Determination of inhomogeneities when producing electrodes for battery cells

Abstract

The invention relates to an apparatus for hyperspectral imaging of electrode tracks (26, 1) when producing electrodes for battery cells. The apparatus comprises: - at least one hyperspectral spectroscopy unit (27) comprising a line scan camera (21) for imaging, wherein the spectroscopy unit (27) is positioned and designed - to capture hyperspectral images of the forward-moving electrode track (26, 1) at a predefined location of the electrode production, - to determine inhomogeneities of the electrode track (26, 1) based on the images, and - to determine and secure the local position of said inhomogeneities in the longitudinal direction of the electrode track (26, 1). The invention also relates to an associated method.

Description

[0001] Description

[0002] Determination of inhomogeneities in electrode production for battery cells

[0003] FIELD OF THE INVENTION

[0004] The invention relates to a device and a method for hyperspectral imaging of electrode strips during electrode production for battery cells.

[0005] BACKGROUND OF THE INVENTION

[0006] Lithium-ion accumulators, also referred to as lithium-ion batteries, are used as energy storage devices in mobile and stationary applications due to their high power density and energy density.

[0007] A lithium-ion battery typically comprises several battery cells. A battery cell, in particular a lithium-ion battery cell, comprises a plurality of layers. These layers typically include anodes, cathodes, separators, and other elements. These layers can be configured as stacks or as windings.

[0008] The electrodes usually comprise metal foils, in particular comprising copper and / or aluminum, which are coated with an active material. A lithium-containing paste, called slurry, is typically applied as the active material. The foils and the coating each have a thickness of a few micrometers. As a result, even a deviation of just a few micrometers in the thickness of the coating or in the material properties, in particular the material composition, has a negative impact on the quality of the electrode. The disadvantage of irregular coating is that battery cells of inferior quality are produced. Furthermore, safe operation of the battery cell cannot be guaranteed. Currently, with the state of the art, defective coatings can often only be identified after the entire battery cell production process has been completed, as part of a so-called end-of-line test.In some cases, defective coatings are only discovered after the battery cell has been in operation for several years.

[0009] Battery materials for battery cell production are deposited or applied to the electrode track as a viscous medium (a coating paste, English: slurry).

[0010] The continuous determination of material parameters (e.g. composition, homogeneity of the distribution of an ingredient, moisture ...) along an (endlessly long) production line / product line (e.g. battery electrodes, paper web, steel, ...) or a series production line (e.g. biscuits, pretzels, brake pads, ...) is of critical importance for the result.

[0011] Battery cell production in particular (here referring to lithium-ion cells) is today very often characterized by an exploratory approach (i.e., a mixture of craftsmanship and art rather than science or a fully understood industrial process). The pasty active material is applied to both sides of a long electrode track (typically sequentially), dried, and processed. Process parameters, in particular, are often stored as recipes (e.g., drying for x hours at y degrees) and are rarely monitored by sensors.

[0012] Since the ingredients for the pasty mixture, the so-called slurry, made of NMC or graphite (as the main components) with a solvent and possibly other components (e.g. conductive carbon black), which is applied to the electrode track, are mixed in a mixer and the homogeneity of the mixture and the viscosity are crucial for the performance of the subsequent battery cell, and thus of the entire subsequent battery, this homogeneity of the coating must be monitored during the manufacturing process.

[0013] With a suboptimally homogenized mixture, areas will appear that arise from coating with more or less solvent (water, NMP). On the other hand, this also creates areas of varying chemical composition. Furthermore, uneven deagglomeration of the solid particles can occur, i.e., a deviation of the particle size or particle size distribution from the target value if the mixing process is subject to fluctuations. This enables closed-loop feedback between the inhomogeneity detection and the slurry mixer.

[0014] During further processing, these areas can then lead to uneven behavior of the manufactured electrodes and / or cause short circuits and inefficiency in the battery cells.

[0015] To date, samples have been extracted by punching out sections of the produced web or removing entire sections from the production line. The material thus extracted is then, ideally, analyzed microscopically. Deviations in the chemical composition of the layers or the particle size distribution can be detected offline (in laboratory tests) using X-ray methods.

[0016] SUMMARY OF THE INVENTION

[0017] The object of the invention is to provide a solution that improves the electrode production for battery cells.

[0018] The invention results from the features of the independent claims. Advantageous developments and refinements are the subject of the dependent claims. Further features, possible applications, and advantages of the invention emerge from the following description. One aspect of the invention relates to the determination of the local position of inhomogeneities in electrode tracks during the production of coated electrodes for battery production.

[0019] The invention claims a device for hyperspectral imaging of electrode strips during electrode production for battery cells. The device comprises: at least one hyperspectral spectroscopy unit having a line camera for imaging, wherein the spectroscopy unit is arranged and configured

[0020] - to take hyperspectral images of the advancing electrode track at a given location during electrode production,

[0021] - to determine inhomogeneities of the electrode path from the images, where an inhomogeneity is a deviation of the chemical composition of the layers or the particle size distribution from specified target sizes, and

[0022] - to determine and secure their local position (ie a spatial coding) in the longitudinal direction of the electrode track.

[0023] The electrode track is a carrier film coated with the slurry. The line scan camera is called a "push-broom scanner." "Save" is defined broadly, also in the sense of "store" or "mark."

[0024] The expert selects the wavelengths of the hyperspectral imaging depending on the inhomogeneities to be detected and the materials from which the slurry, i.e. the coating of the electrode track, is formed.

[0025] For example, the inhomogeneity can be determined by comparing it with previously determined reference images (= target values). Reference images can be obtained, for example, from a previously produced electrode track with a proven homogeneous structure or from an averaging of images taken from the first few meters of a newly produced electrode track.

[0026] The device has a deflection roller over which the electrode track is guided, wherein the line scan camera takes the images at the position of the contact surface between the deflection roller and the electrode track.

[0027] In a further development, the device is designed to secure the local position of the inhomogeneity by marking the electrode track. These markings can be retrieved, for example, when cutting the electrode track and used for sorting.

[0028] In a further development, the device comprises an engraving unit configured to perform the marking by engraving, preferably along the longitudinal edge of the electrode track. This can be done, for example, using a laser engraving machine.

[0029] In a further development, the device comprises a storage unit configured to store spectral characteristics of the inhomogeneity as spatial coding along the electrode path such that the position of the inhomogeneity can be located based on the spectral characteristics. The stored spectra of the inhomogeneities can be retrieved, for example, when the electrode path is cut.

[0030] In a further development, the device comprises a production control system to which the spectroscopy unit is connected and which is configured to trigger the spectroscopy unit for measurement. This allows for subsequent local assignment of the measurement, for example, during the cutting and sorting of electrode tracks.

[0031] In a further development, the device has an evaluation and computing unit at the location of the spectroscopy unit, wherein the evaluation and computing unit is configured to pre-compress spectroscopic imaging data into a vector and assign it to the local position. This determines a measure of the lateral distribution of materials along a considered measurement line. This means that a vector of inhomogeneity is created.

[0032] In a further development, the line scan camera is a SWIR line scan camera. This allows for a cost-effective solution.

[0033] In a further development, the device in the spectroscopy unit has an AI engine that is set up to determine the inhomogeneities with the aid of a computer.

[0034] In a further development, the AI ​​engine is trained using deep learning methods in order to determine the inhomogeneities from the spectra.

[0035] The invention also claims a method for determining inhomogeneities by means of hyperspectral imaging on electrode tracks during electrode production for battery cells, comprising the following steps:

[0036] - Taking hyperspectral images at a given location of the electrode production of the forward moving electrode track,

[0037] - Determination of inhomogeneities of the electrode path from the spectra of the images, where an inhomogeneity is a deviation of the chemical composition of the layers or the particle size distribution from predetermined target sizes, and

[0038] - Determination and securing of the local position of the inhomogeneities in the longitudinal direction of the electrode track.

[0039] The images are taken by the line scan camera at the position of the contact surface between a deflection roller, over which the electrode track is guided, and the electrode track.

[0040] In a further development of the method, the local position of the inhomogeneity is secured by marking it on the electrode path. In a further development of the method, the spectral characteristics of the inhomogeneity are stored as spatial coding along the electrode path in such a way that the position of the inhomogeneity can be located based on the spectral characteristics.

[0041] In a further development of the procedure, the inhomogeneities are determined using a computer-aided AI engine.

[0042] The invention offers the following advantages, among others:

[0043] Measurement on a pulley or above a supporting / stabilizing track segment,

[0044] Direct online evaluation allows closed-loop feedback with, for example, the upstream dryer,

[0045] Field-near pre-compression of the data reduces the load on IT and enables a physically “correct” evaluation (in the sense of a real measurement value instead of a “house number”) through the associated semantic annotation,

[0046] Creation of semantic annotation,

[0047] Integration into manufacturing automation,

[0048] Assignment to a position along the “product”,

[0049] traceability,

[0050] Continuous measurement information,

[0051] Non-destructive without sampling, and continuous recording in the process, which provides a characteristic 2D inhomogeneity indication, which is suitable for the identification / recognition of path segments in subsequent process steps.

[0052] Further features and advantages of the invention will become apparent from the following explanations of an embodiment using schematic drawings.

[0053] BRIEF DESCRIPTION OF THE DRAWINGS

[0054] They show:

[0055] FIG. 1 the coating of the carrier film,

[0056] FIG. 2 the drying of the coated carrier film,

[0057] FIG. 3 the calendering of the dried carrier film

[0058] FIG. 4 shows a block diagram of the hyperspectral measurement setup and

[0059] FIG. 5 shows a section of a hyperspectral image of an electrode track.

[0060] DETAILED DESCRIPTION OF THE INVENTION

[0061] FIGS. 1 to 3 schematically show, by way of example, an electrode production process for batteries. First, a carrier film 1 is coated with the slurry 2, which is conveyed from a reservoir 7, via an application tool 3, e.g., a slotted nozzle, a doctor blade, or an anilox roller (FIG. 1). The carrier film 1 is coated in the coating direction 4 either continuously or intermittently (intermittent coating 6 in plan view 5).

[0062] The foil thicknesses (anode - copper foil and cathode - aluminum foil) vary depending on the battery cell design between

[0063] 10 μm and 25 μm. The carrier film 1 is coated over a width of up to 900 mm in a roll-to-roll process. Depending on the design of the production line, the top and bottom sides of the film are coated either sequentially or simultaneously. The thickness of the coating 6 with the slurry 2 is determined while the film is still wet using a wet thickness gauge 8.

[0064] After coating (FIG. 1), the coated carrier film 1 (also referred to as the electrode sheet) is fed directly into a dryer 9 (FIG. 2). For simultaneous, double-sided coating, a suspended-bed dryer is used. The solvent (vapors 10 via exhaust 12) is removed from the coating 6 by applying heat via air nozzles 11 and is either recovered or recycled for thermal recycling.

[0065] The dryer length determines the achievable throughput speed. The dryer 9 is divided into different temperature zones to implement a customized temperature profile. After passing through the dryer, the dried electrode web (= coated carrier film 1) is cooled to room temperature by a cooling device 13 with the aid of cooling rollers 14.

[0066] During the subsequent calendering step 15 (FIG. 3), the carrier film 1, coated on both sides, is compacted by one or more rotating roller pairs 16. The film is compacted by upper and lower rollers. The roller pair 16 generates a precisely adjustable line pressure. This line pressure determines the porosity of the electrode web. An excessively high line pressure creates a squeezing process and damages the electrode web. The cleanliness of the roller pair 16 is crucial to prevent foreign particles from penetrating the substrate material.

[0067] The carrier film 1 is transported through the calender 15 by means of a pulling device 17. The feeder 18 ensures a clean feed from a storage roll 19 into the calender 15. After calendering, the electrode web is cleaned in a cleaning device 20 and stored as a "mother coil 21" for further processing into battery cells, such as cutting.

[0068] Advantageously, a hyperspectral measurement is carried out after drying (FIG. 2) or after calendering (FIG. 3) with a line scan camera 22 and a downstream evaluation and computing unit 23.

[0069] FIG. 4 shows the block diagram of a hyperspectral measurement arrangement comprising a spectroscopy unit 27 with a line-scan camera 22, which records the hyperspectral images of the coated carrier foil (= electrode track 26) at the location of a deflection roller 25. The line-scan camera 22 is connected to the evaluation and computing unit 23, which in turn is connected to a production control system 24 for the electrode track production. An engraving unit 28 is provided for marking the electrode track 26, for example, at detected inhomogeneities.

[0070] Hyperspectral imaging in the variant of the line camera 22 (= push-broom scanner: line measurement of spectra with high lateral, i.e., spatial, resolution) is preferably used on the deflection roller 25, as shown in FIG. 4, in order to a) observe a stationary electrode track position, b) obtain synchronization with the relative movement of the electrode track, c) enable position traceability, and d) create a stable measurement situation (robust against vibration, can be darkened, etc.).

[0071] In the SWIR (850-1700 nm), imaging (also inexpensive) reveals signatures that vary with the degree of homogeneity of the source material distribution. In the NIR (1200-2500 nm), even more specific bands are possible, but the price of the spectral instruments is high. Therefore, the use of SWIR is advantageous from an economic perspective.

[0072] The spectral measurement system is advantageously connected to the production control system 24 (PLC, Edge, etc. ) and receives triggers for the measurement from here, which facilitates later local assignment of the measurement.

[0073] The advantage here is the early processing of data in the field, close to the measuring point. This reduces the load on the IT system, as high data rates occur due to the high product speed and would block even modern high-speed networks.

[0074] Therefore, immediately after the measurement, the data set is pre-compressed, and a measure of the lateral distribution of the materials along the measurement line under consideration, i.e., a vector of inhomogeneity, is assigned to the path position. This information is already two to three orders of magnitude smaller than the raw data of the actual measurement (one HSI frame: 640 x 224 pixels vs. one evaluation result: 640 x 1 vector or just a scalar indicator).

[0075] FIG. 5 shows a measurement point of one exemplary embodiment. The left image shows a section of a hyperspectral image of an electrode track. Inhomogeneities are represented here by different shades of gray. The right diagram shows the spectral analysis of the pixel with the coordinate 50 / 20. For each spatial coordinate x, y of the left image, there is a spectrum that describes the chemical properties of this pixel.

[0076] The measured spectra can be analyzed preferably in two ways: A) Supervised machine learning: Calibration of a model that either maps the spectra to a target value (e.g., graphite content) or classifies them (e.g., OK, not OK). Methods such as partial least squares, principal component regression, Gaussian processes, or neural networks exist for this purpose.

[0077] B) Unsupervised machine learning: Detection of anomalies, e.g., local inhomogeneities in the coating of the electrode track (e.g., the variance of slurry deposition may increase at the outer edges, changes in layer thickness). Statistical features, such as local variance or entropy, can be calculated for this purpose. Alternatively, residual models (e.g., low-rank PCA approximation, variational auto-encoder neural networks) can be created.

[0078] In both cases, the analyses can be supplemented by preprocessing steps, for example by isolating the coating foreground from the substrate film, filtering out sensor information according to its degree of variance explanation, or increasing the contrast.

[0079] Although the invention has been illustrated and described in detail by the embodiments, the invention is not limited by the disclosed examples and other variations can be derived therefrom by a person skilled in the art without departing from the scope of the invention.

[0080] Reference symbol list

[0081] 1 carrier film

[0082] 2 S lurry

[0083] 3 Application tool

[0084] 4 Coating direction

[0085] 5 Top view

[0086] 6 intermittent coating

[0087] 7 Memory for S lurry 2

[0088] 8 Wet thickness measurement

[0089] 9 dryers

[0090] 10 vapors

[0091] 11 Air nozzle

[0092] 12 deduction

[0093] 13 Cooling device

[0094] 14 cooling rollers

[0095] 15 calenders

[0096] 16 pairs of rollers

[0097] 17 traction mechanism

[0098] 18 Indentation

[0099] 19 Storage roll

[0100] 20 cleaning plant

[0101] 21 Mother coil

[0102] 22 line scan camera

[0103] 23 Evaluation and calculation unit

[0104] 23 . 1 storage unit

[0105] 24 Production control system

[0106] 25 pulley

[0107] 26 Electrode track

[0108] 27 Spectroscopy unit

[0109] 28 Engraving unit

Claims

Patent claims 1. A device for hyperspectral imaging of electrode pathways (26, 1) in the electrode manufacturing of battery cells, comprising: a) at least one hyperspectral spectroscopy unit (27) comprising a line scan camera (21) for imaging, wherein the spectroscopy unit (27) is arranged and configured, - to take hyperspectral images of the forward-moving electrode path (26, 1) at a predetermined location in the electrode manufacturing process, - to determine inhomogeneities of the electrode path (26, 1) from the images, where an inhomogeneity is a deviation of the chemical composition of the layers or the particle size distribution from specified target parameters, and - to determine and secure the local position of the inhomogeneities in the longitudinal direction of the electrode path (26, 1) 25), and b) a deflecting roller (25) over which the electrode path (26, 1) is guided, wherein the line camera (21) records the images at the position of the contact surface between the deflecting roller (25) and the electrode path.

2. The device according to claim 1, wherein the local position of the inhomogeneity is secured by marking on the electrode track (26, 1).

3. The device according to claim 1 or 2, characterized by an engraving unit (28) which is configured to perform the marking by engraving on the longitudinal edge of the electrode track (26, 1).

4. The device according to one of the preceding claims, characterized by a storage unit (23.1) which is configured to store spectral characteristics of the inhomogeneity as a spatial encoding along the electrode path (25, 1) in such a way as to Store the information so that the position of the inhomogeneity can be located based on the spectral characteristics.

5. The device according to one of the preceding claims, characterized by: - a manufacturing control system (24) to which the spectroscopy unit (27) is connected and which is set up to trigger the spectroscopy unit (27) for measurement.

6. The device according to one of the preceding claims, characterized by: - an evaluation and computing unit (23) at the location of the spectroscopy unit (27) , wherein the evaluation and computing unit (23) is configured to condense spectroscopic imaging data into a vector and assign it to the local position .

7. The device according to one of the preceding claims, characterized in that the line scan camera (22) is a SWIR line scan camera.

8. The device according to one of the preceding claims, wherein the spectroscopy unit (27) has an AI engine configured to determine inhomogeneities using computer assistance.

9. The device according to claim 8, wherein the AI ​​engine is trained using deep learning methods to determine the inhomogeneities from the spectra.

10. Method for determining inhomogeneities by means of hyperspectral imaging on electrode traces (26, 1) in the electrode manufacturing for battery cells, characterized by: - Acquisition of hyperspectral images at a predetermined location in the electrode manufacturing process from the forward-moving electrode path (26, 1) , - Determination of electrode path inhomogeneities from the image spectra, where an inhomogeneity is a deviation the chemical composition of the layers or the particle size distribution of predetermined target parameters, and - Determination and securing of the local position of the inhomogeneities in the longitudinal direction of the electrode path (26, 1) , wherein the line camera records the images at the position of the contact surface between a deflecting roller (25) over which the electrode path is guided and the electrode path.

11. The method according to claim 10, wherein the local position of the inhomogeneity is secured by marking on the electrode path.

12. The method according to one of claims 10 or 11, characterized in that the spectral characteristics of the inhomogeneity are secured as a spatial coding along the electrode path in such a way that the position of the inhomogeneity can be located based on the spectral characteristics.

13. The method according to one of claims 10 to 12, wherein the inhomogeneities are determined using an AI engine in a computer-aided manner.

Images