Apparatus and method for profilometry of flat objects with unknown material

Abstract

The invention relates to a method and a device (2) for profilometry of an object surface (40) of a flat object (20) having an unknown material, the device comprising: an optical interferometric measuring system (4); an ellipsometric measuring system (5); a beamsplitter (82) for splitting a light beam of a light source (111) into an interferometric measuring light beam (410) and an ellipsometric light beam (420); and an evaluation unit configured to ascertain a profile height in a measurement region on the object surface (40) from an evaluation light beam (490) evaluated in a detector unit (250) of the interferometric measuring system and a sensor light beam (520) received in an ellipsometric sensor (220). The interferometric measuring system (4) comprises a beamsplitter (91), a reference mirror (60) and a detector unit (250). The ellipsometric measuring system (5) comprises: a polarizer (190) for polarizing the ellipsometric light beam (420) and passing it onto the measurement region on the object surface (40); and an ellipsometric sensor (220) with a polarization filter to determine the polarization state of the received sensor light beam.

Description

Technical Field

[0001] This invention relates to an apparatus and a corresponding method for contour measurement of a flat object made of an unknown material. The apparatus includes an optical measurement system based on interferometric measurement. Background Technology

[0002] In many industries, the surface profile of planar products is measured using optical imaging methods. In the semiconductor industry, these planar products are particularly wafers. Wafers are slices made of semiconductor, glass, or ceramic materials. In some applications, the entire surface or at least most of the surface of the wafer is typically inspected. Circuit "chips" are fabricated on wafers, and these circuit "chips" subsequently serve various purposes in computing, sensor, and control / monitoring electronic devices. During chip manufacturing, various measurement tasks are performed for quality control. For example, the flatness of the chip surface is measured after multiple CMP (Chemical Mechanical Polishing) steps. These steps are implemented to ensure a flat processing surface for subsequent process steps. For this purpose, flatness requirements range from a few nm to 100 nm in current fabrication widths. Typically, defect patterns exhibit lateral extensions from a few μm to a few mm.

[0003] Similar contour measurement tasks need to be performed in other process steps. Therefore, methods for stacking multiple chips to form composites are currently being developed, for example. Here, the use of protruding contact structures should be almost entirely abandoned. These methods are being investigated for integration within wafer composites and for individual chips placed on top of chips still within the wafer composite. For this, accurate and rapid contour measurement is also required to ensure high surface planarity requirements in the low nanometer range. Typical contact structures for these methods have diameters in the range of a few μm and contour heights from a few nm to tens of nm.

[0004] For the tasks exemplified and similar, a rapid and accurate measurement method is required, capable of achieving highly accurate and, as quickly as possible, measurements of the profile. A particular challenge lies in the fact that the surface to be measured is composed of frequently changing materials and may also be covered with one or more transparent layers. However, the profile measurement should accurately reflect the surface of the entire material composite, without errors due to the material or layer system. Similar tasks must be addressed for other structures and in other industries.

[0005] What all these applications have in common is the need for accurate inspection of the surface contours of very flat objects without error due to the underlying material composition, and the requirement for high measurement speeds. These objects include wafers, displays, and the like. Furthermore, these applications share the commonality of often being unable to determine the thickness of the underlying material composition and layered structures with sufficient precision. Another commonality among these applications is the use of sensors to generate contour and / or height data of the structure being probed.

[0006] Traditional height measurement methods use point sensors, line sensors, or surface sensors. The advantage of surface sensors, such as deflection measurements or white light interferometers, is the ability to sense an entire surface. In deflection measurements, the surface to be measured is used as a mirror through which a regularly structured light source can be observed. Changes in the object's height are reflected in the distortion of this mirror image and can be extracted from it. In white light interferometry, a series of records must be made at different distances from the object or the interferometer mirror. In this case, the corresponding profile height at each recorded point is determined from the signal curve of that point within the series. Line sensors use, for example, triangulation, stereometry, or color confocal imaging. What all these methods have in common is that the signals used by these optical methods are affected by the material and any transparent layers that may be present. Without a precise understanding of the thickness of the underlying material composition and layer structure, it is impossible to account for the thickness of the material composition and layer structure sufficiently accurately, for example, in the form of correcting the recorded measurements. This understanding is often unavailable. Existing techniques involve covering key surfaces with a thin metal layer and performing profile measurements on it. It is of great interest to replace this method, which is classified as destructive, with a non-destructive method that can solve measurement tasks.

[0007] US 6,856,384 B1 proposes that, to improve interferometry, a correction factor be determined by post-hoc ellipsometriemessing. Here, multiple ellipsometriemes are performed, requiring only the measurement of the same area on the surface as the interferometry. This combination is advantageous because some components of the measurement system can be used for both measurements. However, the optical measurement system of US 6,856,384 B1 presupposes that the surface gradient to be detected... The material is known, identical in composition, thickness, and flatness, and can be modeled. Furthermore, only individual step heights can be determined within an acceptable measurement timeframe because the measurement system is limited to point sensors. Summary of the Invention

[0008] Therefore, the object of the present invention is to provide an improved apparatus and an improved method, by which reliable and accurate results are provided for measuring the surface profile of a flat object, even when the object has different materials on its surface. In particular, it should be possible to perform measurements with high accuracy in a short time without errors due to structures and layers located beneath the surface.

[0009] The object of the present invention is achieved by an apparatus having the features of claim 1 for contour measurement of a flat object having an unknown material. This object is also achieved by a system having the features of claim 2 and a method having the features of claim 18.

[0010] According to the present invention, this objective is achieved by an apparatus for contour measurement of an object with an unknown material, which includes, in addition to an optical interferometry system, an elliptic polarization measurement system, a beam splitter, and an evaluation unit. The interferometry system and the elliptic polarization measurement system operate in parallel in time, allowing the evaluation unit to process sensor data from both measurement systems simultaneously.

[0011] The beam splitter separates the light beam from the light source into an interferometric beam and an elliptically polarized beam. The interferometric beam is directed to the optical interferometry system, while the elliptically polarized beam is simultaneously directed to the elliptically polarized measurement system.

[0012] Here, the optical interferometry system has a beam splitter that splits the interferometric beam from the light source into a reference beam and a measurement beam. The reference beam is reflected at a mirror (reference mirror) of the system and guided back to the beam splitter as a mirror beam. The measurement beam is guided to the measurement area on the surface of the object to be measured and reflected there. After reflection, the measurement beam is guided back to the beam splitter as an object beam and interferes with the mirror beam to form an evaluation beam, which is supplied to the detector unit of the interferometry system. The detector unit is configured to receive and evaluate the evaluation beam. The profile height of the surface can be inferred from the superposition of the interferometric measurements produced by the two beams and the intensity of the resulting interference signal.

[0013] However, this value is inaccurate due to the different materials of the objects. To correct for these interference signal intensities, parameters obtained using elliptic polarization measurements are used.

[0014] It has proven advantageous that a single ellipsometric measurement is sufficient to determine the correction factor. Therefore, only one additional ellipsometric measurement is performed for each interferometry.

[0015] The apparatus and corresponding system according to the invention are configured and constructed to identify the surface and contour of an object and its structure when the object has unknown materials and / or layered structures. The material and layers of the object need not be homogeneous.

[0016] An elliptically polarized measurement system for calibration includes a polarizer for polarizing an elliptically polarized beam and for transmitting it to a measurement area on the surface of the object to be measured. An elliptically polarized sensor has a polarization filter, allowing the polarization of the received beam to be determined. Therefore, the polarization state of the sensor beam can be analyzed. Here, the elliptically polarized beam is polarized and guided onto the measurement area on the object surface after passing through the polarizer, and after reflection at the measurement area, it is redirected as a sensor beam to the elliptically polarized sensor, so that the elliptically polarized beam, after being filtered by the polarization filter, strikes the polarization sensor.

[0017] The sensor beam received by the ellipsometry sensor is now analyzed in the evaluation unit. From the analysis results, parameters that can be used to correct the measurement results of the interferometry are identified. Preferably, the parameter, referred to as ellipsometry, is a value that reflects the effective refractive index and effective absorption coefficient of the material combination to be measured in the measurement region of the object. Using this ellipsometry, the intensity measured in the interferometry system is corrected so that the actual profile height of the object surface in the measurement region can be inferred and calculated. Therefore, the influence of different materials on the object surface can be corrected.

[0018] Advantageously, this invention enables the measurement of surface structures and profiles on samples of unknown materials or materials with localized variations by simultaneously measuring interferometric measurement signals and elliptic polarization signals, without damaging or altering the material sample, and without requiring explicit modeling of the layer system to be measured. Currently, the practice is to apply a metal layer of constant thickness to the sample. Thus, although the resulting surface structure has a uniform material, it is equivalent to a surface structure without an applied layer. However, the sample being probed becomes unusable after the metal layer is applied. This invention eliminates this drawback of the prior art because it does not damage or alter the object being probed.

[0019] Therefore, by using additional ellipsometric measurements and taking into account the measurable polarization effects, phase shifts caused by local material differences can be eliminated. The correction parameters determined by ellipsometric measurements are used here.

[0020] The object of the present invention is also achieved by a system for contour measurement of the surface of an object, the system comprising: the aforementioned apparatus; a light source for generating a monochromatic light beam; and a moving unit for performing relative movement between the apparatus and the object to be measured.

[0021] Because it is necessary not only to observe the measurement points or measurement areas on the object to be measured, but also to make relative movement between the object and the measuring device. Preferably, the relative movement between the object to be measured and the measuring device can be achieved by moving the measuring device. Here, the object to be measured can be locked in its position. Alternatively or additionally, the object to be measured can also be moved relative to the measuring device. For example, for this purpose, the object can be held on a slide, which is moved by a drive unit in the desired manner and method. Generally, translational movement in one dimension is advantageous. If both the measuring object and the device are made movable, the measurement area can be moved on the surface of the object, thereby allowing, for example, the entire wafer to be measured automatically. This allows the measurement method according to the invention to be automated, as described below.

[0022] In a preferred embodiment, the slide may include a retainer for receiving an object, wherein the retainer is moved by a moving unit. For example, it is possible that the retainer or the slide moves in a back-and-forth motion, while the device moves in a direction perpendicular to this movement.

[0023] The method according to the invention for sensing the surface of an object with an unknown material and for measuring its profile uses interferometry and simultaneously performed elliptometric measurement. A monochromatic beam emitted from a light source is directed toward an optical device and split into an interferometric beam and an elliptically polarized beam by means of a beam splitter. Preferably, the split beams are perpendicular to each other. The interferometric beam is used to perform interferometric measurements, in which the interferometric beam is guided to the measurement area of ​​the object surface to be measured. During interferometric measurements, the interferometric beam can be split into a measurement beam and a reference beam by evaluating the evaluation signal or measuring the light intensity of the interference pattern. The measurement beam and the reference beam interfere with each other after reflection at the measurement area or a mirror and then reach the detector unit.

[0024] A split elliptically polarized beam is used to perform elliptically polarization measurements at a measurement area on the object's surface, preferably using an optical interferometry system. To improve the measurement results, a monitoring signal of the light source is optionally and preferably used to determine the output intensity of the light source and to determine correction parameters by the elliptically polarization measurement. The correction parameters are calculated from the results of the elliptically polarization measurement, taking the output intensity of the light source into account. These correction parameters preferably include the effective refractive index and effective absorption coefficient of the object at the observed measurement point or measurement area. The correction parameters can also be determined without considering the output intensity.

[0025] Further steps of the method according to the invention specify that the measured values ​​identified by interferometry are corrected using correction parameters derived from ellipsometric measurements. The profile height of the measurement area at the object surface is determined from the corrected interferometric measurements and the correction parameters. Therefore, the profile height at the surface of an unknown material of the object can be determined. Possible errors in the interferometry are corrected using the correction parameters.

[0026] By using elliptic polarization measurement, it is not necessary to provide a model of the entire sample, i.e., the object to be measured. It is sufficient to measure only the full effective effects of the surface of the unknown material. These effective optical effects are used to compensate for the effects of interferometric measurement signals based on different material properties and thus allow the actual height of the surface in the measurement region to be inferred.

[0027] In a preferred embodiment of the invention, the beam splitter and / or beam splitter can be designed as a semi-transparent mirror. This makes it possible to achieve simple beam splitting, such as interferometric beams and elliptically polarized beams. With the mirrors arranged at 45°, the two split beams can have an angle of 90° with each other.

[0028] In other preferred embodiments, a lens is provided in the optical path of the device before the measurement area on the surface of the object to be measured. The object directs the beam of light so that the elliptically polarized beam strikes the measurement area on the object surface at a predetermined angle and is reflected. The measurement beam of the interferometric measurement preferably strikes the measurement area on the object surface at a right angle (or approximately a right angle) and is similarly reflected perpendicularly by the object again.

[0029] In a preferred embodiment of the apparatus according to the invention, the elliptometric measurement system includes a mirror in the optical path of the elliptically polarized beam, the mirror being arranged such that the elliptically polarized beam falls perpendicularly onto the polarizer before reaching the object to be measured. A lens is preferably arranged between the polarizer and the object.

[0030] Preferably, the apparatus according to the invention includes a barrel optics in the optical path prior to the detector unit of the interferometric measurement system, which focuses the evaluation beam such that the focal point is located on the detector unit or the sensor surface of the detector unit. Preferably, alternatively or additionally, a barrel optics may be provided in the optical path of the elliptometric measurement system prior to the elliptometric sensor, which focuses the sensor beam onto the elliptometric sensor, i.e., the sensor beam is focused such that the focal point is located on the sensor surface of the elliptometric sensor.

[0031] In a preferred embodiment, the light source of the system according to the invention is a laser. Lasers are particularly suitable for performing interferometric and ellipsometric measurements because they provide monochromatic light. The wavelength of the light is sufficiently constant and can be precisely defined.

[0032] The light source of the system according to the invention preferably includes a monitoring diode, by means of which the output intensity of the light source is determined and monitored. In this way, it can be clearly determined at what intensity the light from the light source is incident on the device according to the invention, so that the light intensity can be used for measurement and evaluation, especially in evaluation in elliptic polarization measurement systems.

[0033] In a preferred embodiment of the system, multiple light sources are provided, each used to generate a monochromatic light beam. Particularly preferably, the light beams from each light source have different wavelengths; most preferably, the wavelength of no single beam is an integer multiple of the wavelength of any other beam. In a particularly preferred embodiment of the system, the light beams are focused in beam-shaping optics. Preferably, the beam-shaping optics are arranged in the optical path before the beam splitter.

[0034] When using multiple light sources, the ellipsometric measurement system preferably includes multiple ellipsometric sensors with polarization filters, which are used to analyze the polarization state of the received sensor beam. Particularly preferably, an ellipsometric sensor is provided for each wavelength used in the existing beam. Therefore, typically an ellipsometric sensor is used for each light source.

[0035] When using multiple elliptically polarized sensors, an optical separation element is preferably provided in the device according to the invention. The optical separation element separates the sensor beams in a wavelength-specific manner according to the wavelengths of all available beams. In this way, portions of each wavelength can be separated. Preferably, the separation element is arranged in the optical path such that the sensor beams strike the elliptically polarized sensors before them.

[0036] In a particularly preferred embodiment, the optical separation element is designed as a dichroic mirror. This dichroic mirror enables precise separation of a specific wavelength or a beam of a specific wavelength in a simple manner. If more than two light sources with different wavelengths are used, it is preferable to arrange multiple separation elements in series to decouple multiple beams of a single wavelength.

[0037] In a preferred embodiment, the device includes a line sensor, which is preferably used in both interferometric and elliptic polarization measurement systems. A line scan camera or a multi-channel line scan camera is preferably used. Using lines with multiple pixels enables parallel measurements and significantly improves the measurement speed of the device.

[0038] Preferably, a four-channel line scan camera can be used, for example, to enable simultaneous measurement of four polarization states, especially when using four polarization filters.

[0039] In a preferred embodiment, the detector unit of the device according to the invention comprises a time-delay integral camera. Most preferably, the detector unit comprises a color time-delay integral camera. In a particularly preferred embodiment, a multi-channel line scan camera is used. Most preferably, they are implemented as time-delay integral multi-channel line scan cameras.

[0040] Other preferred embodiments of the device specify that the elliptic polarization sensor is designed as a time-delay integration camera (TDI camera). Preferably, a line scan camera is used; very preferably, a TDI line scan camera is used; and particularly preferably, a multi-channel TDI line scan camera is used. In a particularly preferred variant, the elliptic polarization measurement system includes a four-channel TDI line scan camera, which particularly preferably has four polarization filters. By using four polarization filters, the polarization state of the beam can be definitively determined, thus requiring only a single elliptic polarization measurement to clearly define the polarization state of the measured beam.

[0041] Therefore, according to the present invention, the proposed objective can be achieved by implementing different method steps, which are hereby reiterated as preferred variations. The steps are as follows:

[0042] a) Continuously illuminate an object with light of at least one monochromatic wavelength, such as from an illumination component or light source;

[0043] b) Relative movement is preferably performed continuously and without interruption between the combined elliptic polarization measurement system / interferometry system and the object;

[0044] c) Split the beam into portions such that the illumination portion (elliptically polarized beam) used for elliptically polarization strikes the surface of the object at a certain angle and is linearly polarized.

[0045] d) The illumination portion (interferometric beam) used for interferometry is redirected such that the illumination portion preferably strikes the surface at at least almost a right angle;

[0046] e) Use the sensors of the elliptic polarization measurement system to record the signal intensity of the beam in at least four polarization directions;

[0047] f) Use the sensors of the interferometric measurement system to record the signal intensity modulated by the surface profile;

[0048] g) Determine the effective refractive index and effective absorption coefficient of the material combination just measured at each measurement point using elliptic polarization data;

[0049] h) Given the influence of the material composition just measured at each measurement point, the intensity measured using the sensor of the interferometric measurement system is corrected by means of the determined effective refractive index and extinction coefficient; and

[0050] i) Calculate the contour height of the object from the corrected interference signal intensity.

[0051] In this arrangement, a row sensor is preferably used for interferometric measurements. This allows for continuous measurement of the width of the object. Preferably, the main focus of the invention is to correct for the effects of different surface materials of the object being measured. For many possible applications, a half-wavelength of light is sufficient for the height measurement range. For a larger operating range, multiple wavelengths can be used simultaneously. The basic concept of this method is disclosed in patent application EP19188318. Details are set forth in the embodiments.

[0052] The distance to the object being measured can be determined clearly with very high resolution by measuring the intensity of a signal modulated by interferometry. However, this only applies when the surface of the object being measured reflects the light 100% or when the reflectivity and possible phase shift are known with corresponding accuracy. This invention is based on the knowledge that the phase shift of reflected light arises from the interference between the light components reflected at the layer interfaces and with the light reflected at the surface. Because deviations in the layer thickness, and sometimes in the material, are generally on the same order of magnitude as changes in the surface being measured, nominal values ​​for the layer system cannot be calculated here. Furthermore, for very thin layers, in addition to the uncertainty in layer thickness, the optical material parameters are not fully known. For layers with a thickness less than the exciton radius (e.g., about 20 nm for Si), the parameters deviate significantly from known volume values ​​due to quantum confinement effects.

[0053] Within the scope of this invention, it is recognized that determining the layer structure solely from the measurement signal itself (e.g., details of the interferogram from a white light interferometer) is very limited and possible only for very simple layer structures. This results in significant inaccuracies. On the other hand, it has been recognized that layer structure is not of great concern for surface profile measurements. Rather, it is sufficient to know or determine the overall effect of the layer stack or the material on the effectiveness of the interferometry.

[0054] This invention is also based on the understanding that the absorption and reflection characteristics of an arbitrary layer system can be explicitly summarized by the effective refractive index n. eff and effective extinction coefficient k eff When the optical parameters, namely the refractive index and absorption coefficient, and the layer thickness in the system are known precisely, the effective refractive index n can be clearly determined using Fresnel formulas and transfer-matrix models that are well known from the literature. eff and effective extinction coefficient k eff Two parameters n eff and k effIt fully describes the reflection at the surface, provided that the sample is made of a homogeneous material or the uppermost layer absorbs so much incident light that the reflection at the interface below it has no significant effect on the measurement signal.

[0055] In the presence of at least partially transparent layers, an effective layer thickness d is additionally required to describe an effective alternative system. eff and through a specific refractive index n subst and extinction coefficient k subst The substrate is described. It has been recognized that this is necessary for representing the phase shift of the reflected light wave caused by interference effects. Therefore, for a complete description of the light wave reflected by the sample to be detected, it is necessary and sufficient to assume a three-layer system (environment, effective layer, and substrate) and use the corresponding parameter n. eff k eff d eff n subst and k subst This is used to characterize the three-layer system. For the environment, without limiting generality, air is assumed in the following text, where n=1 and k=0.

[0056] For the contour measurement to be performed, it can be assumed that it has a parameter n. subst and k subst The substrate material is known. Therefore, the effective parameters n for the refractive index, extinction coefficient, and substrate thickness must be determined. eff k eff and d eff This allows for the accurate correction of interferometric measurements from interferometry without a detailed understanding of the layer system, and thus ascertaines the actual distance between the surface of the object to be measured and the sensor.

[0057] One aspect of the invention is preferably that the parameter n is determined by additional measurements besides recording the interferometric measurement signal. eff k eff and d eff This can be advantageously achieved through parallel ellipsometric measurements. Ellipsometrics measures the change in polarization of a beam of light arriving at the sample in a defined polarization state. For homogeneous materials, representation using a polarization ellipse is sufficient. Here, the polarization ellipse (see...) Figure 2 The polarization variation is shown. The position and shape of the ellipse are determined in ellipsometric measurements and described by the values ​​Ψ and Δ. In this case, the relationship with the complex reflection coefficient ρ is:

[0058] ρ=tan(ψ)e iΔ (1)

[0059] The complex permittivity ε can be calculated directly from ρ.

[0060]

[0061] Here, φ i This represents the angle of incidence of the light beam measured relative to the normal to the object surface (also known as the sample surface) at the measurement point. The relative permittivity is related to the refractive index n and the extinction coefficient k by the following formula:

[0062] (n+ik) 2 =(ε′) r +iε″ r )μ r (3)

[0063] For nonmagnetic materials with n and k (μ) r ~1), the relationship is as follows:

[0064]

[0065]

[0066] These basic relationships are known in principle to those skilled in the art and can be read in detail in several textbooks (e.g., Elliptic Polarization Operation Manual, see reference (2)). The relationships shown apply to each wavelength.

[0067] If one or more transparent layers are to be considered, as is typically required in semiconductor manufacturing for semiconductor wafers or other wafers, conventional methods are insufficient. Therefore, according to the present invention, an effective, resulting layer (through n) is used on a known substrate. eff k eff and d eff The means described in the description.

[0068] Furthermore, information about the basis used (via its parameter n) is also used. subst and k subst The knowledge of (description). If these substrate parameters are not sufficiently known, they can preferably be determined using the apparatus according to the invention in preliminary measurements taken at suitable locations where the substrate is exposed (e.g., at the edge of the wafer or on the back side). Furthermore, the invention utilizes the fact that, with the aid of known transfer matrix models, the optical response of a system consisting of a substrate and one or more material layers can be described as the optical response of a three-layer system (substrate, layers, environment) using effective values ​​for complex refractive index and layer thickness.

[0069] In the mathematical representation of the relationships below, for the sake of brevity, the three layers and their parameters are represented by indices 0, 1, and 2. That is, for the environment, n0 = 1 and k0 = 0; for the substitution layer, n1, k1, and d1 correspond to n eff k eff d eff The group, and for the basis, n2 and k2 represent n respectively. subst and k subst .

[0070] Therefore, according to the present invention, the monochromatic light used is split into two parts, one part (the interferometric beam) for interferometry, and the second part (the elliptically polarized beam) for elliptically polarization measurements to be performed in parallel. For the elliptically polarization measurements to be performed, measurements using a fixed incident angle and a defined polarization direction of the incident beam are sufficient. To definitively determine the polarization state of the reflected beam, four light intensity measurement parameters are preferably required for each of the different polarization directions of the reflected beam.

[0071] The method described here is initially focused on a single detector element. It is understood that it is advantageous for both interferometry and elliptic polarization methods to utilize a line-scan detector to simultaneously measure multiple measurement points. Here, for each zeilenelement, four elliptic polarization intensity measurements and interferometry can be performed individually to determine the parameters being searched. Figure 1 The present invention exemplifies one possible implementation of the measuring device and further elaborates thereon below.

[0072] Below, also for a single detector element and for the wavelength λ used for measurement, the correction for the interferometric measurement signal and the subsequent determination of the correction parameters by elliptic polarization are shown. For high speed and high accuracy in profile measurement, a line-scanning camera operating according to the TDI principle (time delay integration principle) is advantageously used. For elliptic polarization measurements, a suitable TDI line-scanning camera is preferably used as well. These have four TDI blocks, each with four different polarization filters for the incident light. Alternatively, a corresponding area-scanning camera can be used, which has a group of polarization filters in place of the color Bayer template filter before the sensor. It is also possible to perform measurements using a camera with no filters or filters in only one polarization direction by changing the filters and performing measurements sequentially, although with a considerable loss of speed.

[0073] A possible preferred improvement is the use of multiple wavelengths. This preferred structure is advantageous for profile heights exceeding half the wavelength of the light used. It has been recognized within the scope of this invention that using multiple wavelengths is also advantageous for flatter profiles because of the effective material parameters (n) used to determine the alternative layer. eff and k eff ) and layer thickness d eff The number of elliptically polarization measurements will inevitably increase accordingly. This adds two new, undetermined, unknown material parameters for each additional wavelength of light used (because n... eff and k eff (While wavelength-dependent), the effective layer thickness is the same for all wavelengths. Therefore, the ratio of independent measurements to the number of unknown parameters is improved, and this correspondingly contributes to the robustness of the measurement.

[0074] The following text will explain in more detail the correction of interferometry, which is the use of parameters obtained from elliptic polarization measurements to correct the interferometric measurements.

[0075] The reflection of light at the surface of an object in the interferometry method produces a change in the amplitude of the reflected light, described by the actual reflection coefficient R, and a change in the phase shift of the light wave. The phase shift described. The phase shift occurs due to interference effects at the interface located below the surface. The signal S reaching the interferometric detector can be expressed as:

[0076]

[0077] Here, q represents the intensity of the light source; R represents the actual reflectance of the surface of the object being measured at wavelength λ; γ represents the instrument constant, which is derived from the detector sensitivity and the absorption and reflection losses at the optics; and z r and z s This indicates the distance between the interferometer's beam splitter (beam splitter mirror) and the reference mirror (the mirror that reflects the reference beam at that location) or the surface of the object being measured.

[0078] When using substitution:

[0079] Δ z =z r -z s (7)

[0080] The parameter Δz (the fixed distance between the sample surface profile and the reference mirror with respect to the beam splitter) can be expressed as:

[0081]

[0082] Therefore, in addition to the measured signal strength S, the parameters to be determined by measurement are the actual reflection coefficient R at the surface of the measured object and the possible phase shift.

[0083] The determination of correction parameters by elliptic polarization measurement will be explained in more detail below.

[0084] In order to express the reflection used to solve the so-called interference formula (8) as the real reflection amplitude R and phase shift It should be assumed, for example, according to Figure 3 A model layer system that has a known basis L2(n2, k2 or n subst k subst Layer L1 and thickness d1 (=n) described by n1 and k1 on the surface eff k eff d eff In this diagram, air is assumed to be the environment (n0 = 1 and k1 = 0) without limiting its generality. Of course, other materials with corresponding parameters can also be used (in order to potentially take advantage of the immersion effect), such as water or oil.

[0085] The phase factor β for the optical path difference of a wave that has passed through layer L1 once is derived from the following formula:

[0086]

[0087] Where n1*=n1+i k1 is the complex refractive index of L1.

[0088] The angles of refraction at the interfaces of environment / layer L1 and layer L1 / substrate can be calculated using Snell's law of refraction.

[0089]

[0090] Conclusion:

[0091]

[0092]

[0093] For the phase factor, after substituting it into (9), we get

[0094]

[0095] The known Fresnel formula applies to reflection at the first interface:

[0096]

[0097]

[0098]

[0099]

[0100] Here, indices s and p denote the following components of the electrical amplitude of the light wave, whose polarization direction lies in the plane of incidence (p = parallel) or perpendicular to the plane of incidence (s = perpendicular), as in... Figure 4 As indicated by the indexes π and σ. It should be noted that all refractive indices marked with * are typically complex refractive indices. Therefore, for light waves ρ... i,k The reflection coefficient of the amplitude is usually also a composite parameter. In order to account for multiple reflections at the interface between medium 0 / 1 and 1 / 2, a geometric sequence is formed by using the amplitude of the reflections, whose limiting value can be directly calculated.

[0101] Therefore, for the entire reflection at the sample surface (i.e., at the interface of medium 0 / 1, and thus at the interface of the environment / substitute layer), the following can be derived:

[0102]

[0103]

[0104] These field amplitude reflection coefficients can now be directly used to calculate the light intensity at the detector of a radiometric ellipsometer (see [link]). Figure 4 An ellipsometry includes a light source, a polarizer with an angle of α1, a sample (object) to be measured, an analyzer with an angle of α2, and a detector.

[0105] The intensity generated at the detector can be represented using the Jones-Formalismus model. In the Jones model, the polarization state of the electric field is generalized as a vector and the time component of wave propagation is suppressed. The electric field after the polarizer (without a time component) is given by the following formula:

[0106]

[0107] Here E i α1 is the electric field amplitude of the light source, and α2 is the angle of the polarizer.

[0108] Through the matrix:

[0109]

[0110] The reflection at the sample is described by two reflection coefficients calculated in formulas (18) and (19), which are the parallel field component and the perpendicular field component of the light wave reflected at the model system.

[0111] After rotating the system to the analyzer's coordinate system using the rotation matrix R, it can be used through matrix T. aTo describe the role of an analyzer that is assumed to be ideal:

[0112]

[0113]

[0114] In this case, the Jones vector at the detector is

[0115]

[0116]

[0117]

[0118] Therefore, the light intensity at the detector is:

[0119]

[0120]

[0121]

[0122] If we transform this formula into a specific elliptometric measurement case with α1 = 45°, and measure the intensity at four angles α2 = 0°, 45°, 90°, and -45°, the result can be written as:

[0123]

[0124]

[0125]

[0126]

[0127]

[0128] The angles α1 and α2 were chosen simply for ease of representation and calculation. In the context of this invention, five other suitable angles could also be chosen, which would allow for complete determination of the polarization state of the light wave at the detector.

[0129] The product in (30) to (33) I0 is the input intensity of the light source at the device, which corresponds to the output intensity at the light source. This input intensity can be measured in the device according to the invention by means of a preparatory measurement using an ideal mirror or a known material as the sample. Advantageously, this input intensity is additionally monitored by means of a monitoring measurement at the light source itself during the measurement of the unknown sample (object) to be detected. For this purpose, a monitoring diode, for example, is used at the laser used as the light source. Four formulas are derived for the reflectance coefficient:

[0130]

[0131]

[0132]

[0133]

[0134] The redundant fourth measurement can be used to improve stability:

[0135] I d,0 +I d,90 =I d,45 +I d,-45 (38)

[0136] Therefore, it is possible to determine the three unknown parameters n from four measurements (three independent intensity values ​​and one redundant value). eff k eff d eff The complex reflection coefficient ρ can be calculated from this parameter. i,k The input intensity is determined by reference measurement and in parallel by monitoring the light source.

[0137] The same Jones model can be used for interferometric measurement of the beam to describe the wave reflected by the sample. Components, namely the polarizer and analyzer, are omitted here. Furthermore, the parallel polarization component can be neglected because the dominant beam intensity falls perpendicularly onto the sample.

[0138] Therefore, for the returning wave, we can conclude:

[0139]

[0140] Here, the index "IF" represents the relationship of the interferometric beam. This index also represents the light wave at the detector, and index i represents the input light wave. The parameter R for searching the interferometric formula (8) can be determined from this:

[0141]

[0142]

[0143] The following results were obtained under the condition of perpendicular light incidence and air as the ambient medium:

[0144]

[0145]

[0146]

[0147] According to formulas (40) and (41), R and R are represented by n1, k1, and d1. If the three parameters of the effective substitution layer are determined by measuring four elliptic polarization values ​​according to formulas (34) to (37), then R and Therefore, after using (8), the distance difference between the reference mirror (reference mirror) and the sample surface and the beam splitter (beam splitter) can be calculated, taking into account the correction for material-related reflections at the sample (object). The height profile of the surface search is thus obtained.

[0148] When using only one wavelength for interferometry, as in the previously described solution according to the invention, the working range of the measured distance is limited to half the wavelength of the light used. Outside this range, the relationship between intensity and distance becomes ambiguous because the path difference between the reference beam and the object beam in cycles cannot be directly determined from the signal intensity alone. Conventional interferometers solve this problem by determining the reference distance during initialization and subsequent uninterrupted counting of the intensity cycles of the measurement. However, this is unsuitable for use with the line-scan interferometer according to the invention because a reference greater than half a wavelength parallel to the line-scan assembly of the interferometer must be provided, and a step height greater than half a wavelength is not allowed on the object being measured. At least the latter cannot be used, for example, for the measurement of contact structures on a wafer.

[0149] Conversely, by using two or more different wavelengths, the operating range can be extended to allow for explicit measurements without a starting reference and period counts. Different wavelengths have different periodicities. Therefore, explicit measurements can be performed in other ranges based on combinations of signals from the wavelengths used. The basis of this approach is described, for example, in "Multi-Wavelength Interferometry for Length Measurements Using Diode Lasers" by K. Meiners-Hagen et al.

[0150] Therefore, the directly sensing operating range A for the two wavelengths λ1 and λ2 is:

[0151]

[0152] If not only the total intensity curve of the synthesized wavelength Λ is considered, but also the two wavelengths λ1 and λ2 themselves are considered separately, the operating range A can be extended to many times the wavelengths λ1 and λ2 (precise phase separation method). The method shown in the aforementioned publications can be combined with the present invention by merging multiple wavelengths before they enter the component. Parameter group n eff k eff and d eff In this case, it becomes n eff,i k eff,i and d eff The index i refers to the wavelength used, because the refractive index and absorption are different for each wavelength and therefore must be determined for all wavelengths used.

[0153] Therefore, one aspect of the present invention considers the effects of simultaneously measuring interferometric measurement signals and elliptic polarization signals, as well as observing the sample material (the material of the object) and the layer structure in an effective three-layer system including the environment, the effective layer, and the substrate. Attached Figure Description

[0154] The present invention will now be described and explained in detail with reference to some selected embodiments and the accompanying drawings. In the drawings:

[0155] Figure 1 A schematic diagram of a system according to the present invention for interferometric distance measurement by simultaneously recording elliptic polarization data is shown.

[0156] Figure 2 A schematic diagram illustrating the principle of polarization ratio of light waves using a polarization ellipse is shown.

[0157] Figure 3 A schematic diagram of light reflection at a three-layer system consisting of an environment / transparent layer / substrate is shown;

[0158] Figure 4 A schematic diagram of a radiometric ellipsometric measurement system with a fixed polarizer, sample, and analyzer is shown.

[0159] Figure 5 A view showing the method steps for preparing an interferometric measurement system for interferometric distance measurement;

[0160] Figure 6 A view showing the method steps for performing interferometric distance measurements using an interferometric measurement system; and

[0161] Figure 7A schematic diagram of a system for interferometric distance measurement according to the present invention is shown, which simultaneously records elliptic polarization data using three optical wavelengths. Detailed Implementation

[0162] Firstly, according to Figures 1 to 4 and Figure 7 The measurement structure and optical path of the system according to the present invention are described, and then... Figure 5 and Figure 6 The present invention discusses a method flow for preparing a measurement using a preferred embodiment of the invention, and a method flow for performing the measurement.

[0163] Figure 1 and Figure 7 Two different embodiments of a system 1 according to the invention for contour measurement of the surface 40 of a flat object 20 of an unknown material are shown. System 1 includes an optical device 2 for contour measurement of the object surface 20, a light source 111 for generating a monochromatic light beam, and a movement unit 10 by which the object 20 can be moved relative to the device 2. Figure 1 The implementation of system 1 has only one light source 111, while according to Figure 7 System 1 includes three light sources 111. The apparatus 2 varies accordingly in different embodiments.

[0164] In both embodiments, apparatus 2 includes an interferometric measurement system 4 having a beam splitter 91, a reference mirror 60, and a detector unit 250 for receiving and evaluating an evaluation beam 490. Elliptic polarization measurement system 5 includes a polarizer 190 and an elliptic polarization sensor 220 for analyzing the polarization state of the received sensor beam 520. The evaluation unit determines the profile height of the object 20, such as the surface 40 of a wafer, from the results obtained using the interferometric measurement system 4 and the elliptic polarization measurement system 5. Here, the interferometric measurement results are corrected using correction parameters for elliptic polarization measurement to obtain reliable measurement values ​​that account for different and locally varying materials on the object surface.

[0165] A wafer or other flat object 20 moves sequentially relative to the system 1 and the camera of system 1 according to the invention. In this embodiment, the camera is fixed in position, and the object 20 passes beneath the camera by means of a moving unit 10. In another embodiment, the camera is preferably moved. In an alternative embodiment, relative movement is preferably distributed between the camera and the object 20, such that, for example, the camera performs movement in the axial direction, while the object 20 can move in a direction perpendicular to it. The movement between the camera and the object 20 is preferably continuous.

[0166] Figure 1The embodiment uses a laser diode 110 as a light source 111, a TDI line scan camera 251 (TDI = Time Delay Integral) as a detector unit 250 for the interferometric measurement signal (here, the evaluation beam 490), and a TDI multi-channel line scan camera 221 with an integrated polarization filter (not shown) as an elliptically polarized sensor 220 for the sensor beam 520 constructed as an elliptically polarized signal. However, other numbers of light sources 111 may also be used. Similarly, a combination of a wider bandwidth light source and a narrower bandwidth filter may be used as an alternative. A simple (multi-channel) line sensor (without TDI method) or an area scan camera or a group of multiple line scan cameras may also be used instead of a TDI sensor.

[0167] exist Figure 1 In this embodiment, light from the laser diode 110 is supplied to the measurement assembly (device 2) via an optical fiber coupler 100. The light is emitted from the optical fiber coupler 100 as a beam 84. Transmission via the optical conductor is only used to disconnect the laser 110 from the actual recording system, i.e., device 2. Instead of optical conductor transmission, the (laser) beam 84 can also be directly redirected to the recording unit (device 2) via suitable optics, which reduces losses but requires more adjustment work.

[0168] Beam 84 is formed into a parallel optical beam (beam 400) by beam-shaping optics 80, which has a cross-section matched to the recording surface. Beam 400 (optical beam) is split into two sub-beams by a beam splitter 82, which is constructed as a 50% semi-transparent mirror: an interferometric beam 410 and an elliptically polarized beam 420. The beam splitting does not have to be 50 / 50. Other ratios can also be used, and thus matched according to the sensitivity requirements of the optical path and the sensor.

[0169] The interferometric beam 410 is partially focused as a sub-beam by a beam splitter 91, which is a 50% semi-transparent mirror 90, and directed onto the surface 40 of the object to be measured via a lens 50. This sub-beam is the measurement beam 430. The remaining 50% of the interferometric beam 410 serves as a reference beam 440, passing through the semi-transparent mirror 90. The reference beam 440 then strikes a reference mirror 60, where it is reflected. The resulting, returning sub-beam (mirror beam 450) is reflected again by 50% at the semi-transparent mirror 90 and travels as a sub-beam 460 to the interferometric detector 252, i.e., the detector unit 250.

[0170] The measurement beam 430 directed onto the object surface 40 of the wafer is focused into the imaging area by lens 50. Due to reflection at the object surface 40, the measurement beam 430 is reflected back as the object beam 470 through lens 50. Half of the light from the object beam 470 passes through the translucent mirror 90 and continues as a beam 480 to the interferometric detector 252. The light from the two beams 460 and 480 interferes with each other to form an evaluation beam 490. The intensity of the evaluation beam 490 is now modulated, on the one hand, by the distance difference between the translucent mirror 90 and the reference mirror 60, and on the other hand, by the distance difference between the translucent mirror 90 and the object surface 40, through the interference of light waves. The evaluation beam 490 is focused as a beam 495 onto the TDI line scan camera 251 by the lens barrel optics 260. The detector unit 250 thus records the distance-modulated interferometric measurement signal.

[0171] An elliptically polarized beam 420 is focused as a beam 500 through two 100% mirrors 70 and 72, with an angularly adjustable polarizer 190, and then through a lens 50 at a moderate angle of incidence Φ onto the object surface 40. The light reflected from the surface 40 is focused as a beam 510 through the lens 50 and, via mirrors 74 and 76, as a sensor beam 520 to the barrel optics 230, and then focused onto a four-channel elliptically polarized sensor 220. In this example, the elliptically polarized sensor 220 is a four-channel TDI row sensor with polarizing filters pre-installed before each of the four TDI blocks. The polarizing filters are oriented in a suitable manner so that the polarization state of the incoming light (sensor beam 520) can be fully analyzed (e.g., at angles of 0°, 45°, 90°, and -45°). Therefore, four elliptically polarized signals can be recorded in the four detector elements forming the column of the elliptically polarized sensor 220, allowing the polarization state of the light entering in that column to be fully determined from these four signals.

[0172] In order to monitor the intensity of the light source 111, the laser output power is continuously controlled and measured by means of a monitoring diode 140.

[0173] exist Figure 7 The diagram illustrates an alternative arrangement according to the invention, in which three wavelengths are used to improve the robustness of the measurement and / or to extend the defined operating range. These three wavelengths are generated by lasers 110, 120, and 130 (laser diodes) and their intensity is monitored by means of monitoring diodes 140, 150, and 160. The beams from the individual laser diodes are focused, and preferably focused into beam 84, via a preferably spliced ​​fiber optic coupler 100. Figure 1 The layout is different here. Figure 7For interferometric measurements, a multicolor TDI line scan camera 255 is used to record signals separately for three wavelengths. Similarly, unlike other methods, the elliptically polarized signal used for measurement is separated by a separating element 600. The elliptically polarized signal is preferably decomposed into three wavelength portions by means of two dichroic mirrors 610 and 620, and is guided as beams 522, 524, and 528 through lens barrel optics 232, 234, and 230, and finally measured by means of elliptically polarization sensors 220, 222, and 224.

[0174] This diagram illustrates a general arrangement, which can of course be configured in various ways. Thus, for example, the order in which the beam is decomposed (610, 620) into wavelength portions and focused using the lens barrel optics (232, 234, and 230) can be reversed.

[0175] according to Figure 5 and Figure 6 A preferred embodiment of the method according to the invention is described, which preferably includes some preparatory steps. For the purpose of measurement, the invention provides the following preparatory steps (see...). Figure 5 ):

[0176] 1) Measure the dark signal when light source 111 is off to determine the dark signal of the detector;

[0177] 2) Adjust the light source 111 to the predetermined working brightness;

[0178] 3) Measure the reflection at 100% mirror of the sample to determine the instrument constant γ in formula (8);

[0179] 4) Measure the reflection at a 100% mirror or a known, homogeneous material, which serves as the sample, to determine the instrument constant γ for the elliptic polarization measurement system 5. e .

[0180] In order to preferably perform actual measurements using the system 1 according to the invention, the invention provides the following steps for each measurement point.

[0181] 1) Simultaneously measure the interferometric measurement signal, the monitoring signal, and the elliptic polarization signal at the desired measurement point on the surface of the object, i.e., receive and evaluate the evaluation beam 490 and the sensor beam 520.

[0182] 2) Calculate ρ using the four intensity values ​​measured from the elliptic polarization channel and the monitoring value from light source 111. s and ρ p ;

[0183] 3) Utilizing the incident angle Θ0 of the elliptic polarization channel and the predetermined parameter n of the substrate of object 20. Subst k Subst , by ρs and ρ p Calculate the effective parameter n eff k eff and d eff ;

[0184] 4) From n eff k eff and d eff Calculate ρ of perpendicular light incident in the interference channel s ;

[0185] 5) ρ of perpendicularly incident light s Calculate R and

[0186] 6) Calculate the distance difference Δz between the reference mirror and the sample surface and the beam splitter 91 (beam splitter);

[0187] 7) Calculate the profile height at the measurement point (refer to the distance of the reference mirror).

[0188] After measurements are performed using device 2 and the interferometric measurement system 4 and elliptic polarization measurement system 5 included therein, signal evaluation is performed in evaluation unit 700.

[0189] To simplify the illustrations, the calculation of the surface profile will be explained using only one wavelength in the following text. When using multiple wavelengths, the formulas are applied to each wavelength in a similar manner. Similarly, only one pixel group, i.e., one line scan position of the advantageously used line scan camera, is shown. These formulas are applied to each pixel group at the corresponding line scan position in a similar manner. Here, a pixel group represents pixels belonging to the same line scan position in both the interferometric camera and the elliptical polarization camera.

[0190] Therefore, multiple height points of the profile can be simultaneously determined based on the sensor size and arrangement. For a currently available line scan camera with 16,384 points per row, this means that there are 16,384 height values ​​per readout cycle of the camera. Furthermore, in the apparatus and system 1 described herein, the line scan camera can move continuously relative to the wafer (object). Depending on the camera's cycle speed, multiple rows with height information are acquired per unit time. Thus, at a cycle of, for example, 100 kHz for the proposed multi-channel TDI line scan camera 221, more than 1600 Mio. height values ​​per second can be obtained. Such cameras are offered by different manufacturers (e.g., Vieworks and Dalsa Teledyne). Using this type of TDI multi-channel line scan camera is a particularly suitable variant because very high measurement speeds can be achieved. These cameras contain multiple (typically four) TDI blocks in a single camera, which can operate and read out simultaneously. When using such a camera, the wavelengths used for each row on the wafer surface are recorded sequentially. This means that while the first TDI block is recorded in the row region at λ1, the second TDI block identifies the signal at λ2, and the third TDI block identifies the signal at λ3. In principle, this temporal stagger is not important for the calculations described later. It is sufficient to distribute and evaluate the obtained signal maps in a phased manner over time (corresponding to the spatial stagger of the TDI blocks). With this arrangement, high signal quality (and correspondingly high resolution and robust measurements) can be combined particularly well with high speed. Alternative arrangements will be explained below.

[0191] Figure 5 and Figure 6 The process of measurement and its preparation is explained. In order to accurately determine the height value, it is necessary to prepare for the measurement by measuring dark signals and determining the transfer function of optical devices and sensors.

[0192] In dark signal measurement, the signal d is measured at each camera pixel y of the two line scan cameras (elliptic polarization sensor 220 and detector unit 250) with the light source off. The so-called dark noise of the camera is thus determined, representing the offset for each further measurement and subtracted from the signal. This is performed on the two sensor assemblies 220 (index z = 1 to 4) and 250 (index z = 5).

[0193] d yz = Signal read at pixel y of sensor z minus dark value

[0194] To determine the optical and electrical transfer functions of the arrangement, a bright signal h is measured using a 100% mirror. For this purpose, the wafer surface or object surface 40 is replaced with a plane mirror with known reflective properties. Since the signal h at each sensor pixel is determined by the intensity of the light source 111 (signal value q), the transfer function M, the reflectivity R of the mirror, and the path difference l between the two interferometric beams 460 and 480 (impacting the detector unit 250 = sensor, z = 5),... z The transfer function M for each pixel y in each sensor row of sensor row z can be determined by using the dark signal d, provided that the values ​​h, q, R, and d are known.

[0195] For the interferometric sensor, i.e., detector unit 250 (z=5), it is the path difference l z The transfer function is a function of the sample height. For the elliptic polarization sensor 220 (z = 1 to 4), M is independent of the sample height as long as the sensor moves within the focal region. The transfer function is typically different for each pixel y and each sensor z. This transfer function is determined by the sensitivity of each pixel, through illumination, material properties, coatings, and imaging errors of the optics.

[0196] In order to check the output intensity of the laser source 111 and incorporate it into the calculation as a correction or reference value, the signal q of the monitoring diode 140 (on the side of the laser away from the output end) that is typically installed in each laser module 110 is used directly.

[0197] For the interferometric sensor (detector unit 250), the signal h for the bright reference measurement is:

[0198] h yz,href (l z )=q href *M yz (l z )*R href +d yz (46)

[0199] in:

[0200] h yz,href (l z = as path difference l z The measured value at pixel y of the sensor z (interferometric sensor 250 [z=5], elliptic polarization sensor 220 [z=1…4]) is a function of the reference measurement.

[0201] q href =Light intensity of the laser source during the reference measurement period

[0202] M yz (l z = as path difference lz The transfer function of the function at pixel y of sensor z.

[0203] R href = The reflectivity of a mirror in a reference measurement (known material)

[0204] Independent variable l z This represents the path difference between the two interfering beams 460° and 480° (impacting sensor 250, z=5) relative to each other. Here, the reflectivity R can be used as the true value (reflectance). The mirror is used because it is advantageous for reference measurements to select a mirror that will not produce a phase shift due to the layer reference.

[0205] If, for example, a mirror with a protective coating is used to improve long-term stability, the phase shift caused by the protective coating has no effect, because the reference measurement is only used to determine the maximum value of the transfer function.

[0206] Applicable to the elliptic polarization sensor 220:

[0207] h yz,href =q href *M yz *R href +d yz (47)

[0208] For elliptical polarization channels, it is important to note that when determining the transfer function, phase shifts at the mirrors, such as those caused by possible protective coatings, must be considered in the calculations.

[0209] To determine the relationship between the wafer surface and the sensor assembly z sample The distance decomposes the transfer function M(l) into a non-interference factor M. max And interference effects. Factor M max This is obtained by determining the maximum value for each transfer function M(l). In this case, the intensity modulation caused by interference can be directly determined from the values ​​measured during the run.

[0210] The transfer function is applicable to the sensor of detector unit 250:

[0211]

[0212] And for the elliptic polarization sensor 220, M yz It is constant within the focal region.

[0213] in:

[0214] M yz (l z = as path difference l zThe transfer function of the function at pixel y of sensor z.

[0215]

[0216] z mirror = Distance between beam splitter and reference mirror in interferometry

[0217] z sample =Distance between beam splitter and wafer surface

[0218] Therefore, for the interferometric measurement sensor (detector unit 250), the following conclusions can be drawn:

[0219]

[0220] in:

[0221]

[0222] M is determined in the following manner. max That is, continuously changing the distance z of the mirror used for measurement. sample And so on, running for a complete wavelength cycle, recording the signal and identifying the maximum value.

[0223] During the measurement operation on the unknown wafer surface (object surface 40) to be probed, the sensor signal i of the elliptic polarization camera sensor 220 [z = 1…4] is now being recorded simultaneously. y1 to i y4 The sensor signal i of the interferometric measurement sensor (detector unit 250) [z=5] y5 The output intensity q of the laser source 111 is measured using a monitoring diode 140.

[0224] For signals at the sensor block in the camera (elliptic polarization sensor 220), applicable to:

[0225] i yz =q*M yz *r yz,wafer +d yz (50)

[0226] in:

[0227] i yz = Elliptic polarization measurement at pixel y of sensor z / Interferometric measurement

[0228] q = incident light intensity

[0229] M yz =Transfer function at pixel y of sensor z

[0230] r yz,wafer= The reflectance of the wafer at the position of pixel y with polarization angle z

[0231] d yz = The dark signal value read at pixel y of sensor z

[0232] Here, the reflection coefficient r is obtained from the material properties of the sample and under the condition of the correspondingly selected polarization angle, according to formulas (30) to (33). y,wafer Now, parameters describing the sample material are calculated from these measurements using formulas (34) to (37) and (13) to (19). n 1,y =n eff,y k 1,y =k eff,y and d 1,y =d eff,y The intensity value I, transformed by q = I0, is used in the formula system (34) to (37). d,z =I d,Polarisationswinkel :

[0233]

[0234] The material values ​​thus determined can now be used, according to formula (43), to determine the complex reflection coefficient ρ of perpendicular light incident at the position y of pixel y of the interferometric sensor [z=5]. s,y5 Finally, according to formulas (40) and (41), ρ can be derived from... s,y5 Determine the (true) reflectance R for each pixel location y. y and phase offset φ y .

[0235] For the intensity measured at detector unit 250, the following is obtained:

[0236]

[0237] Therefore, the search parameter Δz can be calculated. y =z mirror,y –z sample,y =The difference between the distance from the sample surface 40° to the (beam splitter) mirror 90° and the distance from the reference mirror 60° to the (beam splitter) mirror 90°:

[0238]

[0239] If as Figure 7 If multiple wavelengths, such as preferably three, are used to expand the measurement area and / or improve the robustness of the measurement, then the calculations shown here must be performed separately for each wavelength. Here, the calculations are first performed using formula (51) with the corresponding half-wavelength λ. x Determine the distance difference Δz for each unit.xy , where x is the wavelength index. In this case, according to the so-called precise phase component method, the distance difference can be expressed as the integer part δ. x and fractional remainder f x sum:

[0240]

[0241] By finding out three times δ of the integer part x (Regarding this, the Δz) yx (Minimum average deviation from the corresponding average value) from Δz yx Determine the required distance difference Δz y Next, calculate the average of the remaining three fractional parts.

[0242] By appropriately selecting wavelengths λ1, λ2, and λ3 for x = 1, 2, and 3, a working range of 0.5 mm can be effortlessly established, and the intensity measurement value can be tripled by i. xy5 Clearly related to the distance difference Δz y This correspondence is sufficient to satisfy various contour measurement tasks. This point is discussed at the beginning by K. Meiners-Hagen and R. This is explained in detail in the publications of F. Pollinger and A. Abou-Zeid. In the arrangements disclosed therein, for example, wavelengths of 532nm, 632nm, and 780nm were used to achieve an operating range of 0.6mm with a clear correspondence of distance differences.

[0243] Other preferred embodiments of the invention are described below. These alternative preferred embodiments are intended to be part of the invention but are not intended to limit its general applicability.

[0244] Therefore, the use of TDI technology in the proposed arrangement is only for improving the signal-to-noise ratio and is not essential to the principles of the present invention. Thus, for simpler measurement configurations and requirements, a component consisting of four independent line sensor cameras or four independent single-channel TDI line sensor cameras can be selected instead of the TDI multi-channel camera line sensor used as elliptic polarization sensors 220, 222, and 224.

[0245] Alternatively, a conventional area scan camera or a TDI single-channel line scan camera operating in area readout mode can be used. In this arrangement, the clock frequency is correspondingly reduced to, for example, a clock frequency of 1 kHz (as described for the area readout mode of a TDI single-channel line scan camera), and more than 16 Mio. height values ​​per second can still be obtained.

[0246] Understandably, when using a surface scan camera or a TDI single-channel line scan camera in surface readout mode, the three wavelengths λ1, λ2, and λ3 are imaged as different rows x = 1, 2, and 3 on the camera sensor. The correspondence between the signal maps obtained from observing the same point on the wafer surface 40 at the three wavelengths is determined through spatial correspondence correlation of the camera sensor region. For the above principle evaluation, it is not important whether the correspondence is performed spatially (for surface sensors) or in a phased manner over time (for line sensors).

[0247] It should be noted that for high-resolution lateral applications (e.g., in the range of a few μm), using multiple cameras necessitates lateral and rotational corrections to the recorded signal map, as it is difficult to mechanically adjust the entire measurement assembly to a camera offset of less than 1 μm. This mathematical correction is typically achieved by recording a reference pattern from which the precise location on the wafer observed by each pixel can be determined.

[0248] In other possible variations of the implementation, the number of wavelengths used can be matched to the required operating range. For particularly small operating ranges, implementations using only one wavelength are already possible, or for small ranges, two wavelengths may be used. For larger operating ranges or to improve reliability through measurement redundancy, extending to three or more wavelengths is suitable, which can be achieved in particular by the aforementioned multi-block TDI camera (a camera with seven TDI blocks has already been described).

[0249] Illumination can be adequately achieved using monochromatic light sources with continuous radiation. For this purpose, lasers are as suitable as other broadband beam sources, which can be combined with corresponding interference filters. The only condition is that the coherence length of the light used is sufficiently large for the desired operating range.

[0250] When using different wavelengths, radiation can be converged using spliced ​​optical waveguides or dichroscopes. A spliced ​​optical waveguide is an optical waveguide that is partially fused together or guided close to each other, allowing light to pass from one waveguide to another. In a dichroscope, only light within a specific wavelength range is reflected, while at least one other wavelength range is transmitted. The size of such a mirror can be set for use at different angles. The most common angle is 45°. Therefore, in a proper arrangement, the reflected and transmitted light can have the same propagation direction.

[0251] Instead of a dichroic mirror, one or more prisms, gratings, or other spectral dispersion measures can also be used to separate beams of light at different wavelengths in an elliptically polarized optical path. Spatial separation allows signals to be recorded simultaneously, which enables high measurement speeds. However, temporal separation can also be achieved by performing measurements sequentially at different wavelengths, which requires longer measurement times.

[0252] Similarly, a bonded prism with a square cross-section can be used instead of beam splitter 91 or mirror 90, with a partially mirrored layer disposed on the bonded 45° surface. When using such a prism, the reference arm can be generated by vapor-depositing the mirrored portion directly on the outer surface of the prism. This arrangement results in reduced interference contrast due to the significant difference in dispersion between the reference arm and the measuring arm of the interferometer; however, it is advantageous for simple requirements due to its robustness.

[0253] Specifically, the optical components used to generate the interference can be specified to include a semi-transparent mirror that reflects a portion of the radiation in the direction of a first surface facing the object surface, while allowing another portion of the radiation to pass through. The reflected radiation can then be combined with the already transmitted beam. This can be achieved, for example, by having the transmitted beam reflect back to itself on the mirror and couple into the reflected beam at the semi-transparent mirror. Such a component is called a Michelson interferometer. However, other components can also be used, such as the so-called Mach-Zehnder interferometer, where the incident light is split and the sub-beams are recombined after reflection at the sample surface.

[0254] Other preferred variations have been described in patent application EP19188318, the entire contents of which are incorporated herein by reference.

[0255] literature

[0256] (1) "Multi-Wavelength Interferometry for Length Measurements Using Diode Lasers" K. Meiners, Hagen, R. F. Pollinger and A. AbouZeid, *Measurement Science Review*, Volume 9, Chapter 3, November 1, 2009

[0257] (2) "Handbook of Elliposmetry", Harland G. Tompkins and Eugene A. Irene, Springer, 2005, ISBN 0815514999

Claims

1. A device (2) for contour measurement of the surface (40) of a flat object (20) of an unknown material, comprising: A beam splitter (82) for splitting the beam (84, 400) of the light source (111). An optical interferometric measurement system (4) and an elliptic polarization measurement system (5) are used to simultaneously measure the measurement area on the surface (40) of the object. Evaluation Unit (700) in: a) The optical interferometric measurement system (4) includes: A beam splitter (91) for splitting the interferometric measurement beam (410) of the light source (111) into a reference beam (440) and a measurement beam (430). Reference mirror (60) for reflecting the reference beam (440); Detector unit (250) for receiving and evaluating the evaluation beam (490) of the optical interferometry system (4); in, The measurement beam (430) is directed onto the measurement area on the object surface (40) for reflection, and after reflection, it is directed to the beam splitter (91) as the object beam (470). The reference beam (440) is reflected at the reference mirror and guided as a mirror beam (450) to the beam splitter (91). The object beam (470) and the mirror beam (450) interfere after striking the beam splitter (91) and are supplied to the detector unit (250) as an evaluation beam (490) for evaluation. b) The elliptic polarization measurement system (5) includes: A polarizer (190) is used to polarize an elliptically polarized beam (420) and to transmit the elliptically polarized beam (420) onto the measurement area on the surface (40) of the object. Elliptic polarization sensors (220, 222, 224) with polarization filters configured to analyze the polarization state of a received sensor beam (520); The elliptically polarized beam (420) is guided to the measurement area on the surface of the object (40) and, after being reflected at the measurement area on the surface of the object (40), strikes the elliptically polarized sensor (220, 222, 224) as a sensor beam (520). c) The beam splitter (82) is configured to split the beam (84, 400) of the light source (111) into the interferometric beam (410) and the elliptically polarized beam (420). as well as d) The evaluation unit (700) is configured to simultaneously process the evaluation beam (490) evaluated in the detector unit (250) and the sensor beam (520) received in the ellipsoidal polarization sensors (220, 222, 224) and determine the height of the contour in the measurement region on the object surface (40) without using material parameters of the unknown material of the object (20) to be measured and / or without modeling the material parameters of the unknown material of the object (20) to be measured. In the evaluation unit (700), the measurement results of the detector unit (250) are corrected based on the following parameters: (i) The effective refractive index of the object (20) to be measured using the elliptic polarization measurement system (5), wherein the alternative layer integrates the effective optical effects of one or more layers of the object (20); (ii) The effective absorption coefficient of the surrogate layer of the object (20) to be measured, as determined using the elliptic polarization measurement system (5); and (iii) The effective layer thickness of the replacement layer, The effective refractive index, effective absorption coefficient, and effective layer thickness of the substitute layer are used to calculate the actual reflection coefficient and phase shift of the object beam (470).

2. A system for contour measurement of the surface (40) of a flat object (20) of an unknown material, comprising: A light source (111) for generating monochromatic light beams (84, 400). The apparatus (2) according to claim 1; and A moving unit (10) for performing relative movement between the device (2) and the object (20) to be measured.

3. The system according to claim 2, characterized in that, The moving unit (10) moves the device (2) and / or the object (20).

4. The system according to claim 3, characterized in that, A retainer is provided for receiving the object (20) when the object (20) is moved, and the retainer is moved by the moving unit (10).

5. The apparatus according to claim 1 or the system according to claim 2, characterized in that, The beam splitter (91) and / or the beam splitter (82) are semi-transparent mirrors.

6. The apparatus according to claim 1 or the system according to claim 2, characterized in that, A lens (50) is provided in the optical path, which deflects the light beam (84, 400) so that the polarized elliptically polarized light beam (420) strikes the measurement area on the surface of the object (40) at a predetermined angle and the measurement light beam (430) strikes the measurement area on the surface of the object (40) at a right angle.

7. The apparatus according to claim 1 or the system according to claim 2, characterized in that, Mirrors (70, 72, 74, 76) are provided in the optical path of the elliptically polarized beam (420), which cause the elliptically polarized beam (420) to be vertically turned toward the polarizer (190) and / or cause the sensor beam (520) to be vertically turned toward the elliptically polarized sensor (220, 222, 224).

8. The apparatus or system according to claim 7, characterized in that, A lens optics (260) is provided in the optical path before the detector unit (250), the lens optics focusing the evaluation beam (490) onto the detector unit (250), and / or a lens optics (230, 232, 234) is provided in the optical path before the ellipsoidal polarization sensors (220, 222, 224), the lens optics focusing the sensor beam (520) onto the ellipsoidal polarization sensors (220, 222, 224).

9. The apparatus according to claim 1 or the system according to claim 2, characterized in that, The light source (111) is a laser diode (110, 120, 130).

10. The apparatus according to claim 1 or the system according to claim 2, characterized in that, The light source (111) has monitoring diodes (140, 150, 160) to determine and monitor the output intensity of the light source (111).

11. The system according to claim 2, characterized in that, The system includes multiple light sources (111) for generating monochromatic light beams (84, 400), the light beams (84, 400) of the multiple light sources having different wavelengths.

12. The system according to claim 11, characterized in that, The beams (84, 400) of the plurality of light sources (111) have different wavelengths, and the wavelength of one beam (84, 400) is not an integer multiple of the wavelength of the other beam (84, 400).

13. The system according to claim 12, characterized in that, The wavelength of none of the beams (84, 400) is an integer multiple of the wavelength of the other beams (84, 400), and the beams (84, 400) of the plurality of light sources are focused by means of beam shaping optics (80).

14. The system according to claim 13, characterized in that, The beams (84, 400) of the plurality of light sources (111) are focused by means of beam shaping optics (80).

15. The system according to claim 11, characterized in that, The system (1) includes a plurality of ellipsoidal polarization sensors (220, 222, 224) for determining the polarization state of the received sensor beam (520).

16. The system according to claim 15, characterized in that, Elliptic polarization sensors (220, 222, 224) are provided for each wavelength of the existing beam.

17. The apparatus according to claim 1 or the system according to claim 2, characterized in that, The device (2) includes an optical separation element (600) for separating the sensor beam (520) in a wavelength-specific manner, the optical separation element being arranged such that the separation of the sensor beam (520) is achieved before the sensor beam strikes the elliptic polarization sensor (220, 222, 224).

18. The system according to claim 2, characterized in that, The elliptic polarization sensors (220, 222, 224) include multiple polarization filters, enabling the recording and evaluation of multiple polarization directions for analyzing the polarization state of the received sensor beam (520).

19. The system according to claim 18, characterized in that, The elliptic polarization sensors (220, 222, 224) include four polarization filters, which record and evaluate the four polarization directions in order to analyze the polarization state of the received sensor beam (520).

20. The system according to claim 18, characterized in that, Simultaneously record and evaluate multiple polarization directions.

21. The apparatus or system according to claim 17, characterized in that, The optical separation element (600) is a dichroic mirror (610, 620).

22. The apparatus according to claim 1 or the system according to claim 2, characterized in that, The detector unit (250) includes a time-delay integration camera.

23. The apparatus or system according to claim 22, characterized in that, The detector unit (250) includes a color time-delay integration camera.

24. The apparatus according to claim 1 or the system according to claim 2, characterized in that, The detector unit (250) includes a time-delay integral multi-channel line scan camera.

25. The apparatus according to claim 1 or the system according to claim 2, characterized in that, The elliptic polarization sensors (220, 222, 224) include time-delay integration cameras.

26. The apparatus according to claim 1 or the system according to claim 2, characterized in that, The elliptic polarization sensors (220, 222, 224) include a time-delay integral line scan camera.

27. The apparatus according to claim 1 or the system according to claim 2, characterized in that, The elliptic polarization sensors (220, 222, 224) include a multi-channel time-delay integral line scan camera.

28. The apparatus according to claim 1 or the system according to claim 2, characterized in that, The elliptic polarization sensors (220, 222, 224) include a four-channel time-delay integral line scan camera, which includes four polarization filters.

29. A method for sensing the surface profile of an object surface (40) of an object having an unknown material by means of interferometric measurement and simultaneously by means of elliptic polarization measurement, comprising the following steps: A monochromatic light beam (84, 400) is emitted in the direction of the optical device (2) by means of the light source (111); The beam (84, 400) of the light source (111) is split into an interferometric beam (410) and an elliptically polarized beam (420) by means of a beam splitter (82). Interferometric measurement is performed at the measurement area on the surface (40) of the object using the interferometric beam (410) with the aid of an optical interferometric measurement system (4); The monitoring signal of the light source (111) is measured to determine the output intensity of the light source (111); Simultaneously with the interferometric measurement, the ellipsometric measurement is performed at the measurement area on the surface (40) of the object using an optical ellipsometric measurement system (5); The correction parameters are calculated using the results of the ellipsometric measurement and taking into account the output intensity of the light source (111); The measured values ​​obtained by interferometry are corrected using the aforementioned correction parameters; Taking into account the corrected measurements, the height of the contour of the measurement area at the surface of the unknown material of the object (20) is determined by means of an evaluation unit without the need to use the material parameters of the unknown material of the object (20) to be measured and / or without modeling the material parameters of the unknown material of the object (20) to be measured. The measured values ​​are corrected based on the following correction parameters: (i) The effective refractive index of the object (20) to be measured using the optical ellipsometric measurement system (5), wherein the alternative layer integrates the effective optical effects of one or more layers of the object (20); (ii) The effective absorption coefficient of the surrogate layer of the object (20) to be measured, as determined using the optical ellipsometric measurement system (5); and (iii) The effective layer thickness of the replacement layer, The effective refractive index, effective absorption coefficient, and effective layer thickness of the substitute layer are used to calculate the actual reflection coefficient and phase shift of the object beam, wherein the object beam is the beam obtained by guiding the measurement beam in the interferometric measurement beam (410) to the measurement area on the object surface (40) for reflection.

30. The method according to claim 29, characterized in that, The object contains exactly the material being measured at each measurement point.

Images