of view of an apparatus according to the present invention. The figure also shows multiple structured segments of the sample. As illustrated, the different segments may have different scattering intensities at different wavelengths.

Fig. 10 shows schematically how the diffraction efficiencies varies as a function of wavelength for a segment of a sample with multiple structured segments. The right part of the figure shows schematically how multiple segments of a sample have different diffraction efficiencies. The figure further illustrates how a segment at three different wavelengths may have different scattering intensity.

According to the present invention, a method is also provided for obtaining structural dimensions of micro/nano-structures. The method is outlined schematically as a flow diagram in Fig. 8. The method is based on the description throughout the present document.

General methods for finding optimal structural parameters and measurements uncertainties e.g. as known to a skilled person in the art can be comprised in the method of the invention. Preferably, a method is used, where

electromagnetic solvers are applied in a two-step procedure together with a global and local optimization method to give the best fit between calculated and measured diffraction efficiencies. In the global optimization step the structural parameters are adjusted, and in the local optimization step also the associated covariance matrix is calculated in order to obtain the best estimates of the structural parameters and their associated uncertainties. This procedure is described in Materials Science and Engineering B 165 (2009) 165-168 written by P.-E. Hansen and L. Nielsen.

In another preferred embodiment, the scatterometer is used when a sample has moved between each image acquisition. The movement can, for example, be due to thermal drift. For movements where the area-of-interest/segment is still visible in the same field-of-view, the analyzer unit can compensate for the movement if characteristic features are visible. Characteristic features comprise, but are not limited to, alignment marks, edges, border regions, corner of sample etc. For large movements, where the area-of- interest/segment moves out of the field-of-view, it is preferred to compensate for this by physical moving the area-of-interest/segment back into the field-of- view before data acquisition. With a motorized table, this movement is preferably controlled by a computer unit.

In another preferred embodiment, a scatterometer scan is used when vibrations are present. For example, large machines in a production

environment give vibrations to nearby equipment. Isolation of vibrations is expensive and cumbersome to implement in a production environment. For periodic structures, it is preferred to perform an averaging (binning pixels), in order to increase the area-of-interest to compensate for the vibrations. To compensate for the vibrations the area with periodic micro/nano-structures has to be larger than the amplitude of the vibrations. For instance, a 500 μιη area with periodic structures can be imaged with a 100 μιη area-of-interest with vibrations up to 200 μιη in the lateral directions. Vibrations affecting the height, and thus the focus, are preferably compensated for by using a larger area-of-interest.

The scatterometer system is preferably used also when both movement, e.g. drift, and vibrations are present, for example for roll-2-roll production.

In a preferred embodiment of the invention, a scatterometer comprises three or more individual or separate image recorders that obtain separate images. The images are preferably processed by a single image analyzer unit, alternatively two or more image analyzer units. Preferably, three image recorders each are arranged for obtaining an image at a specific wavelength, alternatively at two or more wavelengths. Images are processed using a similar method as when acquired by scatterometer with a single image recorder. It is an advantage that a scatterometer comprising multiple image recorders that a system with parallel processing can lower acquisition time. This is relevant for industrial uptake of the described techniques. For example, for roll-to-roll fabrication, is it an advantage to can obtain several images, such as 3, 10, or 20 images, at a time. Many different types of samples are relevant to measure using the present invention. For example, reference components, such as MetroChip from the commercial vendor TedPella. Other examples include, holographic

components, such as for counterfeiting of bank notes or other valuable assets, components with structured coloring, microfluidic channels, components comprising overlayer alignment marks. In a preferred embodiment as shown in fig. 1 1 the scatterometer is used in a 'stray through' configuration for semi-transparent samples. A sample 16 comprising a 1 D grating is illuminated by a white light source 1 1 via a linear polarizer 18 from the backside (rear side) and the imaging system comprising a low NA objective system 15, an imaging lens element 17, an optical filter 12 and an imaging recorder 19

In a preferred embodiment there is no beam splitter on the output side. When used in the stray through configuration the formula for calculating the diffraction efficiencies is slightly modified to:

_ I' sample 0 ^~ I' dark (^)

Where the reference l'_ref is the measured intensity through the sample on a non-textured area. The reference free measurements are also available for this measurement scheme. An advantage of the stray through configuration is often a higher light intensity, such that it is possible to lower the acquisition time. As an example one can measure in the production line of roll-2-roll fabrication. A thin semi-transparent foil is illuminated from the backside and an imaging system placed on the other side of the foil. In one configuration this is combined with the multiple detector setup, where each camera records one or more wavelengths. As the samples passes on to the next detector system, the images are combined to a full multi-spectral image.

In a preferred embodiment as shown in Fig. 12 the illumination system comprises a light source 21 , a collimating lens 22, a polarizer 23 and an imaging system comprising a low NA objective system 25, a spectral separator 28 and a camera (CCD) is split into two separate parts to achi an angle of incidence different than normal incidence. The illumination system may in an embodiment further comprise a not shown compensator. The illumination system is arranged to illuminate a sample 26 arranged on a XYZ stage 27. The angle of incidence can vary from 0° to 90°. In a preferred embodiment the illumination system and imaging system is placed

symmetrically around the normal of the sample. The incoming light is

polarized using e.g. a linear polarizer 23. A single wavelength or selections of wavelengths are selected either in the illumination system or in the imaging system. When selected at the imaging system side of the scatterometer, the selection can take place at any place before the camera unit, such as in front of the objective, such as in front of the camera unit. The selection can also take place inside the camera unit. The advantage of this scatterometer configuration is that diffraction efficiencies vary with the incoming angle of light. For some structures a more unique diffraction efficiency curve are found at an angle different from normal incidence.

The scatterometer apparatus of FIG. 13 is a variation of the scatterometer apparatus of Fig. 12, where the position of the NA objective system 25 and the spectral separator 28 have been switched.

The scatterometer can be used to measure on small samples (less than 1 mm²) to large samples (up to 100 cm² or more, such as a 4 inch substrate) covering the full sample in a single image. The scatterometer can also measure on part of an infinite sample, such as a segment of a roll-2-roll production.

Experimental data using the instrument are presented in Figure 14. A

scatterometer system where the illumination source and imaging system is split by 15° has been used to obtain the images. A silicon sample partly patterned with 1 D gratings (16 fields in total, each having a size of 4 mm x 4 mm) and large areas without structures is used for the experiment. The illuminated area of the sample is around 2 cm in diameter. First a reference image was obtained at 51 wavelengths in the range 440 nm to 690 nm. The reference image for at 440 nm is also shown in Figure 14. The reference image is obtained on a flat silicon area in between two areas with nanostructures. In some configurations it is preferably to obtain the reference images on the same substrate to ensure that alignment is maintained. One could also have measured the reference image on another sample, e.g. a flat silicon sample. Only the area without structures can be used for the reference image, hence the other areas have to be discarded in post-processing. Please note the advantage that the user can wait to discard data until the postprocessing. In other words, if part of the image is valid, the measurement can still be used using the method described above. The next step is to obtain the series of 'dark' images for the same wavelengths as the reference image. The dark image is shown as an example. This image is obtained by turning off the light source. Preferably, but not always, should the acquisition time and gain setting of the camera be the same for the dark image as for the reference image. The third and last measurement is of the patterned area. Again, one should preferably, but is not limited to, use the same camera settings and wavelengths as for the reference image. All images obtained at the same wavelength are now grouped together and used to find the diffraction efficiency, η. For reducing the noise in the images, one can choose to bin several pixels together. In the presented images 4 pixels has been binned together (2x2) to one pixel. One pixel is now selected, and the wavelength diffraction efficiencies are found. These data points are used to fit a theoretical model, for instance modelled using RCWA, FEM or scalar diffraction theory. Experimental data points are shown as dots and the best fit is shown as the solid line in Fig.15. The best fit parameters are found to: height = 192 nm, width = 376 nm, and sidewall angle = 88°. Using state-of-art conventional microscopes, the height has been measured to 189 nm with AFM, the width to 380 nm using SEM, and the sidewall angle to 90° using tilted AFM imaging. A detailed description of the reference measurements can be found in

(Reference: Madsen MH et al, 'Fast Characterization of Moving Samples with Nano-Textured Surfaces', Optica, 2, 301 -306 (2015)). It should be stressed that three separate measurements using different equipment is necessary using the conventional microscopy techniques. The scatterometer can for example be used for quality control or dimensional information of a structured surface or structures under a semi-transparent layer. For quality control one can compare the diffraction efficiencies from a sample that has passed conventional quality control with an unknown sample. The advantage of this configuration is that the measurements using the scatterometer can be much faster than conventional characterization techniques. Furthermore it is possible to integrate it into a production line. Reference free scatterometry measurements can be performed by measuring both TE and TM polarized light at the same area of interest on the sample. The intensities of the two polarizations components (l^TE and I™) from the sample may be measured individually or simultaneously. Simultaneously measurements are possible, but not limited to, using a polarization splitter (e.g. Wollaston prism polarizer or Rochon prism polarizer). The wavelength dependent diffraction efficiency, (λ), can be calculated from the measured intensities of the two polarizations components (l^TE and I™) using the expression:

Ι™(λ) - Ι^ΤΕ(λ)

η(λ) =

1™(λ) + Ι^ΤΕ(λ) - 2I^Dark(A)

This formula is the same for reflection and transmission measurements, in contrast to the ordinary diffraction efficiency formulas.

It is envisaged that the term "reference free" refer to a measurement method that does not need information on the intensity of the light source, also referred to as l_ref(A). The light intensity is typically acquired using a reference standard, with filters or directly. This step can be omitted using the reference free measurement method. A measurement of the 'dark signal', Idark(A), might still be needed for reference free measurements.

Data for reference free measurements can be obtained in a single

measurement using a scatterometer with double imaging as sketched in Fig. 16. A Wollaston prism polarizer 38 is used to split the incoming light in its TE {transverse-electric) and TM {transverse-magnetic) components. The TE and TM light can be imaged simultaneously using two detectors 40 and 41 . A single wavelength is selected from a light source 31 using an optical filter 39. The light is directed normal to the sample surface using a 50/50 beam splitter 34 and focused on the sample 36 with a low NA objective 35. The dual imaging setup can also be used in the split configuration. The selection of wavelengths can also take place after the light has interacted with the sample.

In Fig. 17 are different objectives sketched. A low NA objective only collects light within a small angle from normal incidence. A high NA objective collects light within a large angle with respect to normal incidence. By introducing an aperture in front of a high NA objective, its NA will be reduced. This is referred to as an objective with an effective low NA.

Under normal circumstances an objective is completely filled with light. This is sketched in the left part of Fig. 18. By only partially filling the objective with light, center Fig. 18, an incidence angle different than normal can be obtained using a non-split scattero meter. Furthermore, the partially filling of the objective reduces its effective numerical aperture. This can for instance be used to change a high NA objective to an effectively low NA objective. It can also be used for a further reduction of the NA of a low NA objective. The right part of Fig. 18 shows an example of an insert for partially filling the objective with light. Two holes symmetrically around the center of the insert is used for guiding the light in and out, respectively. Other shapes than holes can also be used for the aperture. Non-symmetrically placed holes can also be used.

Experimental data, diffraction efficiency, obtained using a 488 nm optical filter at the output side is shown in Fig. 19. A white LED cold light source (CLS-LED USB, Qioptiq) is used as the light source. A CCD camera with a resolution of 1392 x 1040 pixels is used for data acquisition. A 5x infinite focus objective with a numerical aperture of 0.13 was used for the acquisition and a further 12x magnification applied after the beam splitter in the scatterometer, given a total magnification of 60x. The sample is a Si(100) with etched line gratings, 1 D gratings, with a period of 700 nm. Two different orientations of the 700 nm gratings can be observed on the same image. Also a blank area of Si(100) are observed on the image. The diffraction efficiency image has been found by measuring on a reference sample blank, Si(100), obtaining a dark

measurement and measuring on the sample. All measurements were performed with the same settings of gain and acquisition time of the camera unit.

Claims

PATENT CLAIMS

1 . A scatterometer apparatus suitable for determining dimensional information of a sample, the scatterometer apparatus comprising: a light source operable for emitting illuminating light, preferably by a light beam, such as in a single beam path, the light source is preferably a broad band light source, a low NA objective system, an image recorder comprising a 2D array of detectors, and an image analyzer operatively connected to said image recorder for analyzing recorded images, wherein, said light source is arranged for illuminating a sample by an illuminating spot size with a dimension larger than 10 μιη, such as at least about 0.1 mm, such as at least about 0.5 mm, said low NA objective system has an effective numerical aperture (NA) of less than 0.3 and is arranged for collecting scattered light from the sample upon illuminating the sample by said light source and said image recorder is arranged for recording at least a portion of said collected scattered light comprising at least one wavelength in form of at least one image.

2. The scatterometer apparatus according to claim 1 , wherein said apparatus comprises a polarizer arranged for providing a single polarization state of light illuminating the sample or for polarizing scattered light collected by said low NA objective system, said polarizer is optionally incorporated into said light source, is arranged between the light source and the sample and/or between the low NA objective system and said image recorder.

3. The scatterometer apparatus according to claim 1 or claim 2, wherein the image recorder is configured for recording one or more images of said collected scattered light comprising recording images comprising at least 3 different wavelengths, such as at least 5 different wavelengths, such as at least 10 different wavelengths, such as at least 20 different wavelengths.

4. The scatterometer apparatus according to any one of the preceding claims, wherein the image recorder is configured for recording a plurality of images of said collected scattered light comprising recording images at different wavelengths, or wavelength profile.

5. The scatterometer apparatus according to any one of the preceding claims, wherein scatterometer apparatus comprises 2 or more image recorders, such as at least 3 separate image recorders, such as at least 5 separate image recorders, such as at least 10 separate image recorders, such as at least 20 separate image recorders, wherein each separate image recorder preferably is arranged for recording images at separate

wavelengths/and or at separate polarization state of said collected scattered light, said apparatus optionally comprises one or more splitters and/or one or more filters and/or one or more polarizer arranged between said low NA objective system and said respective image recorders.

6. The scatterometer apparatus according to claim 5, wherein said apparatus comprises a polarizer arranged between said low NA objective system and said respective image recorders, wherein the polarizer is configured for directing one polarization state of said collected scattered light towards a first of said image recorders and for directing a second polarization state of said collected scattered light towards a second of said image recorders, preferably the polarizer is selected from Wollaston, Glan Taylor, Rochon and/or Glan-Laser, .

7. The scatterometer apparatus according to claim 5 or claim 6, wherein said image analyzer is a common image analyzer operatively connected to each of said image recorders for analyzing recorded images.

8. The scatterometer apparatus according to any one of the preceding claims, wherein said image recorder is a camera, such as a CCD camera or a CMOS camera.

9. The scatterometer apparatus according to any one of the preceding claims, wherein said image recorder is a 2D recorder comprising a 2D array of detectors in form of pixel detectors, preferably the image recorder comprises at least 1000 pixels, such as at least 10.000 pixels, preferably the image recorder is a mega pixel recorder comprising at least 1 mega pixels.

10. The scatterometer apparatus according to any one of claims 1 -8, wherein said image recorder is a 1 D recorder comprising a 1 D array of detectors in form of pixel detectors, preferably the image recorder comprises at least 50 pixels, such as at least 300 pixels, such as at least 1000 pixels.

1 1 . The scatterometer apparatus according to any one of the preceding claims, wherein said image recorder is a multispectral or hyperspectral imaging camera.

12. The scatterometer apparatus according to any one of the preceding claims, wherein said image recorder comprises a plurality of detectors, said image recorder is configured to simultaneously record at least two

wavelengths by at least two different sensors, such as pixel sensors.

13. The scatterometer apparatus according to any one of the preceding claims, wherein the image recorder is a monochrome image recorder.

14. The scatterometer apparatus according to any one of the preceding claims wherein the image recorder is a color image recorder configured for imaging wavelengths within the visible range.

15. The scatterometer apparatus according to any one of the preceding claims, wherein the scatterometer apparatus further comprise an optical filter, such as a monochromator, a passive filter for a fixed range of wavelength, or a configuration of interchangeable passive filters, or a liquid crystal tunable filter, or a variable optical filter, or a multiple segment optical filter.

16. The scatterometer apparatus according to claim 15, wherein the optical filter is arranged to filter at least a portion of the collected scattered light in front of the image recorder.

17. The scatterometer apparatus according to claim 15 or claim 16 where the apparatus comprises two or more image recorder and wherein separate optical filters are positioned in front of the respective separate image recorders, in order to record images comprising different wavelength(s) for the respective separate image recorders.

18. The scatterometer apparatus according to any one of claims 15-16, wherein the optical filter or the separate optical filters is/are configured for filtering at least 3 different wavelengths from the broadband light source, such as at least 10 different wavelengths, such as at least 20 different wavelengths.

19. The scatterometer apparatus according to any one of claims 15-18, wherein the optical filter or the separate optical filters is/are synchronized with the image analyzer(s) with respect to at least wavelength(s) and/or light intensities, preferably the analyzer are synchronized with said optical filter(s) at least with respect to wavelength(s) and/or light intensities.

20. The scatterometer apparatus according to any one of the preceding claims wherein the light source(s) is/are synchronized with the image analyzer(s) with respect to at least wavelength(s) and/or light intensities, preferably the analyzer IS synchronized with said light source(s) at least with respect to wavelength(s) and/or light intensities.

21 . The scatterometer apparatus according to any one of the preceding claims, wherein the light source intensity is modulated.

22. The scatterometer apparatus according to any one of the preceding claims, wherein the image analyzer is configured for providing information of at least two segments having different structures of the sample.

23. The scatterometer apparatus according to any one of the preceding claims, wherein the image analyzer is programmed for determining

dimensional information of at least two distinguished segments having different structures of the sample.

24. The scatterometer apparatus according to any one of the preceding claims, wherein the image analyzer comprises a database comprising reference diffraction efficiency data of known references, preferably said diffraction efficiency data comprises information of diffraction efficiencies of one or more reference features.

25. The scatterometer apparatus according to any one of the preceding claims, wherein the image analyzer is programmed for analyzing said recorded images, said image analyzer is preferably programmed for determining diffraction efficiency or diffraction efficiency pattern for a pixel sensor and/or a group of pixel sensors and/or at least one wavelength for one or more of said recorded images.

26. The scatterometer apparatus of claim 25 wherein the image analyzer is programmed for comparing said determined diffraction efficiency or diffraction efficiency pattern and for determining dimensional information of the sample based on said comparison.

27. The scatterometer apparatus according to any one of the preceding claims, wherein the light source is a broad band light source spanning at least 100 nm, such as at least 250 nm, such as at least 500 nm.

28. The scatterometer apparatus according to any one of the preceding claims, wherein the light source comprises a broad band light source selected from a broad area laser diode, a super luminescent diode laser, or a supercontinuum laser, optionally the laser is tunable.

29. The scatterometer apparatus according to any one of the preceding claims, wherein the light source comprises a narrow band light source having a bandwidth of up to 100 nm, such as up to 50 nm, such as up to 10 nm, optionally the narrow band light source comprises a diode and/or a tunable laser.

30. The scatterometer apparatus according to claim 29, wherein the light source comprises a laser array, such as an array of narrow band light lasers, such as a diode array.

31 . The scatterometer apparatus according to any one of the preceding claims, wherein the apparatus comprises a monochromator arranged to filter light from the broadband light source.

32. The scatterometer apparatus according to claim 31 , wherein the monochromator is arranged to filter light from the broadband light source before the light is incident on the sample.

33. The scatterometer apparatus according to any one of the preceding claims, wherein said light source is arranged for illuminating substantially collimated light onto said sample with a spot size with a dimension larger than 10 μιη.

34. The scatterometer apparatus according to any one of the preceding claims, wherein said light source is arranged for illuminating said illuminating light onto said sample without any intermediate lens.

35. The scatterometer apparatus according to any one of the preceding claims 1 -32, wherein apparatus comprises a focusing lens arranged for focusing said illuminating light onto said sample in a spot size with a dimension larger than 10 μιη.

36. The scatterometer apparatus according to any one of the preceding claims 1 -33 and 35, wherein the apparatus comprises a low NA objective system arranged for focusing said illuminating light onto said sample in a spot size with a dimension larger than 10 μιη, said low NA objective system preferably has a numerical aperture (NA) of less than 0.3.

37. The scatterometer apparatus according to claim 36 wherein said low NA objective system arranged for focusing said illuminating light onto said sample and said low NA objective system arranged for collecting light scattered light from the sample is a common low NA objective system, said apparatus further comprises a splitter arranged for enabling a common beam path through said low NA objective system of light illuminating the sample and of light scattered from the sample, said splitter preferably being a beamcube.

38. The scatterometer apparatus according to any one of the preceding claims 1 -33, wherein the apparatus is configured for operating in a stray through configuration for analyzing an at least semi-transparent sample where the light source is arranged on a rear side of the sample and the low NA objective system for collecting scattered light and the image recorder is arranged on a front side of the sample.

39. The scatterometer apparatus according to any one of the preceding claims 1 -36, wherein the apparatus is configured for operating in a split configuration where the light source is arranged for impinging onto the sample in an impinging center direction and the low NA objective system for collecting scattered light is arranged for collecting scattered light in a collecting center direction, which is different from the impinging center direction, preferably the impinging center direction has an angle of the impinging center direction up to 175 degrees, such as from 5 to 160 degrees.

40. The scatterometer apparatus according to any one of the preceding claims, wherein the low NA objective system comprises a low NA objective lens.

41 . The scatterometer apparatus according to any one of the preceding claims, wherein the low NA objective system comprises a lens assembly comprising multiple lenses coordinated to obtain an effective numerical aperture less than 0.3

42. The scatterometer apparatus according to any one of the preceding claims, wherein the low NA objective system comprises a lens-light blocking substrate assembly comprising at least one lens and at least one light blocking substrate coordinated to obtain an effective aperture less than 0.3

43. A scatterometry method of determining dimensional information of a sample by use of the scatterometer apparatus according to any one of the preceding claims, the method comprising: • illuminating said sample by impinging illuminating light from said light source onto said sample an illuminating spot size with a dimension larger than 10 μιη,

• collecting light scattered from the sample by said low NA objective system,

• recording one or more images of at least a portion of said collected scattered light comprising at least one wavelength by said image recorder and

• analyzing said recorded images by said image analyzer for thereby determining said dimensional information.

44. The scatterometry method of claim 43, wherein the sample comprises one or more structured segments and the method comprises determining dimensional information of said one or more structured segments.

45. The scatterometry method of claim 43 or claim 44, wherein the method comprises recording images of one or more structured segments of the samples at at least three different wavelengths.

46. The scatterometry method of any one of the claims 43-45, wherein the analyzing of said recorded images comprises · providing a priori information of the micro/nano-textured surface,

• measuring images of scattering intensities, preferably in form of

diffraction efficiencies or diffraction efficiency patterns, and

• determining a best set of dimensional parameters from database

lookup and/or from electromagnetic calculations and/ or from experimentally measured reference data.

47. The scatterometry method of claim 46, wherein the electromagnetic calculations are performed using a numerical solver.

48. The scatterometry method of any one of the claims 43-47, wherein the method further comprises at least one of: • generation of a database of reference data comprising computer modeled scattering intensities, preferably in form of diffraction efficiencies or diffraction efficiency patters,

• optimization of parameter fit and check for convergence,

· theoretical profile matching of the profile of the structured sample.

49. The scatterometry method of any one of the claims 43-48, wherein the method comprises performing of two or more steps in an iterative process.

50. The scatterometry method of any one of the claims 43-49, wherein the method comprises estimating a measurement uncertainty.

51 . A quality control method for determining at least one structural quality parameter of a sample by use of the scatterometer apparatus according to any one of the preceding claims 1 -42, the method comprising:

• providing reference data comprising sets of dimensional parameters describing said structural quality for one or more reference samples;

• determining a set of dimensional parameters of the sample according to the method described in any one of claims 43-50;

• comparing said determined set of dimensional parameters of the

52. The quality control method of claim 51 , wherein said reference data comprises experimentally measured data using the same or any other instrument capable of providing dimensional parameters e.g. in form of scattering intensities.

53. The quality control method of claim 51 or claim 52, wherein said quality parameter is in form of a pass-or-fail parameter or in form of percentage of how close the sample or device is to either pass or fail.