Arrangement for determining the depth of depressions formed in surfaces of a substrate on which at least one layer of a material different from the substrate material is formed

Abstract

Arrangement for determining the depth of depressions formed in surfaces of a substrate on which at least one layer of a material different from the substrate material is formed, in which several detectors designed for spatially resolved spectral analysis of electromagnetic radiation within a wavelength interval, arranged in a row or a row and column arrangement and the detectors are connected to an electronic evaluation unit and arranged so that electromagnetic radiation emitted by a broadband radiation source is diffusely directed onto a surface and impinges on the detectors either after reflection at the surface of the substrate or a layer formed on the surface of the substrate and / or after radiating through the substrate transparent to the electromagnetic radiation or the substrate with at least one layer present on a surface, wherein the irradiation is carried out in such a way that a laterally and temporally homogeneous intensity of the electromagnetic radiation is maintained on a surface from which the electromagnetic radiation is reflected or through which the surface is transmitted, or electromagnetic radiation emitted by the substrate and / or at least one layer formed on the substrate strikes the detectors as a result of an energy input and the electronic evaluation unit is designed such that the spatially and wavelength-resolved measurement signals of the detectors can be detected within a wavelength interval for individual locations on a surface of the substrate or a layer formed on a substrate, and measurement signals detected at several positions can each be assigned to a sub-area of the detected area, and Based on these spatially and wavelength-resolved measurement signals, a comparison is made with previously determined measurement values stored in an electronic memory, which can be carried out in an analogous manner on comparison samples produced with the same materials and the same layer thicknesses and in which depressions of a defined depth have been formed beforehand, whereby If there is sufficient agreement between spectra recorded with the arrangement on a substrate and spectra of a comparison sample, the depth of a depression can be determined.

Description

[0001] The invention relates to an arrangement for determining the depth of depressions formed in the surfaces of a substrate on which at least one layer made of a material different from the substrate material is formed. These substrates can have only one layer on one surface or can have multiple layers one above the other on the surface of a substrate. All of the layers can be made of different materials. The depth of depressions that are individually distributed over the surface and have a wide variety of geometries of their internal free cross-sections, such as circular (blind holes) or polygonal shapes, can be determined. However, depressions can also be trench-shaped, elongated, straight or curved.

[0002] A trench-shaped formation of depressions is required in the production of organic photovoltaic elements. In a multilayer system suitable for photovoltaic applications, formed on a substrate, trench-shaped depressions (so-called scribes) are typically formed to electrically interconnect individual photovoltaic cells in a specific manner, so that a desired electrical voltage can be achieved during use. Through locally targeted material removal of the multilayer system on at least one layer of the multilayer system, defined electrical insulation can be achieved with the trench-shaped depressions formed.

[0003] In this application, the trench-shaped depressions are preferably formed with a laser beam. This has the following advantages: ◯ Realization of very narrow “isolation trenches”, ◯ Thus low “consumption” of active usable PV area ◯ higher geometric form factors can be achieved.

[0004] However, it is problematic to maintain an optimal depth for the trench-like depressions. The depressions must not be too deep to avoid damaging an underlying layer (e.g., a barrier layer). A film serving as a substrate would then no longer be stable over the long term. However, depressions should not be too shallow either. If the depressions are not deep enough, adequate segmentation of the areas of individual photovoltaic elements cannot be achieved, which in turn would lead to incorrect electrical voltages and significant losses due to increased electrical resistance, as well as lower electrical voltage.

[0005] In order to be able to check that an optimal depth (the quality of the scribes) is maintained, simple test methods should be available that can, if possible, also be implemented as inline-capable test systems and therefore continuously record the current scribe depth during the process and can be used to control the material removal process based on their measurement signals.

[0006] Optical microscopes, scanning electron microscopes (SEM) or atomic force microscopes (AFM) are commonly used for this purpose.

[0007] Camera-based systems, such as those described in DE 10 2006 000 946 A1, are also possible. However, these are intended only for optical inspection of periodic structures. Using high-resolution cameras, a maximum resolution of approximately 6 µm can be achieved. However, only information about lateral geometric parameters (shape and width of the depressions, relative position of the depressions) can be obtained, but no chemical or depth information.

[0008] Another possible option is to record the emission spectrum of the material (vapor) removed by the laser ablation process. This can be used to determine the material quality and to assign it to the current "depth" of a depression or to the current ablated layer. However, the disadvantage is that this is not possible in all process pressure ranges. Sufficient laser power / energy is also required to excite the emission. This is usually significantly greater than that of the laser beam required to create the trench-like depressions. This also does not allow the actual depth to be determined, but only the current material area in which the laser ablation is currently being performed.

[0009] For example, DE 103 19 843 A1 describes a method for determining the depth of a structure buried in a semiconductor wafer.

[0010] In “Detection of physical defects in solar cells by hyperscanning imaging technology”; Optics & Laser Technology, Vol. 42; No. 5; September 6, 2010; pages 1010 and 1013, Quing Li et al. explain ways of detecting defects.

[0011] DE 10 2012 102 826 A1 relates to a device and a method for determining a depth of a region having a high aspect ratio and protruding into a surface of a semiconductor wafer.

[0012] DE 10 2014 106 974 A1 concerns a hyperspectral camera with spatial and spectral resolution. The camera is said to have two modules: one module is a high-resolution imaging module, and a second module is a spectral resolution module.

[0013] DE 10 2014 009 372 A1 describes an arrangement for determining properties and / or parameters of a sample and / or at least one layer formed on a surface of a sample.

[0014] It is therefore an object of the invention to provide possibilities for determining the depth of depressions formed on substrates on which at least one layer, preferably several layers, are formed, wherein the determination is carried out in a short time for large areas and preferably in such a way that a control of the process of forming depressions is possible.

[0015] According to the invention, this object is achieved by an arrangement having the features of claim 1. Advantageous embodiments and further developments of the invention can be realized with features designated in subordinate claims.

[0016] In the arrangement according to the invention for determining the depth of depressions formed in surfaces of a substrate on which at least one layer made of a material different from the substrate material is formed, a plurality of detectors which are designed for the spatially resolved spectral analysis of electromagnetic radiation within a wavelength interval are arranged in a row or a row and column arrangement.

[0017] The detectors are connected to an electronic evaluation unit and arranged in such a way that electromagnetic radiation emitted by a broadband radiation source impinges on the detectors either after reflection at the surface of the substrate or a layer formed on the surface of the substrate and / or after irradiation through the substrate transparent to the electromagnetic radiation or the substrate with at least one layer present on a surface, wherein the irradiation takes place in such a way that a laterally and temporally homogeneous intensity of the electromagnetic radiation is maintained on a surface from which the electromagnetic radiation is reflected or through which the surface is transmitted.

[0018] As an alternative to irradiation, electromagnetic radiation emitted by the substrate and / or at least one layer formed on the substrate can be used to impinge on the detectors as a result of energy input. Further possibilities for this are explained below.

[0019] The electronic evaluation unit is designed such that the spatially and wavelength-resolved measurement signals of the detectors can be detected within a wavelength interval for individual location points on a surface of the substrate or a layer formed on a substrate, and measurement signals detected at several positions can each be assigned to a sub-area of the detected surface (hypercube).

[0020] Based on these spatially and wavelength-resolved measurement signals, a comparison is made with previously determined measurement values stored in an electronic memory. These values were obtained analogously on reference samples manufactured with the same materials and the same layer thicknesses, in which depressions of defined depths had been previously formed. This allows the depth of a depression to be determined if there is sufficient agreement between the spectra recorded with the arrangement on a substrate and the spectra of a reference sample.

[0021] For the individual spectra, the respective intensities of individual wavelengths can be taken into account.

[0022] The irradiation of the surface can be carried out at least at an angle in the range 0 ° to < 90 ° with respect to the normal of the surface of the irradiated surface of the substrate.

[0023] The detection and evaluation can advantageously be carried out using at least one polarizer with at least one defined known polarization plane with respect to the plane of incidence or combinations thereof.

[0024] The detectors and the substrate can be moved relative to each other along at least one axis and preferably at a constant distance from each other.

[0025] The radiation source can comprise optical elements that shape electromagnetic radiation. However, it can also be a radiation source that emits electromagnetic radiation diffusely onto the surface, which is arranged, in particular, within a hollow body (e.g., an integrating sphere). Particularly preferably, a diaphragm is arranged in front of the detectors in the beam path of the electromagnetic radiation to prevent the incidence of scattered electromagnetic radiation.

[0026] The row and column arrangement of detectors with optical elements and evaluation electronics can be formed with a hyperspectral camera.

[0027] The memory should contain spectra of comparison samples in which depressions were formed using the same procedure as those used to form depressions in at least one layer of the substrate. It is advantageous for each depression to be formed entirely by a single layer, so that the base of a depression is formed by the material of the next layer or the substrate material. However, it is also possible to partially remove the layers in order to obtain more comparison samples for a more precise characterization of the depression formed in each case, in particular its depth. Accordingly, the material of one or more layers in a depression should be removed to a predefined depth before the depth of the respective depression is determined and used for a comparison sample.If the thickness of the respective layer(s) is known, the depth can preferably be determined without additional measurement if the material of one or more layers has been completely removed and the base or bottom of the depression is formed by the material of the next layer or the substrate material. Otherwise, the depth of the reference depression must be determined using an alternative reference method.

[0028] However, it is also possible to determine the depth using a different measuring method.

[0029] If the depth of the depressions is known, a spectrum is then acquired with the detectors of an inventive arrangement and stored in memory as a reference pattern. The spectra acquired from depressions of different depths can then be used for comparison in tests on a real substrate.

[0030] The electronic evaluation unit can calculate the spectral angle, or another measure of similarity, between the spectra of the hypercube and all measured reference spectra. Each spectrum can be assigned to the class of the reference spectrum with which it achieves the greatest agreement.

[0031] The totality of the wavelength-resolved intensities recorded at all locations of the respective irradiated area forms a three-dimensional data structure consisting of one wavelength-resolved and two spatially resolved dimensions (hypercube).

[0032] According to the invention, a spatially resolved optical spectroscopic analysis (hyperspectral imaging, HSI) of the respective sample is evaluated. The spectra acquired at individual locations, which have been subjected to the energy of the applied energy beam, are used to determine the material composition or material of a layer. From this, the depth of a depression formed in one or more layers can be determined.

[0033] Furthermore, the quality of the formed depression can be assessed. This includes its homogeneity, the chemical composition of the depression base, the texture of the depression base, the influence of the depression edges (droplets, "ejecta"), the influence of the adjacent sample areas (changes due to energy input), other geometric parameters of the respective depression, and the width and shape of the depression. The position and course of a trench-shaped depression, or the parallelism and / or spacing of depressions to one another, can also be determined. Areas where no depression has formed can also be identified.

[0034] The use of hyperspectral technology or imaging spectroscopy enables the areal (laterally resolved) characterization of the areas in which at least one depression is formed and their immediate surroundings based on the evaluation of spectra (transmission and / or reflection) measured simultaneously at different locations on the sample, for example along a line.

[0035] Alternatively, the entire sample or a selected area of the sample for determination must be homogeneously irradiated with electromagnetic radiation; otherwise, lateral intensity fluctuations will overwhelm the intensity fluctuations caused by the sample. A laterally homogeneous light field should be realized.

[0036] For example, a microscope optics / illumination can be used alone or in conjunction with an integration sphere (integrating sphere) or similar structures to maintain a homogeneous light field.

[0037] Depending on the hardware conditions and the sample surfaces, different optics, working distances and magnifications can be used or maintained to realize an optical setup.

[0038] The spectra can be recorded in transmission with an angle of 0° relative to the normal of the respective surface, or in reflection with an angle in the range between 0° - < 90° relative to the normal of the respective surface.

[0039] Polarized electromagnetic radiation can be used for irradiation. Detection can be performed by reflection or transmission, from both sides of the substrate. In this case, the substrate and layer(s) should be transparent.

[0040] Electromagnetic radiation emitted by a coated substrate can also be detected. For this purpose, the substrate and / or at least one layer can be exposed to energy so that it emits electromagnetic radiation. The energy can be introduced by heating the sample. In this case, electromagnetic radiation from the NIR and IR spectrum is predominantly emitted. Alternatively, fluorescence excitation can lead to emission. In this case, at least one layer or the substrate is made of a suitable material or contains such a material.

[0041] The invention enables the continuous acquisition of spatially resolved chemical and geometric information, such as the depth of the depression(s) or the number of layers still present on a substrate in the region of a depression after material has been removed by the energy beam. The depth of the base of a depression and the material forming the base of a depression can also be determined. With suitable alignment and arrangement of the detectors, the depth and / or the material at the base point, i.e. where material is currently being removed by the energy beam, can be determined. Simultaneously, information can also be obtained from the edge regions of a depression by optical examination without simultaneous exposure to energy from the energy beam.

[0042] Using an arrangement according to the invention, continuous, complete monitoring and / or control of the depression formation process can be achieved. However, it can also be used solely for quality control.

[0043] The current processing depth during decoating processes of optically sufficiently transparent material systems, especially multi-layer systems, can be determined by classification. Using a spectral analysis of the currently exposed material and knowledge of the layer structure, the respective processing depth can be continuously determined. This can be done across the entire surface.

[0044] In addition to the aforementioned laser beam, an electron beam or plasma beam can also be used as an energy beam. However, recesses can also be created mechanically using machining, for example, a drill or milling tool. This is also possible chemically using an etching process, which can advantageously be combined with a photolithographic process.

[0045] It can be used to inspect so-called insulating scribes in photovoltaic and LED production (thin-film electronics in general), particularly on the P1, P2, and P3 layers of PV modules. It can also be used to detect so-called "hidden contacts" in layers made of insulating, transparent, and / or electrically conductive oxides, as well as in edge insulation or edge finishing of substrates, especially semiconductor substrates. Example

[0046] For the investigation, a polyethylene (PET) film is being investigated, on which a layer of an organic material (organic) with a layer thickness of 20 µm is formed, a layer of SiN with a layer thickness of 150 nm, a layer of indium tin oxide (ITO) with a layer thickness of 130 nm, and on top of that, an organic layer (HDR) with a layer thickness of 100 nm is formed. The multilayer system formed on the film is intended to form several organic photovoltaic elements.

[0047] The film is processed with a pulsed laser to partially remove the individual layers on the film, thereby creating electrically isolated areas ('scribing' the layer). Various settings are selected, both for the power of the laser beam used or the individual laser pulses, as well as for the overlap of the individual laser spots, in order to obtain differently shaped trench-shaped depressions ('scribes') through the desired material removal. The average width of the resulting depression is approximately 50 µm.

[0048] The aim of the study is to determine the depth of depressions based on the spectral signature of the respective depression.

[0049] The prepared wells are illuminated with a halogen light source directed onto the surface of the film on which the multilayer system is formed. The examination is carried out using an optical microscope with 10x magnification. The microscope is connected to a total of 1000 x 190 optical detectors arranged in a row-column array. The optical detectors enable wavelength-resolved recording of the intensities of the radiation reflected by the sample for all locations on the sample.

[0050] The coated film is moved perpendicular to the optical axes of the detectors to allow larger sections of the wells to be analyzed. The resulting spectral image of the sample can also be referred to as a hypercube. The spatial resolution is approximately 1 µm and the spectral resolution approximately 3 nm. The detected wavelength range is between 400 nm and 1000 nm.

[0051] Furthermore, reference spectra were determined for later evaluation. For this purpose, spectra of the following layer systems were determined: • [1] PET, 20 µm organic' / 150 nm SiN / 130 nm ITO / 100 nm 'HDR' • [2] PET '20 µm organic' / 150 nm SiN / 130 nm ITO • [3] PET, 20 µm organic' / 150 nm SiN

[0052] The resulting hypercube of the examined wells is then processed and evaluated.

[0053] The spectra were processed using the following steps: • Mean smoothing (11 points) of all spectra of the hypercube and the reference spectra • Calculation of the spectral angle between the spectra of the hypercube and all measured reference spectra according to: α=cos−1(x→⋅r→‖x→‖⋅‖r→‖) • with: ◯ x ... spectrum of the hypercube ◯ r ... Reference spectrum ◯ α ... 'angle' between the spectra

[0054] The respective spectrum is assigned to the class with whose reference spectrum it achieves the highest α value. If sufficient agreement is not found with any of the reference spectra, for which a threshold is set around a mean value that must not be exceeded or undershot, the spectrum is classified as a "defect." A larger number of reference spectra could increase the accuracy of the classification.

[0055] Based on the classified spectra, the depth to which the layer system was ablated by the laser can then be determined. It is also possible to calculate general parameters of the depressions, such as width and shape.

[0056] The respective depression can be defined, for example, by the previously performed classification. All spectra that cannot be assigned to the untreated coating system (here [1] PET, 20 µm organic / 150 nm SiN / 130 nm ITO / 100 nm HDR) are considered to be the depression area.

[0057] The width of a depression can be determined by the number of image points (pixels) detected by the detectors at a specific position of a sample perpendicular to the longitudinal direction of the respective depression.

[0058] The absolute width of each depression can be calculated from the number of pixels captured by the detectors and the known width of the pixels. The width of the pixels is known due to the optical elements used in the beam path between the respective sample and the respective detector. In this example, this width is 0.8 µm. To improve accuracy, the mean and standard deviation of the calculated depression width can be calculated for a specific area of the depression captured by the detectors.

[0059] Faulty depressions can be detected if the parameters mentioned, e.g. the mean value and the standard deviation of the calculated depression width for a certain area of the depression detected by the detector, exceed or fall below a threshold value of these parameters.

[0060] The shape of a depression can be assessed by calculating geometric parameters, for example the area as the total number of pixels that can be assigned to a depression (for a specific area of the depression detected by the detector), the circumference with the number of pixels detected on the outside of a depression (for a specific area of the depression detected by the detector), or the spatial coordinates of the centroid of the depression surface (for a specific area of the depression detected by the detector), by comparison with previously determined parameters for a depression rated as good. For example, deviating ratios of circumference to area, chipping or delamination, or the deviation of the spatial coordinate of the centroid of the surface indicate a curvature of the depression.

Claims

[1] Arrangement for determining the depth of depressions formed in surfaces of a substrate on which at least one layer of a material different from the substrate material is formed, in which several detectors designed for spatially resolved spectral analysis of electromagnetic radiation within a wavelength interval, arranged in a row or a row and column arrangement and the detectors are connected to an electronic evaluation unit and arranged so that electromagnetic radiation emitted by a broadband radiation source is diffusely directed onto a surface and impinges on the detectors either after reflection at the surface of the substrate or a layer formed on the surface of the substrate and / or after radiating through the substrate transparent to the electromagnetic radiation or the substrate with at least one layer present on a surface, wherein the irradiation is carried out in such a way that a laterally and temporally homogeneous intensity of the electromagnetic radiation is maintained on a surface from which the electromagnetic radiation is reflected or through which the surface is transmitted, or electromagnetic radiation emitted by the substrate and / or at least one layer formed on the substrate strikes the detectors as a result of an energy input and the electronic evaluation unit is designed such that the spatially and wavelength-resolved measurement signals of the detectors can be detected within a wavelength interval for individual locations on a surface of the substrate or a layer formed on a substrate, and measurement signals detected at several positions can each be assigned to a partial area of the detected surface, and Based on these spatially and wavelength-resolved measurement signals, a comparison is made with previously determined measurement values stored in an electronic memory, which can be carried out in an analogous manner on comparison samples produced with the same materials and the same layer thicknesses and in which depressions of a defined depth have been formed beforehand, whereby If there is sufficient agreement between spectra recorded with the arrangement on a substrate and spectra of a comparison sample, the depth of a depression can be determined. [2] Arrangement according to claim 1, characterized by that the irradiation of the surface occurs at least at an angle in the range 0 ° to < 90 ° with respect to the normal of the surface of the substrate. [3] Arrangement according to one of the preceding claims, characterized by that the detection and evaluation can be carried out using at least one polarizer with at least one defined, known polarization plane relative to the plane of incidence. [4] Arrangement according to one of the preceding claims, characterized by that the detectors and the substrate are movable relative to each other along at least one axis and preferably at a constant distance from each other. [5] Arrangement according to one of the preceding claims, characterized bythat the radiation source has optical elements that form the electromagnetic radiation or that the radiation source is a radiation source that emits electromagnetic radiation diffusely onto the surface and is arranged within a hollow body. [6] Arrangement according to one of the preceding claims, characterized by that in the beam path of the electromagnetic radiation, a diaphragm is arranged in front of the detectors to prevent the incidence of scattered electromagnetic radiation. [7] Arrangement according to one of the preceding claims, characterized by that the row and column arrangement of detectors with optical elements and evaluation electronics is formed with a hyperspectral camera. [8] Arrangement according to one of the preceding claims, characterized bythat spectra of comparison samples in which depressions were formed using the same procedure as those used to form depressions in at least one layer of the substrate are stored in the memory. [9] Arrangement according to one of the preceding claims, characterized by that the electronic evaluation unit can be used to calculate a distance measure between the spectra, which have each been assigned to a sub-area of the detected area using measurement signals recorded at several positions, and all measured reference spectra, and that the respective spectrum can be assigned to a class with whose reference spectrum the smallest distance is achieved.