The present invention relates to a method and apparatus for characterising a material. In particular, the present invention relates to a method of characterising planar material such as material layers or webs. Materials may include layers of material deposited during an industrial process, such as a process to manufacture battery electrodes. Planar materials may also include sheet goods, such as PET foam.

Batteries such as lithium ion batteries are used in an increasing number of application areas allowing small scale and large storage of electrical energy. As the demand for lithium ion batteries has increased, manufacturing processes are required to increase volume and quality so as to deliver an increased number of battery cells. Lithium ion batteries typically comprise one or more battery cells which can be arranged in different forms e.g. pouch, cylindrical, prismatic. Regardless of the type of cell used, the formation of the individual components typically requires a standard structure. That is, each cell comprises electrodes separated by a separator. An ion conducting electrolyte is provided within the battery which fills pores within the electrodes and can pass through the separator. Each electrode is provided with a current collector which typically comprises aluminium or copper.

Industrial processes are used to deposit the electrode materials onto the collector. Industrial processes are often monitored with rigorous quality control tests performed on the resulting electrodes before assembly into a battery cell.

In lithium-ion battery manufacture electrode materials are deposited onto the collector. It will be appreciated that defects within an electrode structure can result in deficiencies in the performance of a resulting battery cell. For example, variations in thickness and/or composition of the electrode materials can result in reduced energy storage capacity, reduced battery life, battery cell failure, or any number of different failure mechanisms. As such, accurate control of the processes involved in depositing the electrode materials is extremely important.

More generally, industrial processes are used to deposit materials onto a substrate, or to manufacture generally planar sheets of material. Such industrial processes are often monitored with rigorous quality control tests to ensure that the material conforms to suitable standards.

In another example, polymerfoam, such as, for example polyethylene terephthalate (PET) foam is typically manufactured in sheets for further processing. It will be appreciated that the internal defects such as voids within the porous structure of the PET foam may result in a defective product. In particular, voids within the foam may compromise the structural performance of the PET foam. PET foam is used in a wide variety of structural applications including, but not limited to, wind turbines and shipping. As such, accurate quality-control of the manufactured sheets is crucial.

Various techniques are known to monitor material layers coated on a substrate and/or planar materials more generally. For example, ionising radiation techniques (e.g. using X-rays, or p-radiation) may provide high accuracy when measuring the thickness or weight of deposited material, and can be integrated into an in-line process. However, such techniques typically involve a high cost and require extensive safety measures to be installed on a production line. Furthermore, components are typically required to be installed on both sides of a substrate material, which can put some restrictions on where such techniques can be applied.

Backscattering techniques can also be used to monitor the thickness of material. Such techniques can be easier to integrate than radiation-based techniques, and are typically lower cost. However, backscattering techniques are typically very sensitive to movement or vibrations, and can therefore be somewhat unreliable.

Ultrasound measurement techniques can also be used to monitor the thickness of some materials without the use of ionising radiation. However, such techniques typically involve a high cost and require double sided sensing.

Infra-red (e.g. near-IR) sensing can also be used to monitor the thickness of some materials without the use of ionising radiation. However, near-IR techniques typically require sensing components to be installed on both sides of the collector material, and cannot penetrate opaque materials. In general, monitoring the thickness of a material and/or the presence of internal defects (e.g. voids) can also be achieved by destructively sectioning production samples of the material. However, such a technique necessitates the destruction of the produced material and thus cannot be integrated into an in-line process.

It is an object of the present invention to provide an improved or alternative method of characterising material layers or webs which overcomes one or more of the problems associated with known methods, whether discussed above or otherwise.

According to a first aspect of the invention there is provided a method of characterising a material. The method comprises detecting a characteristic of the material at a monitoring location while the material is moved past the monitoring location in a movement direction. The method also, or alternatively, comprises detecting a characteristic of the material at a monitoring location while the material is moved relative to the monitoring location in the movement direction. The detecting comprises providing a plurality of electrodes comprising an array of electrodes parallel to and separated from the material and generating a plurality of measurement data items. Generating each measurement data item comprises i) selecting a pair of the electrodes of the plurality of electrodes, ii) applying an energisation signal to a first one of the selected pair of electrodes causing a field to be generated in a region extending from the plurality of electrodes towards the material, and iii) measuring an electrical parameter at a second one of the selected pair of electrodes in response to the energisation signal. The detecting further comprises generating data indicative of a characteristic of the material based upon the plurality of measurement data items.

According to a second aspect of the invention, there is provided a method of characterising a material. The method comprises detecting a characteristic of the material at a monitoring location. The detecting comprises providing a plurality of electrodes comprising an array of electrodes parallel to and separated from the material. The detecting further comprises generating a plurality of measurement data items. Generating each measurement data item comprises i) selecting a pair of the electrodes of the plurality of electrodes, ii) applying an energisation signal to a first one of the selected pair of electrodes causing a field to be generated in a region extending from the plurality of electrodes towards the material, and iii) measuring an electrical parameter at a second one of the selected pair of electrodes in response to the energisation signal. The detecting further comprises generating data indicative of a characteristic of the material based upon the plurality of measurement data items.

According to a third aspect of the invention, there is provided a method of processing a material. The method comprises detecting a characteristic of the material at a monitoring location during said processing. The detecting comprises providing a plurality of electrodes comprising an array of electrodes parallel to and separated from the material. The detecting further comprises generating a plurality of measurement data items. Generating each measurement data item comprises i) selecting a pair of the electrodes of the plurality of electrodes, ii) applying an energisation signal to a first one of the selected pair of electrodes causing a field to be generated in a region extending from the plurality of electrodes towards the material, and iii) measuring an electrical parameter at a second one of the selected pair of electrodes in response to the energisation signal. The detecting further comprises generating data indicative of a characteristic of the material based upon the plurality of measurement data items.

Each of the optional features described below can be combined with one or more of the first, second or third aspects of the invention.

In accordance with the techniques described herein, it is possible to use electrical capacitance tomography (ECT) to monitor characteristics (e.g. thickness, mass, moisture content, porosity) of the material in real time, allowing a material process (e.g. a coating process) to be assessed and, if required, controlled dynamically. The methods are non-invasive and non-intrusive and can thus be carried out while processing is ongoing, without causing any interruption to processing. In this way, a simple and direct determination of the characteristic of the material at a particular location can be made.

Tomographic techniques often provide a ring of electrodes spaced apart around the periphery of a body to be imaged. However, rather than using tomography to provide detailed information relating to the material within the centre of the body (e.g. to monitor the flow of material within the centre of a pipeline) the method applies tomographic techniques to the inspection of the (generally planar) surface of a material, where an extended imaging depth may be of little importance. Thus, unlike in conventional tomographic processes, detection depth can be relatively small, and a sensor having a shallow depth of field can provide useful diagnostic information relating to a characteristics of a thin layer of material (or thin web of material).

The method may comprise moving the material past the monitoring location in the movement direction. The material may be arranged to lie in a plane parallel to the movement direction.

The material may comprise any generally planar material. The term planar material comprises any material having at least one substantially planar surface.

The material may comprise a web or layer of material. The layer of material may comprise a layer of the material deposited on a substrate. Moving the material past the monitoring location may comprise moving the substrate.

The web of material may comprise a continuous film of the material.

The method according to any preceding claim wherein for each of the plurality of measurements a different pair of electrodes is selected.

The (or each) array of electrodes may be a two-dimensional array. The array may comprise a planar array.

The plurality of electrodes may comprise the array of electrodes and at least one further electrode. The at least one further electrode may comprise a substrate configured to support the layer of material, or may be applied to such a substrate.

The plurality of measurements may comprise at least one measurement for each of the electrodes in the array.

The array of electrodes may be a one-dimensional array. The one-dimensional array may extend in a direction perpendicular to the movement direction.

The data indicative of a characteristic of the material may comprise first data indicative of a characteristic of the material at a first time. The method may further comprise generating second data indicative of a characteristic of the material at a second time, after the first time. In this way, it is possible to monitor changes in the material characteristics over time.

Generating data indicative of a characteristic of the material based upon the plurality of measurement data items may comprise processing the plurality of measurement data items based upon model data. The model data may be referred to as an inversion model, or reconstruction model.

The model data may comprise data indicative of the geometry of the plurality of electrodes. The model data may comprise data indicative of sensor configuration, sensor geometry, excitation pattern, and/or energisation signal.

The model data may be generated based upon the geometry of the electrodes and/or the geometry of the apparatus. The model data can then be used allow information relating to properties of the material (e.g. permittivity) to be extracted from measurements taken from the electrodes, which may, for example, be indicative of induced charge at locations in response to the applied energisation. That is, data relating to the geometry of the electrodes and/or the geometry of the apparatus can be used to allow sensed voltages to be related to properties of the material in particular locations.

Knowledge of the energisation signal can be used in conjunction with the measured electrical parameters to determine relationships therebetween. Such relationships allow data indicative of properties of material (e.g. permittivity data) around the electrodes to be derived.

Generating data indicative of a characteristic of the material may be further based upon reference data.

The reference data may comprise material reference data. The material reference data may comprise data indicative of the bulk permittivity of the material. The material reference data may comprise data indicative of the composition and/or rheology of the material. The material reference data may comprise material temperature data, material homogeneity data and/or material flow rate data. Such reference data may be measured. The method may thus further comprising obtaining the material reference data by performing a further measurement. The further measurement may comprise an ECT or ERT measurement at an upstream process. The further measurement may comprise an ECT measurement performed by the monitoring apparatus when a known material, or material having a known characteristic is present. The reference data may be referred to as calibration data.

The data indicative of a characteristic of the material may comprise a plurality of data values, each of the data values being indicative of the characteristic of the material at a respective one of a plurality of locations within the region extending from the plurality of electrodes towards the material.

The plurality of locations may be arranged in a two-dimensional array. Each data value may comprise a pixel. The data indicative of a characteristic of the material may comprise a tomogram.

The data indicative of a characteristic of the material may comprise permittivity data.

The data indicative of a characteristic of the material may comprise at least one of: thickness data, mass data, moisture content data, and porosity data.

The data indicative of a characteristic of the material may comprise permittivity data and one or more of thickness data, mass data, moisture content data, and/or porosity data. The thickness data, mass data, moisture content data, and/or porosity data may be generated based on permittivity data and material reference data. Permittivity data may be generated based on the plurality of measurement data items and model data.

The method may further comprise identifying defects in the material based upon the data indicative of the characteristic.

The method may comprise moving the material relative to monitoring location in a plane parallel to the array.

The monitoring location may be referred to as a static monitoring location. The material may be moved past the monitoring location. The method may comprise moving the two-dimensional planar array of electrodes relative to the material in a plane parallel to the material.

The monitoring location may be referred to as a dynamic monitoring location. Advantageously, this allows characterising materials post-manufacture in quality control process. Moving the two-dimensional planar array of electrodes may comprise moving the array by one or more actuators.

The method may further comprise moving the two-dimensional planar array of electrodes in a direction substantially normal to the material.

By moving the array relative to the material in a normal direction it is possible to adjust the separation (e.g. to achieve a target separation distance).

The method may further comprise monitoring a separation distance between moving the two-dimensional planar array of electrodes and the material in a direction substantially normal to the material.

By monitoring the separation distance it is possible to calibrate the characterisation process based on the separation and/or to adjust the separation (e.g. to achieve a target separation distance).

The data indicative of a characteristic of the material may comprise a plurality of data values, each of the data values being indicative of the characteristic of the material at a respective one of a plurality of locations within the region extending from the plurality of electrodes in a direction normal to the plane of the material.

There is also provided a method of qualifying a material. The method may comprise characterising the material according to the method of the first aspect, and generating qualification data based upon the data indicative of a characteristic of the material.

The qualification data may be generated based on one or more of: thickness data, mass data, moisture content data, density data, and porosity data. The qualification data may comprise data indicating at least one of: the presence of a void or defect within the material, the configuration of a void of defect within the material, the position of a void or defect within the material and the size of a void or defect within the material.

The method may further comprise performing a first characterising method to generate first data indicative of a characteristic of the material and a second characterising method to generate first data indicative of a characteristic of the material. The first characterising method may be performed before a material processing step, and the second characterising method may be performed after the material processing step. The method may further comprise comparing the first data indicative of a characteristic of the material and second data indicative of a characteristic of the material. Alternatively, the second data indicative of a characteristic of the material may be based on a second plurality of measurement data items generated by the second characterising method, and the first data indicative of a characteristic of the material generated by the first characterising method.

There is also provided a method of processing a material. The method may comprise processing the material by a processing apparatus, and characterising the processed material according to the method of the first aspect of the invention.

The method of processing a material may further comprise controlling the processing apparatus based upon the data indicative of the characteristic.

Controlling the processing apparatus based upon the data indicative of the characteristic may comprise controlling a process parameter based upon the data indicative of the characteristic.

The method may further comprise identifying defects in the material based upon the data indicative of the characteristic.

The processing may comprise coating a layer of the material onto a substrate.

The data indicative of a characteristic of the material may comprise thickness data. The thickness data may comprise thickness uniformity data. Identifying defects may be based on thickness data. The data indicative of a characteristic of the material may comprise mass data. The mass data may comprise mass uniformity data. Identifying defects may be based on the mass data. Mass data may comprise data indicative of a mass of material at a particular location (e.g. the mass of material deposited on a substrate).

The method may further comprise controlling the coating thickness based upon the data indicative of the characteristic.

Controlling the coating thickness may comprises adjusting a gap between a component of the coating apparatus and the substrate. Adjusting the gap between the component of the coating apparatus and the substrate may comprise adjusting a gap between the component of the coating apparatus and a guide roller configured to support the substrate.

Where the process is a coating process, the process parameter may comprise one or more of: a coating speed, a substrate tension, a process temperature, and a slurry pump flow rate.

The processing may comprise a drying process for drying the material.

The method may further comprise controlling the drying process based upon the data indicative of the characteristic. Where the process comprises a drying process, the characteristic may comprise a moisture content of the material. Where the process comprises a drying process, the process parameter may comprise one or more of: a drying temperature, a drying time, and a substrate speed.

The processing may comprise a calendaring process for compressing the material.

The method may further comprise controlling the calendaring process based upon the data indicative of the characteristic. Where the process comprises a calendaring process, the characteristic may comprise a porosity or the material. Where the process comprises a calendaring process, the process parameter may comprise one or more of: a pressure, and a substrate speed. The method may comprise performing a first characterising method according to the first aspect of the invention to generate first data indicative of a characteristic of the material before a first process. The method may further comprise performing a second characterising method according to the first aspect of the invention to generate second data indicative of a characteristic of the material after the first process.

For example, the method may comprise, generating data indicative of a moisture content before and after the drying process based upon the first characterising method and the second characterising method. Alternatively, or additionally, the method may comprise, generating data indicative of a porosity before and after the calendaring process based upon the first characterising method and the second characterising method.

The first characterising method may be performed using a first array of electrodes. The second characterising method may be performed using a second array of electrodes.

The method may comprise generating data indicative of the first process based on the data indicative of the characteristic generated by the first characterising method, and the data indicative of the characteristic generated by the second characterising method.

The method may comprise performing a plurality of characterising methods at a corresponding plurality of locations (e.g. before and after drying, and/or before and after calendaring), and generating data indicative of the process in question based on a comparison of outputs from the plurality of characterising methods.

The process may comprise a process for manufacturing a battery electrode.

The coating and/or drying and/or calendaring processes may comprise a process for manufacturing a battery electrode.

The method may further comprise generating data indicative of a characteristic of the battery electrode based on the data indicative of the characteristic of the material.

Data indicative of a characteristic of the battery electrode may comprise expected electrode performance data (e.g. a defective electrode, or an electrode storage capacity), and/or a defect density. The method may further comprise generating data indicative of an upstream process based on the data indicative of the characteristic of the material.

For example, electrode slurry composition or rheology data may be obtained based on the generated characteristic data.

The array of electrodes may comprise a first array. The plurality of electrodes may further comprise a second array of electrodes parallel to and separated from the material. The second array of electrodes may face the first array or electrodes and be provided at the opposite side of the material than the first array. Generating the plurality of measurement data items may comprise selecting a pair of the electrodes of the plurality of electrodes. A first electrode of the pair may be in the first array, and a second electrode of the pair may be in the second array.

By obtaining measurement from electrodes arranged on either side of the material, it is possible to improve the spatial and depth sensitivity.

The method may further comprise generating first data indicative of a characteristic of the material based on a first plurality of measurement data items, and generating second data indicative of a characteristic of the material based on a second plurality of measurement data items and model data and/or reference data.

In this way, it is possible to generate an approximate value for a characteristic (e.g. thickness) based on raw capacitance measurements (e.g. the first plurality of measurement data items), and separately to generate a more refined estimate of the same characteristic based on a full set of measurements (e.g. the second plurality of measurement data items). The second plurality of data items may comprise a greater number of data items than the first plurality of data times.

Data indicative of a characteristic of the material may comprise a plurality of data values, each of the data values being indicative of the characteristic of the material at a respective one of a plurality of locations within a three-dimensional sensing region extending between the first and second arrays of electrodes. The data indicative of a characteristic of the material may comprise a three-dimensional tomogram. The data indicative of a characteristic of the material may comprise a plurality of two-dimensional tomograms, each representing the characteristic of the material at a different location in a third dimension.

The method may further comprise identifying at least one of: the presence of a void, the configuration of a void, the position of a void and the size of a void within the material.

Where a three-dimensional tomogram is generated, or a plurality of two-dimensional tomograms, it may be possible to identify details of voids within the material that cannot be seen by visual inspection since they may not extend to an external surface of the material.

The methods of the first to third aspects of the invention can be carried out in any convenient way. In particular, the method may be carried out by a controller and such a controller is therefore provided by the invention. The controller may be provided by any appropriate hardware elements. For example, the controller may be microcontroller which reads and executes instructions stored in a memory, the instructions causing the controller to carry out a method as described herein. Alternatively, the controller may take the form of an ASIC or FPGA. The method may be carried out by a plurality of controllers.

According to a fourth aspect of the invention, there is provided a characterising apparatus arranged to detect a characteristic of a material at a monitoring location while the material is moved past the monitoring location in a movement direction. The apparatus comprises a plurality of electrodes comprising an array of electrodes arranged parallel to and separated from the material, an energisation source arranged to apply an energisation signal to one or more of the plurality of electrodes, a detector arranged to measure an electrical parameter at one or more of the plurality of electrodes, and a controller. The controller is configured to cause the apparatus to generate a plurality of measurement data items. Generating each measurement data item comprises: i) selecting a pair of the electrodes of the plurality of electrodes, ii) applying an energisation signal, by the energisation source, to a first one of the selected pair of electrodes causing a field to be generated in a region extending from the plurality of electrodes towards the material, and iii) measuring, by the detector, an electrical parameter at a second one of the selected pair of electrodes in response to the energisation signal. The controller is further configured to generate data indicative of a characteristic of the material based upon the plurality of measurement data items.

The material may comprise a web or layer of material.

The material may be moved relative to the monitoring location in a movement direction.

The plurality of electrodes may comprise a two-dimensional planar array.

The array of electrodes may comprise a first array. The plurality of electrodes may further comprise a second array of electrodes parallel to and separated from the material. The second array of electrodes may face the first array or electrodes and be provided at the opposite side of the material than the first array. The controller may be configured to generate the plurality of measurement data items by selecting a pair of the electrodes of the plurality of electrodes, a first electrode of the pair being in the first array and a second electrode of the pair being in the second array.

The characterising apparatus may further comprise one or more planar actuators configured to move the two-dimensional array of electrodes in a plane parallel to the material.

The characterising apparatus may further comprise a position controller configured to actuate at least one of the one or more planar actuators. The position controller may be configured to cause the one or more planar actuators to move the two dimensional array of electrodes over the material, the material being at rest.

The characterising apparatus may further comprise one or more separation actuators configured to move the two-dimensional array of electrodes in a direction substantially normal to the material.

The characterising apparatus may further comprise a position controller configured to actuate at least one of the one or more separation actuators. The position controller may be configured to actuate the one or more separation actuators so as to maintain a constant separation of the material and the two-dimensional array of electrodes. The characterising apparatus may further comprise a separation sensor configured to generate data indicative of a separation distance between the two-dimensional array of electrodes and the material.

The position controller may be configured to receive the data indicative of a separation distance between the two-dimensional array of electrodes and the material and to actuate the one or more separation actuators based upon the received data.

The data indicative of a characteristic of the material may comprise a plurality of data values, each of the data values being indicative of the characteristic of the material at a respective one of a plurality of locations within a three-dimensional sensing region extending from the array of electrodes.

The data indicative of a characteristic of the material may comprise a three-dimensional tomogram. The data indicative of a characteristic of the material may comprise a plurality of two-dimensional tomograms, each representing the characteristic of the material at a different location in a third dimension.

The controller may be configured to generate data indicative of at least one of: the presence of a void, the position of a void and the size of a void based upon the data indicative of a characteristic of the material.

Where a three-dimensional tomogram is generated, or a plurality of two-dimensional tomograms, it may be possible to identify details of voids within the material that cannot be seen by visual inspection since they may not extend to an external surface of the material.

According to a fifth aspect, there is provided a processing apparatus arranged to process the material. The processing apparatus further comprises a characterising apparatus according to the fourth aspect of the invention. The characterising apparatus is configured to generate data indicative of the characteristic of the material processed by the processing apparatus. The processing apparatus may comprise a process controller. The process controller may be configured to control the process based on the data indicative of the characteristic of the material.

It will be appreciated that features of the methods of the first to third aspects of the invention may be used in combination with the apparatus of the fourth and fifth aspects of the invention.

Embodiments of the invention will now be described, by way of example, with reference to the accompanying drawings in which:

Figure 1 is a schematic illustration of a lithium ion battery cell;

Figure 2 is a schematic illustration of various processes for manufacturing components of the lithium ion battery cell shown in Figure 1 ;

Figure 3 is a schematic illustration of an apparatus for monitoring one or more of the processes shown in Figure 2;

Figure 4 is a schematic illustration of a sensor for monitoring one or more of the processes shown in Figure 2;

Figure 5 is a schematic view of a controller arranged to perform processing according to an embodiment of the invention;

Figure 6 is a schematic illustration of an electrode coating with a superimposed sensor;

Figures 7a and 7b are tomograms indicative of permittivity generated by the apparatus of Figure 2;

Figure 8 is a 3D surface plot of thickness data derived from the permittivity data shown in Figures 7a and 7b.

Figure 9 is a flow chart, illustrating processing performed by the apparatus of Figure 3; Figure 10 is a schematic illustration of an alternative apparatus for monitoring one or more of the processes shown in Figure 2;

Figures 11a and 11b show low resolution tomograms of a relatively uniform thickness coating obtained by the apparatus of Figure 10;

Figures 12a and 12b show low resolution tomograms of a relatively non-uniform thickness coating obtained by the apparatus of Figure 10;

Figure 13 is a schematic illustration of a drying process;

Figure 14 is a schematic illustration of a calendaring process;

Figure 15 is a schematic illustration of an apparatus for qualifying sheet goods;

Figure 16 is a schematic illustration showing the traversal of a sensor of the apparatus of Figure 15 over a sheet of material.

Figure 17 shows a series of tomograms obtained from a dual plane monitoring apparatus when monitoring a low capacitance material (e.g. air);

Figure 18 shows a series of tomograms obtained from a dual plane monitoring apparatus when monitoring a high capacitance material (e.g. PET);

Figures 19a and 19b show a schematic illustration of PET foam sample having a vertically aligned defect, and a series of tomograms obtained from a dual plane monitoring apparatus when monitoring the PET foam sample, respectively; and

Figures 20a and 20b show a schematic illustration of PET foam sample having a horizontally aligned defect, and a series of tomograms obtained from a dual plane monitoring apparatus when monitoring the PET foam sample, respectively.

Figure 1 illustrates a typical lithium-ion battery cell construction. The battery cell 1 comprises an anode 2, the anode 2 comprising a current collector 3 and an anode electrode material 4. The anode is separated by a separator 5 from a cathode 6. The cathode 6 comprises a current collector 7 and a cathode electrode material 8. A region 9 spanning the anode and cathode materials 4, 8, and the separator 5 is provided with an electrolyte liquid, which fills voids therein, and allows ions to travel. During the manufacture of the anode 2 and cathode 6, the electrode materials 4, 8 are deposited on the respective collectors 3, 7 in continuous deposition processes.

Figure 2 schematically illustrates a typical electrode coating process. In a first step S1 a formulation is created comprising various active materials in specific mass ratios. The active materials typically comprise a binder, conductive agent, solvent and other additives. The composition will depend upon the nature of the electrode (e.g. whether it is an anode or a cathode) and the specific requirements for the battery cell to be produced. As the second step S2, the components are mixed together to provide a homogenous electrode slurry of mixed components. At step S3 the electrode slurry is coated onto a current collector in a thin layer. At step S4 the coated electrode material layer is dried, causing solvent to evaporate from the coated electrode slurry. At step S5 the porous electrodes are calendared. Calendaring comprises compression of the electrode material to increase its density. Once the electrodes have been processed in this way they can be cut, assembled and packaged before being provided with an electrolyte filling, as shown in Figure 1.

Various techniques are typically used to monitor the performance and condition of battery electrodes. As described above, such techniques can involve costly process steps, can require the use of ionising radiation, and can have limited monitoring speed. Alternative monitoring techniques may be performed more cheaply or without the use of radiation, but may require off-line and/or destructive testing of battery components once the process steps shown in Figure 2 have been completed.

It is believed that a significant proportion of the manufacturing costs of lithium-ion batteries are incurred as a result of high scrap rates of electrodes. It is desirable, therefore, to provide a more rapid monitoring process that allows real-time or in-line material characterisation. Such characterisation may enable the dynamic identification of defects and process variable upsets, thereby allowing process adjustments to be made. Such characterisation may enable real-time process control and optimisation, enhancing product conformity, and allow manufacturing costs to be reduced. It has been realised that by applying tomographic techniques to the monitoring of processes used in the manufacture of planar battery electrodes material characteristics and defects can be identified, and the processes controlled more accurately.

Tomography generally refers to the use of some form of penetrating wave to image a region of interest. Generally, an image is constructed by the combination of a plurality of image sections. The penetrating waves may be generated by electric or magnetic fields. Electrical Resistance Tomography (ERT) generally uses electrodes placed around the region of interest to monitor resistivity within the region of interest, and can be used for example to monitor characteristics of mixtures of conductive fluids in pipes. Electrical Capacitance Tomography (ECT) also uses electrodes placed proximate to a region of interest, but monitors electrical permittivity within the material of interest. ECT can thus be used to distinguish between materials having different electrical permittivities.

Tomographic techniques based on both ERT and ECT have become established methods for analysis and visualizing the flows of liquid and particulate (e.g. pneumatically conveyed) materials in pipelines. Such tomographic techniques can provide visualization of a cross-section of a flow stream. Tomography has also been used to develop ‘probe’ type arrays that once immersed in a medium can ‘see’ around the probe to provide a three-dimensional visualization of the medium. Such probes may, for example, be used to identify any stratification that may be present in the medium surrounding the probe.

Existing tomographic techniques often provide a ring of electrodes spaced apart around the periphery of a body or region to be imaged. Various combinations of the electrodes are energised, and signals are detected from either those electrodes which are energised, or other ones of the electrodes, depending on the energisation scheme used.

Figure 3 illustrates a coating apparatus 10 comprising a slot die 11 arranged to deposit a layer 12 of electrode material 13 onto a collector 14. The collector is a continuous sheet of material which is advanced in a direction 15 (referred to as a movement direction) during the continuous coating process. The collector may be guided by one or more rollers 16. The electrode material 13 may be any material suitable for use in an electrode of a metal-ion battery (e.g. lithium iron phosphate (LFP), lithium cobalt oxide (LCO), lithium nickel cobalt manganese oxide (NMC). The collector 14 may be a metal foil (e.g. copper or aluminium). It will be appreciated, however, that different materials and material combinations can be used as appropriate, and that the techniques described herein are not limited to use with a particular type of battery, or battery material.

A monitoring apparatus 20 is arranged proximate to the coating apparatus 10. The monitoring apparatus is a tomographic monitoring apparatus comprising a sensor 21. The sensor 21 is an ECT sensor, and is described in more detail below. The monitoring apparatus 20 further comprises an energisation source 22, and a detector 23. The energisation source 22 and the detector 23 are each connected to the sensor 21. The energisation source 22 and the detector 23 are both connected to, and controlled by, a controller 24.

Figure 4 shows the sensor 21 in more detail. The sensor 21 comprises a plurality of electrodes 25. The electrodes 25 are arranged in a two dimensional array, each electrode being of substantially equal size and shape. In the illustrated arrangement, each electrode 25 is rectangular in shape. The two dimensional array comprises twelve electrodes 25, in three rows of four electrodes each, forming a rectangular array. The electrodes 25 may each be switchably connected to the energisation source 22 and detector 23 to allow energisation and detection, as described in more detail below.

In an exemplary embodiment, the sensor 21 has dimensions of around 150 mm by around 150 mm, with each of the electrodes having dimensions of around 25 mm by around 35 mm. Of course, larger or smaller sensor dimensions may be selected based on a particular monitoring application. For example, in an alternative arrangement, a sensor may have overall dimensions of around 50 mm x 200 mm, and may include an array of 2 x 6 electrodes, each of around 20 mm x 30 mm. In other examples, sensor dimensions may be significantly larger than this (e.g. around 1000 mm, or greater).

The number and sizing of electrodes 25 may be selected based upon a number of criteria. For example, a reduced number of electrodes may allow an increased temporal resolution (given that less time will be required to energise and detect signals for each electrode combination). On the other hand, an increased number of electrodes may provide improved spatial coverage (for a given electrode size). An electrode number of twelve may be considered to provide an acceptable compromise between spatial coverage and temporal resolution in some embodiments. It will be appreciated, however, that only two electrodes are required to obtain capacitance data. Furthermore, for a given number of electrodes, the size of each electrode determines the spatial resolution available. In particular, smaller electrodes result in a higher imaging resolution, but also result in a reduced spatial coverage. Again, a compromise should be found which is appropriate for each particular application.

The sensor further comprises a guard electrode 26, which extends around the perimeter of the rectangular array, and may be referred to as a peripheral guard electrode. The guard electrode 26 shields the electrodes 25 from the electrical influence of material outside the region of interest. The guard electrode may be connected to ground.

The guard electrode 26 acts to shield the electrodes from interference, reducing measurement noise. It will be appreciated that a plurality of separate or connected guard electrodes may be provided in some embodiments. Alternatively, the guard electrode 26 may be omitted entirely. In some embodiments, guard electrodes may be provided between adjacent ones of the electrodes 25. Such inter-electrode guard electrodes may, for example, have a width of around 1 mm.

The gap between each of the electrodes 25 and each of the adjacent electrodes 25, and also between each of the electrodes 25 and the guard electrode 26 (where present) may, for example, be around 1 mm where the electrode dimensions are as described above. Such spacing may help in reducing image reconstruction artefacts. For example, if large gaps are left between the electrodes 25, then there will be large areas (i.e. the areas corresponding to the gaps) to which only a limited sensing field can be applied, and in which limited detection of material characteristics can be made. That is, the sensitivity of the sensor 21 is reduced in areas corresponding to the gaps. As such, the spacing between electrodes may be maintained at a small distance (e.g. 1 mm) so as to avoid large gaps in the sensing ability of the sensor.

In the illustrated embodiment the sensor 21 is substantially planar, lying in a plane which is substantially parallel to, and separated from, the surface of the collector 14 onto which the layer 12 is deposited.

By “separated from” it is meant that the electrodes 25 of the sensor 21 are not in direct physical contact with the collector 14 or the material 12. As described in more detail below, the electrodes are configured to generate an electrical field in a region extending from the electrodes 25 towards the material 12.

It will be appreciated that alterations to the sensor geometry may be made. For example, a sensor may have a different number of electrodes than that shown in Figure 4. Moreover, each electrode may have a non-rectangular shape. For example, the electrodes may have shapes such as: square, strip-shape (i.e. narrow fingers), hexagonal, trapezoidal, and triangular. Similarly, it is not necessary that the electrodes are arranged in two-dimensional arrays. The electrodes may be arranged in a onedimensional array, the array extending in a direction perpendicular to the movement direction. The shape of the sensor and/or each electrode may be determined based upon the location of the sensor with respect to the geometry of the apparatus with which it is intended to operate.

It is preferred that the sensor is arranged such that it is parallel to the surface of the collector 14. By parallel, it is meant that the array is arranged such that the electrodes span at least one dimension that is substantially or generally (although not necessary strictly) parallel to the surface of the collector. Where a two dimensional array is provided, both array dimensions may be substantially, or generally, parallel to the surface of the collector. Where the collector 14 is planar, this will generally result in a planar sensor arrangement. However, in general terms, the sensor geometry may be adapted to the geometry of the apparatus, and in particular, to the geometry to the surface (or surfaces) with which it is operatively associated.

The sensor 21 is arranged to sense characteristics of the material layer 12 adjacent to the sensor 21. A sensing region 30 is shown in Figure 3 extending from the sensor 21 towards and around the collector 14, encompassing the layer 12.

Since the sensor 21 is configured to perform ECT, no direct electrical (or physical) contact is needed between the electrodes 25 and the material layer 12. However, it is preferred that the separation between the electrodes 25 and the material layer 12 is maintained at a small (and constant) distance (e.g. around 5 to 10 mm). Maintaining a small separation distance allows the signal to noise ratio to be increased. A small distance may be small relative to the overall size of the sensor. For example, a separation of around 10 mm may be considered to provide an acceptable compromise between ease of operation and providing an enhanced signal to noise ratio.

In view of the absence of any direct electrical contact, a coating layer may be applied to the electrodes 25 so as to protect the electrodes 25 from contaminating the coating process (or vice versa). However, such a coating may be thin (e.g. around 0.1 -0.2 mm), so as to minimise any electric field distortion, and should be made of a dielectric material. It will be understood that a conductive material (or even a material having a moderately high level of resistance) would, if applied continuously over and between the electrodes 25, allow conduction between the electrodes 25, and would effectively screen the electrodes from the sensing region 30 significantly reducing sensitivity. It is noted that while it may be advantageous to apply a coating over the electrodes in terms of minimising possible contamination, such a coating may have a negative effect on the signal quality. That is, any dielectric material will cause some distortion to the field established by electrodes during measurements. As such, the omission of any dielectric coating may provide optimal signal quality.

It will be appreciated that ECT is typically a voltage driven technique in which capacitance measurements are taken between two terminals (electrodes). In order to perform ECT, the energisation source 22 is arranged to apply an energisation to a first electrode 25a, and the detector 23 is arranged to detect a signal at a second electrode 25b. The energisation may, for example, comprise the application of an energisation signal comprising a sinusoidal signal to the first electrode 25a. The energisation signal may, for example, have a frequency of around 1 MHz and an amplitude of around 18 Volts peak-to-peak, although can be varied depending upon the characteristics of the materials being monitored.

A voltage signal is then detected at the second electrode 25b whilst the energisation signal is applied to the first electrode 25a. The amplitude and phase relationship between the induced voltage signal at the second electrode 25b and the energisation signal applied to the first electrode 25a is monitored and processed as described in more detail below.

It will be appreciated that the frequency and amplitude of applied energisation may be optimised for each application, for example based upon a geometry of the sensor 21 and/or the monitored material layer 12, by experimentation, and/or by reference to known properties of material components being processed. For example, where a material component being processed is known to have a particularly low permittivity, an energisation signal with an increased amplitude may be used. In general terms, the frequency and amplitude of the applied energisation signal may be selected so as to produce a detectable potential at the monitored electrodes in response to the applied energisation signal.

The amplitude and phase of the detected voltage are electrical parameters which can be measured in response to an applied energisation. It will be appreciated that other parameters may be measured. For example, in some embodiments only the amplitude of the detected voltage is measured.

In general terms, an energisation signal is applied to at least one of the electrodes which causes a field to be generated in a region (e.g. sensing region 30) extending beyond the surface of the sensor 21 towards the material. An electrical parameter is then measured at at least one of the electrodes in response to the applied energisation signal, generating a measurement data item. Measurement of the electrical parameter in response to the applied energisation signal may be referred to as detection of a signal. Finally, the measured electrical parameter (or measurement data item) is processed as described in more detail below.

Where a new sensor configuration is used, or where a sensor is re-configured so as to sense a different material (e.g. when switching between cathode and anode production), a calibration process may be performed. Such a process may enable an optimum energisation and detection process to be determined for a particular component material or material combination. Moreover, calibration (or re-calibration) may be performed at regular intervals and/or where a component material or component material property is changed.

Figure 5 shows the controller 24 in further detail. It can be seen that the controller 24 comprises a CPU 24a which is configured to read and execute instructions stored in a volatile memory 24b which takes the form of a random access memory. The volatile memory 24b stores instructions for execution by the CPU 24a and data used by those instructions. For example, in use, measured voltage values may be stored in the volatile memory 24b.

The controller 24 further comprises non-volatile storage in the form of a solid state drive 24c. Measured voltage values, data relating to the energisation of electrodes, and data relating to the sensor geometry may be stored on the solid state drive 24c. The controller 24 further comprises an I/O interface 24d to which are connected peripheral devices used in connection with obtaining the measured signals. More particularly, the energisation source 22 and the detector 23 are connected to the I/O interface 24d.

A network interface 24h allows the probe controller 24 to be connected to a computer network, so as to receive and transmit data from and to other computers in that network (e.g. via the Internet). The CPU 24a, volatile memory 24b, solid state drive 24c, I/O interface 24d, and network interface 24h, are connected together by a bus 24i.

The above described process of energisation and detection is repeated for each combination of the electrodes 25. Each of the combination of electrode pairs may be selected in turn. That is, when performing ECT, while an energisation is applied to the first electrode 25a, a signal is measured at each of the remaining electrodes 25b-25l in turn. An energisation signal is then applied to the second one of the electrodes 25b, and a signal is measured at each of the remaining electrodes 25c-25l in turn. This process continues until each electrode combination has been tested.

It will be appreciated that reverse combinations may be omitted. That is, if an energisation has been applied to a first electrode 25a and a signal measured at a second electrode 25b, there is no need to measure a signal at the first electrode 25a when the second electrode 25b is energised.

Thus, eleven measurements are taken for the first energisation, ten measurements are taken for the second energisation, and so on, until a single measurement is taken for the eleventh energisation. As such, an array of twelve electrodes has sixty-six unique electrode combinations that can be selected, and sixty-six measurement data items that can be generated. Each of those measurements may be repeated several times and an average taken to provide a degree of noise immunity. Of course, some combinations may be omitted (e.g. if less sensitivity is required, or higher speed processing is required). Similarly, alternative energisation protocols may be used.

Once each of the unique measurements has been taken (or while measurements are ongoing), the controller 24 processes the measurements (also referred to as measurement data items) to generate a value indicative of the capacitance associated with each particular electrode combination. That is, a set of capacitance values can be generated which are indicative of the capacitive coupling between each pair of electrodes within the electrodes 25. Each of the generated capacitance values will depend upon the particular geometry of the sensor 21 , and properties of any material in a region within the sensing region 30 generally adjacent to and/or extending between the particular electrodes involved in the measurement.

By performing measurements using different electrode pairs, it is possible to generate information relating to the distribution of matter in the region around the sensor 21 , and in particular to generate data indicative of the permittivity of material adjacent to the sensor 21 , with a degree of spatial resolution. That is, the known geometry of the sensor 21 , and the geometrical relationship between the various electrodes 25 is processed in combination with the capacitance data to generate an ECT image which represents the permittivity of material adjacent to the sensor 25.

Prior to such processing, a model may be used to simulate an electric potential distribution as a function of dielectric permittivity at each sensed location based upon the sensor configuration, sensor geometry and excitation pattern. Such modelling may be referred to as using a “forward problem”, and may use finite element analysis to generate a mesh of values indicative of a relationship between the charge at a particular surface position, an applied excitation, and a dielectric permittivity at each sensed location.

Such a forward problem can then be processed using an inversion model, so as to create a so-called “inverse problem”. The use of an inverse problem, or a reconstruction model (also referred to herein as model data) allows dielectric permittivity information to be extracted (or reconstructed) from a set of measurements indicative of induced charge at surface locations in response to the applied excitation. That is, the reconstruction model takes into account sensor configuration, sensor geometry and an excitation pattern, and allows permittivity data to be obtained from measurement data. This processing may, for example, be carried out as described in “Evaluation of planar 3D electrical capacitance tomography: from single-plane to dual-plane configuration”, Measurement Science and Technology, Volume 26, Number 6, by Hsin-Yu Wei, Chang-Hua Qiu and Manuchehr Soleimani, which is herein incorporated by reference.

In this way, the measured data may be processed so as to generate a visual tomogram, which represents the distribution of the permittivity of material within sensing region 30. That is, the measured data is processed to generate data indicative of a characteristic the layer of material 12 coated onto the collector 14 during the coating process shown in step S3.

The generated tomogram has a plurality of pixels. Each of the pixels in a tomogram has an associated signal strength value. The signal strength value associated with each pixel is indicative of the permittivity of material at a corresponding two-dimensional location. Where no material is present, this signal would have a value of zero (provided suitable calibration has been carried out). However, this value rises as the quantity (or thickness) of material at the location corresponding to the pixel increases. In this way, tomograms can be used to visually represent the quantity of material within the sensed region of the layer 12, an increase in permittivity (as compared to free space) being indicative of an increase in layer thickness.

In the arrangement described above, the sensor 21 comprises a two-dimensional array of 12 electrodes. Sixty-six unique measurements are conducted, corresponding to sixty- six capacitance values, each representing a capacitive coupling between a respective pair of the electrodes 25. A single set of capacitance values may be processed, as described above, to generate a permittivity tomogram comprising 1024 pixels arranged in a 32x32 array. Of course, other tomogram dimensions can be used as appropriate.

It will be understood that in many applications of tomography an increased imaging depth is of critical importance. For example, in a key application of tomography the technique is used to provide detailed information relating to the flow of material within the centre of a pipeline. However, in the present application, that is the application of tomographic techniques to the characterisation of relatively thin layers of material deposited onto a planar surface, imaging depth may be of limited interest. In particular, any material present will necessarily be adhered to or resting upon the surface of the collector. Thus, detection depth can be relatively small, and a sensor can still provide useful diagnostic information.

The sensor 21 is arranged to look out into the sensing region 30 immediately adjacent to the sensor 21 , and to allow visualisation of properties of the material in that region based upon the detected signals. The sensor 21 may, for example, have an effective measurement depth (i.e. in a direction normal to the surface of the sensor 21) of around 30 mm. Thus, any material coated to or resting upon the surface of the collector 14 will fall within this region, provided the sensor is suitably located. A sensing region may typically extend around 20 % of the depth of the sensor dimension.

It will also be appreciated that whereas a two dimensional tomogram may be produced, and the signal strength value associated with each pixel is indicative of the permittivity of material at a corresponding location, each signal strength value may be influenced by material within the sensitivity depth range. That is, each signal strength value may be an average of permittivity of material within a region extending up to the sensitivity depth from the sensor surface. It will further be appreciated that the sensor sensitivity may be greatest at the surface of the sensor, and may reduce as the distance from the surface of the sensor (i.e. towards the collector 14) increases. Thus, each pixel value may be dominated by the effect of material closest to a corresponding location at the surface of the sensor 21.

Figure 6 shows schematically an arrangement in which a region of the collector material 14 shown in Figure 3 is coated with electrode material 12. In a first sub-region 12A the electrode material has a good coating, whereas in a second sub-region 12B the electrode material has a poor coating. The poor coating is considered to be poor (i.e. of an unacceptable quality for use in a battery) since the coating thickness is too thin, and the uniformity is poor (i.e. unacceptable thickness variation within the sub-region). On the other hand, the good coating is coating considered to be good (i.e. of an acceptable quality for use in a battery) since the coating thickness is at the target thickness, and the uniformity is good (i.e. acceptable thickness variation within the sub-region). ECT sensor 21 of the general type shown in Figure 4 is arranged adjacent to the collector material 14 in such a way that the sensing region spans both of the sub-regions 12A, 12B. Figure 7a shows a tomogram which is generated as an output of the processing described above, with the arrangement of Figure 6. The tomogram intensity varies between the sub-regions 12A, 12B, with a thicker and more uniform film being evident in the region 12A as comparted to the region 12B. The tomogram shown in Figure 7b shows the same data, but with regions having approximately the same intensity having similar hatching.

It will be appreciated that in Figure 7a and 7b (and each of Figures 11a to 12b - which are described in more detail below), the intensity shown corresponds to permittivity, and that increased permittivity represents an increased thickness of material adjacent to the sensor 21 , where the material in question has a relative permittivity that is greater than air. Thus, the intensity values shown are indicative of thickness of material adjacent to the sensor 21 , and are generated based upon measurements taken from the electrodes 25.

Figure 8 shows a representation of the thickness of the material in the sub-regions 12A and 12B as derived from the permittivity data shown in Figures 7a and 7b. As can be seen clearly, the thickness is generally greater in first sub-region 12A than the second sub-region 12B.

It can be seen from Figures 7a, 7b and 8 that material thickness can be monitored by the use of ECT sensing techniques using a planar sensor 21 .

The coating apparatus 10 of Figure 3 may be controlled based on a number of different process parameters. One such process parameter is a coating speed (e.g. speed of movement of the collector 14). A further process parameter is a coating gap (e.g. a gap between the slot die - or, in other embodiments, a comma bar or doctor blade - and the collector surface, or a component - e.g. a roller - supporting the collector). A further process parameter is web tension (e.g. a tension in the portion of the collector material extending between rollers as it passes the coating location). A further process parameter is a temperature (e.g. a temperature of the electrode slurry, the collector material, and/or the ambient temperature in which the process is performed).

In the illustrated configuration, the coating apparatus 10 comprises slot die 11. Alternative coating apparatus configurations may use a different type of coater, such as a doctor blade or a comma bar. Further, or alternatively, a tensioned web arrangement may be provided. Each of the various apparatus configurations may have additional or alternative process parameters that can be adjusted in real-time, or selected at the time of process commissioning. For example, a slot width may be selected (or even adjusted) for a slot die coater. Similarly, a pump flow rate may be varied to adjust the rate at which electrode slurry material is supplied from the slot.

In an embodiment, one or more process parameters (e.g. the process parameters described above) may be controlled or adjusted based on an output of the monitoring apparatus 20. For example, during a coating process, thickness data may be obtained at regular intervals (e.g. at a sampling frequency of around 1 Hz). In other examples, higher frequency sampling may be used (e.g. if required by a high speed production line).

If it is detected that a thickness is at an unacceptable level (e.g. to thick, or too thin) a process parameter (e.g. coating gap) may be adjusted in order to correct the thickness. That is, the coating thickness may be controlled based upon data indicative of the characteristic of the material (e.g. thickness data). To perform such control, the coating apparatus may comprise a process controller (not shown) configured to receive the thickness data (or other data indicative of a characteristic of the coated material), and to control the process based on the received data. In some embodiments, the process controller may comprise part of the controller 24.

Alternatively, or additionally, if a measured characteristic indicates that a quality issue may exist, a warning (e.g. a siren or visible warning) may be generated, allowing visual inspection of the process to be performed. In a further alternative, if a measured characteristic indicates that a critical quality issue exists, production (or other processing) may be suspended to allow defective components to be removed from the production cycle or processing line.

Figure 9 shows a process running on the controller 24 in which a coating process is monitored and controlled. At step S10 a first one of the electrodes 25 is energised according to a predetermined measurement protocol (e.g. as described in more detail above). At step S11 , an electrical parameter measurement (e.g. voltage) is performed at a second one of the electrodes 25 and a measurement data item is generated. At step S12 it is determined if the sequence of measurements is complete. If not, processing returns to step S10 where an electrode is energised (either the same one or a different electrode, depending on the measurement sequence), and a further measurement is performed at step S11. In this way, a predetermined measurement sequence is performed (e.g. comprising 66 measurements - if the above described protocol is followed for a sensor having 12 electrodes).

Once it is determined at step S12 that the sequence of measurements is complete, processing passes to step S13, where the plurality of measurement data items is processed to generate dielectric permittivity data D1. The permittivity data D1 may be generated based upon the electrical parameter measurements obtained at step S11 , in combination with model data D2. The model data D2 takes into account sensor configuration, sensor geometry and an excitation pattern, and allows permittivity data to be obtained from measurement data, as described in more detail above. In other words, the plurality of measurement data items are processed based upon the model data D2 to generate the permittivity data D1.

The model data D2 allows the geometry of the plurality of electrodes to be taken into account, as well as sensor configuration (e.g. number and arrangement of electrodes), sensor geometry (e.g. overall size and shape of the sensor), excitation pattern (e.g. the identity of the electrodes at which each energisation signal was applied, and each electrical parameter was measured), and energisation signal (e.g. details of the frequency and amplitude of the energisation signal).

Once the permittivity data D1 has been generated, at step S14 material characteristic data D3 is generated. The material characteristic data D3 may be generated based upon the permittivity data D1 and reference data D4. For example, reference data D4 may comprise data indicative of a bulk permittivity of the electrode material 12, or permittivity of components of the electrode material 12, and composition information. Such data may be referred to as material reference data. As such, by combining the permittivity data D1 with reference data D4, it is possible to determine a characteristic of the material 12, such as, for example thickness. Alternatively, or additionally, the reference data D4 may comprise data indicative of the composition and/or rheology of the material. The material reference data may comprise material temperature data, material homogeneity data and/or material flow rate data.

It will be appreciated, however, that, the permittivity data D1 obtained by techniques described above may itself be indicative of a material characteristic, even if not combined with reference data D4. For example, while absolute thickness (or other characteristic) data may not be available if the dielectric properties (or other reference data) of the materials under measurement are not known, changes or variations in the distribution of the material (either in the direction of transport 15, or across the width of the collector 14) can be determined without reference data D4. Thus, in some circumstances, the use of reference data may be omitted.

Alternatively, or in addition, the reference data D4 may comprise calibration data. For example, the reference data D4 may comprise data previously obtained by the sensor 21 (or a similar sensor) based on a reference sample under ideal conditions (e.g. with a known, or target material thickness). The characteristic data D3 generated at step S14 may thus be based on a comparison of the permittivity data D1 with calibration or expected permittivity data. Such a comparison may allow variations in the material characteristic (e.g. thickness, distribution) to be identified, and isolate them from measurement artefacts (e.g. aspects of the permittivity data that are indicative of the collector, or of the measurement environment).

The characteristic data D3 may comprise a characteristic (e.g. thickness) tomogram, or another representation of the characteristic of the material (e.g. as shown in Figure 8). However, there is no requirement for the characteristic data D3 to comprise a visual representation.

The characteristic data D3 is then assessed at step S15, to determine if it meets a predetermined criterion. For example, the criterion may be a thickness range and/or uniformity. If the characteristic is acceptable (“Yes”), processing passes back to step S10 where further monitoring is performed, and the coating process continues.

Alternatively, if it is determined at step S15 that the characteristic is not acceptable (“No”), processing passes to step S16, where a process control parameter is modified, in order to improve the coating process. Processing then passes back to step S10, where further monitoring of the modified coating process is performed.

In this way, by performing real-time monitoring and adjustment of the coating process, it is possible to provide a reliable consistent material characteristic, and to minimise the risk of poor quality electrode layers being manufactured.

It will, of course, be appreciated that the process described above with reference to Figure 9 is one example of a way in which a coating process may be monitored and controlled. In some examples, no control modification is performed at step S16. Instead (or as well), characteristic errors may be flagged to an operator, and/or production may be suspended. Alternatively, characteristic data may be stored, and associated with the produced electrode for subsequent use. In some circumstances, steps S10 to S14 may be performed as a characterisation method, with steps S15 and S16 being omitted entirely.

Similarly, in some examples, the permittivity data may be used directly as the basis for comparison or assessment, rather than data that has been converted to units of thickness (or other measure of characteristic). That is, since it may be known that a desired characteristic (e.g. thickness) of material will have a particular permittivity characteristic (via reference data D4) the assessment at step S15 may simple consist of a comparison between the permittivity data and reference data D4, with a deviation threshold being applied (either in terms of uniformity, or average value).

While the processing described above generally reference to thickness, it will be appreciated that other quantitative measures of an amount or distribution of material may be used. For example, the data indicative of a characteristic of the material may comprise mass data, that is, data indicative of a mass of material present at a particular location (e.g. the mass of material deposited on a substrate). Data indicative of a mass of material may be generated based on thickness data and material density data (e.g. slurry density data).

Figure 10 shows schematically an alternative arrangement in which a region of collector material is coated with electrode material in a layer. The coating apparatus and material layer are generally as described above with reference to Figure 3. Like parts are shown with like reference numerals, and are will not be described again. A first ECT sensor 42 is arranged adjacent to the surface of the collector material 14 upon which the coating 12 is applied. A second ECT sensor 43 is arranged adjacent to the opposition surface of the collector material 14 to which the coating 12 is applied. The first and second ECT sensors 42, 43 are of a similar size and shape, and are arranged to face each other, both being substantially parallel and separated from the material 12 and collector 14. Such an arrangement may be referred to as a dual-plane ECT sensing arrangement.

Each of the ECT sensors 42, 43 are connected to an energisation source 22 and a detector 23, which may be generally of the sort described above with reference to Figure 3. It will be understood that when operated individually (e.g. as described above, with up to sixty-six different measurement combinations), the ECT sensors 42, 43 will each have a respective sensing region extending towards the collector 14.

By providing dual-plane sensors, it is possible to perform additional measurements. For example, by providing two sensors each having twelve electrodes, twenty-three measurements can be taken for each electrode energisation. If reverse measurements are excluded (as described above), the dual-plane configuration using two arrays of twelve electrodes has 276 unique measurements, each measurement being indicative of the capacitive coupling between a pair of the electrodes.

Once this measurement data has been obtained, processing can be performed as described above in order to generate permittivity data. That is, the measured data can be processed by a reconstruction model so as to generate a high-resolution visual tomogram (or other form of characteristic data), which represents the distribution of material within a sensing region 44 extending between the sensors 42, 43. By using two sensors in this way, it is possible to generate visual tomogram, (or a non-visual representation of the same data) with a higher resolution image than would be possible using a single measurement plane.

The permittivity data can be converted, if required, into thickness data, with each pixel representing a small sub-region within the sensing region 44. By providing high- resolution imaging, thickness uniformity and defect detection can be enhanced. Moreover, whereas the sensing depth of a single measurement plane can be limited, dual-plane sensing offers significantly enhanced depth sensitivity. Further, dual-plane sensing allows a measurement system to discriminate between a coating applied to a front side of the collector 14 (e.g. material layer 12) and a coating applied to a rear side of the collector 14 (not shown). In this way, double sided materials can be characterised.

Of course, it will be understood that different sub-sets of a set of measurements can be used in different ways. That is, a full set of 276 measurements contains 66 measurements relating to the first ECT sensor 42, 66 measurements relating to the second ECT sensor 43, and 144 inter-sensor measurements.

Indeed, it will also be understood that the time required to obtain 276 measurements, and to process this data to obtain thickness data may be greater than for a single-plane sensing arrangement. In some circumstances, such processing may not be possible in real time. However, in an alternative operational configuration, electrodes can be energised according to a simpler measurement protocol in which measurements are taken between each opposite electrode pair, providing 12 measurements in total (for the described arrangement). Such a measurement process can provide a simple representation of capacitance at each of 12 sub-regions within the sensing region, with such raw data being obtainable at high speed. In some circumstances, such data may be sufficiently high resolution to provide useful control information.

Figures 11a and 11 b show a low resolution “raw” data tomogram obtained by the system shown in Figure 10 in which the detected thickness is relatively uniform. On the other hand, Figures 12a and 12b show a low resolution “raw” data tomogram obtained by the system shown in Figure 10 in which the detected thickness is relatively non-uniform, with the lower-right-hand quadrant showing a thickness below acceptable levels.

In both the high-resolution processing described above, and the lower-resolution processing described with reference to Figures 11a-12b measured electrical parameter data is used to generate data indicative of a characteristic the material deposited onto the collector 14 during a coating process. The processing described with reference to Figure 9 can be applied to either high-resolution processing in which model data is used to reconstruct permittivity data from measurement data, or low-resolution processing in which capacitance measurements are directly obtained from “raw” data. In such processing, step S13 may be omitted.

Further, while low-resolution processing is described with particular reference to a dualplane sensing arrangement, it may also be performed using a single-plane sensor. For example, raw data may be obtained from an adjacent pair of electrodes and used to provide high speed, but low resolution, information regarding material characteristics immediately adjacent to that electrode pair. Such sensing may typically be used to supplement higher resolution sensing, but can also be used independently.

As described above with reference to Figure 2 the battery electrode manufacturing process typically comprises several processing steps. The coating process performed at step S3 has been described in some detail above. The drying process performed in step S4 will now be described in more detail with reference to Figure 13.

After the material layer 12 has been coated (e.g. by coating apparatus 10) onto the collector 14, the collector 14 is moved in a direction 51 into a drying apparatus 50. The drying apparatus 50 may comprise a drying chamber 52. A heat source 53 is provided to allow the drying temperature within the chamber 52 to be controlled. An air supply 54 is provided to the drying chamber 52. Air extraction 55 is also provided to extract the air, which may carry solvent vapour, from the drying chamber 52. The extracted air may be provided to a recovery unit which separates air from the solvent vapour. Air can then be returned via air supply 54 to drying chamber 52. Recovered solvent vapour may be passed to a solvent recovery unit before being returned to the coating apparatus 10 for re-use. Several drying chambers may be provided, allowing a variety of different temperature zones to be provided.

As described above with reference to Figure 3 a monitoring apparatus 20 may be provided shortly after the coating apparatus 10. Such a monitoring apparatus is shown in Figure 13 as monitoring apparatus 20a.

It will also be appreciated that additional or alternative monitoring stages may be implemented. For example, a further monitoring apparatus 20b may be positioned after the drying apparatus 50 so as to monitor the dried material. That is, a first monitoring apparatus 20a may be provided after the coating apparatus 10 but before the drying apparatus 50, with a second monitoring apparatus 20b being provided after the drying apparatus 50. In this way, it will be possible to monitor a change in material characteristics as the material has passed through the drying apparatus 50. Monitoring in this way may enable the moisture content of the electrode material to be monitored in real-time. More particularly, a reduction in the moisture content (e.g. due to the removal of solvent) of the material can be determined.

Monitoring in this way also allows the drying process to be controlled. For example, in a similar manner to that described above with reference to figure 9, various aspects of the drying apparatus 50 may be controlled based upon a monitored material characteristic. To perform such control, the drying apparatus may comprise a process controller (not shown) configured to receive data generated by the first and /or second monitoring apparatus 20a, 20b, and to control the process based on the received data. In some embodiments, the process controller may comprise part of the controller 24.

One such process parameter that could be varied is drying time, which could be modified by varying the material speed. Alternatively, or additionally, a further process parameter that could be varied is the temperature in one or more of the drying chambers, or the rate of air input and/or extraction.

After the drying process has been performed the material progresses to the calendaring process described above with reference to step S5. Figure 14 illustrates the calendaring process in more detail. A calendaring apparatus 60 comprises a pair of rollers 61 , 62. The current collector material 14 is provided with the coated layer 12 of electrode material (as described above). The collector 14 along with the coating 12 is advanced in a direction 65 towards and between the rollers 61 , 62. The rollers 61 , 62 are provided on opposing sides of the collector 14 and are configured to press the collector material and the coated electrode material 12 together so as to compress the electrode material 12. Once the material has passed through the calendaring apparatus 60 a thin uniform layer of material is provided which can then be further processed by being cut and packaged as described above.

The calendaring process is intended to reduce the coating porosity, thereby allowing the energy storage density to be optimised. The (second) monitoring apparatus 20b described above with reference to Fig. 13 may be provided before the calendaring apparatus 60. A further (third) monitoring apparatus 20c may be provided at an outlet of the calendaring apparatus 60.

That is, a monitoring apparatus which is similar to the monitoring apparatus 20 described with reference to figures 3 and/or 10 may be provided after a material has passed through the pressing rollers 61 , 62. When such a monitoring apparatus 20c is used in conjunction with additional monitoring either at the output of the coating stage (i.e. monitoring apparatus 20a) or at the output of the drying apparatus 50 (i.e. monitoring apparatus 20b) it may be possible to provide a comparison between characteristics of the material 12 at different stages. In this way, it is possible to adjust various control parameters of the calendaring process so as to improve the electrode characteristics.

A characteristic which can be monitored at the output of the calendaring stage is the electrode material porosity. A process parameter which can be controlled based upon the monitored characteristic is the pressure applied by the two rollers 61 , 62. A further process parameter that can be controlled and adjusted is the speed of material through the calendaring apparatus 60. To perform such control, the calendaring apparatus may comprise a process controller (not shown) configured to receive data generated by the monitoring apparatus 20c and to control the process based on the received data. In some embodiments, the process controller may comprise part of the controller 24.

It will be appreciated that it is possible to monitor the process at different positions: e.g. after coating (20a), after drying (20b), and after calendaring (20c). By comparing data obtained at different monitoring points, it is possible to obtain contrast information between wet and dry layers (e.g. by comparing moisture content at different points), offering information relating to the drying process. It is also possible to provide contrast information between uncompressed and compressed layers (e.g. by comparing porosity at different points), offering additional monitoring parameters relating to the calendaring process.

In each case, processing similar to that described with reference to Figure 9 may be performed. It will of course be appreciated that different reference data D4 may be used at each stage of the process. For example, at an initial monitoring stage (e.g. directly after the coating has been applied), reference data relating to the material properties of the electrode slurry (e.g. the slurry permittivity, and/or slurry density) may be used in order to allow the permittivity data D1 to be converted into thickness data (or some other measure of a material characteristic).

Thereafter, when characterising a drying process, characteristic data D3 obtained by processing according to Figure 9 can be used as reference data for a subsequent process to allow an effective comparison between wet and dry layers to be made. When characterising a drying process, it may be noted that the amplitude of voltage signals obtained from electrodes varies in proportion to the moisture content.

In addition to the obtained electrical parameter measurements, additional reference data may be used in the calculations performed at this stage in order to obtain more detailed characteristics relating to the dry material characteristics, or relating to the drying process.

Similarly, data obtained from the dry (but uncompressed) layer (e.g. data obtained by the monitoring apparatus 20b) may be used as reference data for a monitoring process that is performed after the calendaring operation has been performed. When characterising a calendaring process, it may be noted that the amplitude of voltage signals obtained from electrodes may vary in inverse proportion to the porosity.

Generally speaking, it will be appreciated that two or more monitoring processes may be performed on the same portion of material at different points within a multi-stage manufacturing or material processing process, with data indicative of a particular process (e.g. a coating process, a drying process, a calendaring process, etc.) being generated by processing (e.g. comparing) the data generated at the output of each of the monitoring processes. Moreover, permittivity data D1 and/or characteristic data D3 obtained by performing the processing as described above with reference to Figure 9 may be used as an input (e.g. as reference data D4) for a similar process performed by another monitoring apparatus at a later process stage.

In addition to any of the processing described above, it will of course be appreciated that additional processing steps may be included. For example, additional storage and/or cooling processes may be included between one or more of the above described processes. In this way, it may be possible for the different processes to be controlled to a certain extent independently from one another. Of course, where a continuous processing line is in operation, it will be understood that altering the speed of one process may have consequences for other processes. As such, it may be preferable either to provide a variable length storage facility between adjacent processes, or to avoid adjusting certain process parameters that would otherwise influence upstream or downstream processing.

Furthermore, alternative monitoring and processing steps may be performed before or after the processes described above. For example, additional process monitoring may be performed to provide information regarding a composition or rheology of the material 13 entering the coating apparatus 10. Such a measurement may be performed by apparatus substantially as described in WO2017/077293, which is herein incorporated by reference. In particular, data indicative of the rheology of a material within the mixing apparatus may be obtained. Such data may be used to control the mixing process and/or to control process parameters of subsequent coating, drying, or calendaring processes. In this way, rheological information may be used to provide feed-forward control of subsequent processing steps.

Of course, it will be appreciated that whereas the rheology monitoring device described above may be configured substantially as described in WO2017/077293, alternative configurations may be used. For example, rather than first and second sets of electrodes having different sensitivity depths, a single set of electrodes may be used and configured to provide sensing within a mixing apparatus.

Additional monitoring may also be performed using one or more secondary sensors at various points during the processing sequence. For example, one or more of the monitoring techniques described above (e.g. ultrasound, backscattering, infrared) may be used to provide an input (e.g. as reference data D4) to the process described above with the reference to Figure 9. In this way, the sensitivity and/or accuracy of ECT monitoring can be improved. For example, in some circumstances, an independent secondary measurement can be made which provides secondary measurement data, the secondary measurement data comprising an accurate measurement of a material characteristic (e.g. material thickness, weight, moisture content, porosity, etc.). This secondary measurement data can then be used to refine characteristic data D3 generated at step S14. The secondary measurement data may be obtained in any convenient way, and many comprise an accurate local measurement (e.g. at a single point) or a large area measurement (e.g. a single measurement representing a large area). In either case, such data can allow a tomogram to be calibrated more accurately.

Furthermore, data relating to the slurry properties (e.g. composition, permittivity, temperature, viscosity, homogeneity, flow rate) may be provided by monitoring apparatus provided within the slurry mixing apparatus, or other inputs (e.g. sensors). Such data may be used as reference data to be input to the downstream processing performed by monitoring apparatus 20 after coating or subsequent processing steps. In this way, up-to-date material properties can be used as a basis for subsequent material characterisation. This avoids the need to use potentially inaccurate previously obtained reference data.

In a further alternative or additional arrangement, rheological monitoring may be performed by using ECT. That is, it is not necessary to use electrical resistance tomography (ERT) to monitor the slurry properties.

In a further alternative arrangement, data obtained by one or more monitoring steps performed after the coating, drying, and/or calendaring processes may be used to control or assess processing performed in an upstream process, such as, for example, the mixing apparatus. In this way, it is possible that defective processing can be identified and/or avoided, since issues such as, for example, improper mixing, may be identified by non-uniform coating thickness, and adjustments made to the mixing process to improve the homogeneity of the slurry. In this way, enhancements of the overall process can be made while waste can be minimised.

As described above, an ECT sensor 21 and associated controller 24 can be used to detect and characterise material present upon the collector 14. Such processing can be carried out in real time, and with relatively inexpensive general purpose processor. In this way, a simple and direct determination of the quality of a processed layer of material at a particular location can be made. Moreover, the monitoring is non-invasive and nonintrusive. That is, monitoring can be carried out while processing is ongoing, and without interrupting or affecting the processing. The real time nature of the monitoring can allow continuous monitoring to be carried out, for example with a refresh rate (or frame rate) in excess of 1 Hz.

Although reference has been made to processes involved in the manufacture of batteries, these are intended to serve as non-limiting illustrative examples. In another example, tomographic methods of the present invention can be used to qualify materials post-manufacture.

Polymers have many industrial applications, for example when provided in the form of polymer foams. For example, PET foam is a material having many industrial applications, including wind turbine applications, railroad applications, structural composite manufacture and shipping applications, amongst others. PET foam may range in density from 80-235 kg/m ³ .PET foam is typically supplied in sheets of thicknesses between 20- 70mm, having planar dimensions of 1220mm x 2440mm. PET foam may also be supplied and/or manufactured as a continuous extrusion of varying dimensions.

Although PET is porous, larger voids within the structure may constitute an unacceptable defect. It is desirable to manufacturers, therefore, to provide a rapid and non-destructive monitoring process to perform qualification and/or quality control on sheets of PET foam, such that defective sheets can be rejected at the manufacturing facility.

Polymer foam in a continuous extrusion could be monitored by a monitoring apparatus generally similar to the monitoring apparatus 20 of Figure 3 or Figure 10 (with appropriate modifications to the geometry to accommodate the material being monitored), largely as described with reference to battery manufacture. Polymer foam sheets could also be monitored by an apparatus generally similar to the monitoring apparatus 20 of Figure 3 or Figure 10, provided the sensor 21 or sensors 42 and 43 can span the planar extent of the polymer foam sheet to be monitored.

Given that detection and characterisation of internal defects (e.g. voids) is desired through the depth of the material, data indicative of characteristics of the material at a plurality of locations within a three-dimensional sensing region adjacent to, or extending from, the array(s) electrodes may be obtained. Data indicative of characteristics of the material within a three-dimensional sensing region (e.g. sensing region 30) may comprise three-dimensional tomograms and/or a plurality of two-dimensional tomograms. Three-dimensional tomograms and/or a plurality of two-dimensional tomograms allow one or more of the presence of a void, the configuration of a void, the position of a void and the size of a void within the material to be determined. In some situations, involving characterisation of internal defects (e.g. voids), a dual-plane sensor apparatus (e.g. monitoring apparatus 20 of Figure 10) may be preferred, so as to provide for an enhanced depth sensitivity.

In cases where the sensor of a monitoring apparatus is smaller than the area of material over which monitoring is desired to be performed, a dynamic monitoring location, as opposed to the static monitoring locations of Figure 3 and Figure 10, may be provided.

The sensor 21 of Figure 3 and sensors 42 and 43 may be referred to as static monitoring locations, in the sense that they do not move relative to the material - the material being monitored is moved or conveyed past the monitoring location. A dynamic monitoring location, on the other hand, may be understood to mean a monitoring location that is moved, by moving the two-dimensional planar array of electrodes relative to the static material in a plane parallel to the material. In some embodiments, a dynamic monitoring location may also be used where the material being monitored is also moved or conveyed past the monitoring location.

Figure 15 illustrates an alternative monitoring apparatus 100 (which may also be referred to as a qualification apparatus) arranged proximate to a polymer foam sheet 101 with an internal void 112. Qualification apparatus 100 comprises a sensor 21 , position controller 102 and a movement superstructure 104. The movement superstructure 104 itself comprises a sensor carriage 106, planar actuators 108 and a separation actuator 110. The sensor 21 is coupled to the sensor carriage 106. The sensor carriage 106 is coupled to the planar actuators 108 and the separation actuator 110.

In some embodiments, the planar 108 and separation actuators 110 may comprise one or more position-controlled motors.

Although omitted from Figure 15 in the interests of clarity, it will nevertheless be appreciated that, in use, the sensor 21 is connected to an (external) energisation source, detector and controller of the sort described above with reference to Figures 3 and 10. Actuation of the planar actuators 108 induces movement of the sensor carriage and sensor 21 substantially parallel to the plane of the polymer foam sheet 101 . The planar actuators 108 induce movement along mutually non-parallel directions allowing the sensor to traverse the entire extent of the polymer foam sheet 101 (as shown by the horizontal arrow and dashed lines of Figure 15). Put another way, the sensing region 30 passes through the entirety of the foam sheet 101 . Actuation of the separation actuator 106 advances and retracts the sensor in a direction substantially normal to the material.

The planar actuators may be controlled by the position controller 102 so as to traverse the polymer foam sheet in a raster-like fashion, as illustrated in Figure 16. Figure 16 shows a plan view of the sensor 21 of the qualification apparatus 100 superimposed on a polymer foam sheet over a selected number of positions, P1-P6. The movement superstructure 104 and position controller 102 have been omitted in the interests of clarity. The sensor 21 first traverses in ‘rows’ from P1 to P6, in the manner indicated by the arrows of figure 16. At P6, the entire sheet has been traversed. It will be appreciated that in some cases, the relative dimensions of the sensor 21 and the polymer foam sheet may require overlap of the sensor positions.

Due to manufacturing tolerances and/or misalignments of the polymer foam sheet 101 among other causes, the surface of the polymer foam sheet 101 may deviate from its nominal plane. Tomographic measurements are highly sensitive to sensor-material separation distance. It is thus desirable to provide a system maintaining a pre-selected sensor-material separation A (best seen in Figure 15). The qualification apparatus may thus further comprise a distance sensor 111 which indicates the sensor-material separation A such that the controller can actuate the separation actuator to maintain a constant separation distance A. In an example, this might be achieved by means of a closed loop feedback or a feedforward control scheme.

The distance sensor may be optical, infrared or ultrasonic, although alternative forms of sensor may be selected (e.g. mechanical).

In some embodiments, sensor motion may be only provided in one or two directions. In the case of static monitoring locations, for example as described with reference to Figures 3 or 10, sensor motion may be provided in a single direction. The single direction of movement may be to advance and retract the sensor in a direction substantially normal to the material (i.e. separation control), the movement being induced by a separation actuator of the sort described with reference to Figure 15.

In other embodiments, the planar sensor may fully span one dimension of a sheet material under qualification, in which case, the planar sensor may be required to move along one direction to traverse the sheet material, in addition to the separation control movement described above.

It will be appreciated that in some embodiments, for example where internal defects (e.g. voids) are monitored for, it may be desirable to provide the qualification apparatus with a dual-plane sensor of the sort described with reference to Figure 10.

In the following section, the detection of a void using the monitoring apparatus 20, which is a dual plane ECT apparatus, as described with reference to Figure 10 will be described. The processing performed may be generally similar to that described above with reference to Figure 9 (albeit without any process control feedback).

In order to successfully distinguish voids from the surrounding material, calibration may be performed. The calibration may be based on suitably obtained calibration data. For example, calibration data may be obtained by placing materials having a known, or uniform permittivity within the sensing region. In one example materials having the highest and lowest capacitance values expected to be encountered may be provided. During calibration a uniform sample of such a material is placed between the first and second sensors.

It will be appreciated that calibration for a particular set of process conditions may only have validity over similar process conditions. Process conditions may vary in use. As such, multiple calibrations at various process conditions may be carried out. Process conditions may comprise temperature, moisture, sensor separation and product composition (amongst others). In some embodiments, the monitoring apparatus may further comprise process condition indicators, configured to determine the process conditions at any given time, such as temperature or moisture sensors. Process condition indicators may be used to inform adjustments in sensor calibration, according to variations in process condition. As such, monitoring apparatus may dynamically compensate for process variation. Figure 17 shows a series of two-dimensional tomograms or ‘slices’ parallel to the Y-Z plane, obtained from a dual-plane monitoring apparatus (e.g. similar to the monitoring apparatus 20 shown in Figure 10) in which a low capacitance material was placed between the sensors. In this case, the low-capacitance material was air. The resulting tomogram is homogeneously low-capacitance, as indicated by the near-uniform shade.

Figure 18 shows a series of two-dimensional tomograms or ‘slices’ parallel to the Y-Z plane, obtained from a dual-plane monitoring apparatus in which a high capacitance material was placed between the sensors. In this case, the high-capacitance material was a sample of PET foam with no voids or other defects. The resulting tomogram is homogeneously high-capacitance, as indicated by the near-uniform shade.

Figure 19a schematically illustrates a cross section of a sample 113 of PET foam placed between the sensors 115a, b of a dual-plane monitoring apparatus. Sample 113 has a vertically aligned (i.e. parallel to the z-axis) void defect 114. Note that the void 114 extends into the page, forming a substantially bore-like defect, aligned with the z-axis, extending through more than half of the PET foam sample’s z-extent.

Figure 19b shows the resulting series of two-dimensional tomograms 118, 120, 122, 124, 126 obtained by the dual-plane monitoring apparatus. A region of low capacitance 121 can be seen in tomogram 120, clearly indicating the presence and location of void 114. It will be appreciated that shape and size of the region of low-capacitance 121 is indicative of the three-dimensional extent and configuration of the void 114.

Figure 20a schematically illustrates a cross section of another sample 128 of PET foam placed between the sensors 130a, b of a dual-plane monitoring apparatus. Sample 128 has a horizontally aligned (i.e. parallel to the z-axis) void defect 132. The void 132 has a bore-like configuration and extends through the x-extent of sample 128.

Figure 20b shows the resulting series of two-dimensional tomograms 134, 136, 138, 140, 142 obtained by the dual-plane monitoring apparatus. Regions of low capacitance 137, 139 and 141 can be seen in tomograms 136,138 and 140 respectively, clearly indicating the presence and location of void 132. It will be appreciated that shapes and size of the regions of low-capacitance 137, 139 and 141 are indicative of the three-dimensional extent and configuration of the void 132.

Data indicative of a characteristic of the material (e.g. permittivity data, as shown in the tomograms described above, or on one or more of: thickness data, mass data, moisture content data, density data, porosity data, etc. which has been derived from the permittivity data) may be further processed by a processor (e.g. controller 24, or another controller), in order to generate qualification data. The processor may be part of the monitoring apparatus, or provided at a remote location.

The qualification data may comprise data indicating at least one of: the presence of a void or defect within the material, the configuration of a void of defect within the material, the position of a void or defect within the material and the size of a void or defect within the material. Such processing may comprise image-processing, and may including subprocesses such as, for example, edge detection. An identified region of lower capacitance (e.g. below a threshold value) may be identified as being indicative of the presence of a void. The location of the identified region may be used to indicate the likely position of a void. The configuration of the identified region may be used to indicate the likely configuration of a void (e.g. the direction along which a void extends). The size of the identified region may be used to indicate the size of a void.

While such processing may be performed by a processor, it may also be performed by an operator, viewing data or visual representations of the data (e.g. tomograms) generated by the monitoring apparatus in real time. The operator may identify one or more of the presence of a void or defect within the material, the configuration of a void of defect within the material, the position of a void or defect within the material and the size of a void or defect within the material.

The qualification data may further be used as an input to a process control system. For example, products or materials that are identified as being defective may be removed from the processing line automatically, in real-time. Alternatively, a process parameter may be adjusted in real-time based upon the qualification data, so as to attempt to remedy the defect. Qualification data may also, or alternatively, be associated with a product (e.g. by being associated with a unique product identifier in a digital record), so as to allow off-line checks to be carried out. In this way, a quality control process may be performed, allowing defective products to be identified.

It will be appreciated that the above-described methods may find application on any number of other planar materials, and are not restricted to polymer foam, or even polymer materials

The above-described methods may be applied to planar polymer materials in general, not just PET foam. For example, the material under tomographic qualification and/or characterisation may be any kind of polymer (e.g. polyethylene), whether in an expanded foam state or not.

Another application of the above-described method is the detection of surface contamination (e.g. dust or chemical residue) on silicon wafers, for example in semiconductor manufacturing. The mechanical tolerances and cleanliness standards of such applications may benefit from tomography sensors with higher spatial resolution.

The above-described method may also be applied to the inspection of walls in structures and buildings, for example in residential houses. In an example, the above described method might be applied to detect both surface and sub-surface moisture in walls.

Where applied to a fixed material such as the walls of a building, a mobile monitoring apparatus may be provided, which may be manually traversed over a material of interest. In other embodiments traversing may be assisted by powered motion.

Data can be generated that is indicative of a characteristic of the material, with a spatial resolution of a few millimetres. For example, in some of the above described methods, a 32x32 pixel tomogram is generated representing an overall area of around 150 mm x 150 mm. In such an arrangement, each pixel represents around 4.7 mm x 4.7 mm of material. Where such processing is repeated at regular intervals (e.g. at 1 Hz), the evolution of the material characteristics over time can be closely monitored. Indeed, knowledge of the speed of transport of the material relative to the sensor 21 allows a particular region of material to be tracked and monitored at various points in time. That is, first data indicative of a characteristic of the material (e.g. a first tomogram) may be generated at a first time and second (or further) data indicative of a characteristic of the material (e.g. a second tomogram) may be generated at a second time. The first and second tomograms may be generated by different monitoring apparatuses

Alternatively, or additionally, a one-dimensional tomogram (or non-visual representation of the data) may be generated (e.g. by a sensor having linear - i.e. one-dimensional - array. Such an output may be referred to as a “slice”. Data may be re-obtained from the sensor at regular intervals, and new “slices” generated. By combining the slices, a two- dimensional tomogram can be reconstructed. In this way, a smaller sensing footprint can still provide two-dimensional information, with the dimension in the direction of travel provided by virtue of movement and time delays, rather than by the inherent dimensionality of the sensor.

As described above, data indicative of a characteristic of the material (e.g. permittivity, thickness, mass, solvent content, porosity, etc.) can be generated based upon a plurality of electrical parameter measurement data items. While various examples of types of characteristic data have been identified, it will of course be appreciated that other forms of data indicative of a characteristic of the material may be used as appropriate for a particular process or monitoring task.

Such data indicative of a characteristic of the material may enable secondary data to be generated, or process insights to be obtained. For example, permittivity data can be processed to determine a measure of thickness variation. Moreover, permittivity data can be processed to identify defects, (e.g. voids). Such processing may involve identifying regions of the material where a predetermined criterion is satisfied. For example, it may be determined that a defect exists if the thickness (or permittivity, or other variable indicative of a characteristic of the material) in a first sub-region differs from that in a second sub-region by more than a predetermined amount or percentage. Such processing may involve a preliminary step of generating thickness, or density, data. However, the permittivity data itself can allow comparisons to be made between relative thickness or density, meaning that conversion to absolute thickness or density data is not essential.

The permittivity data and/or other characteristic data can also be used to identify various types of defect. Examples of types of defect may be longitudinal thickness variations (i.e. thickness variation along the direction of material movement) or lateral thickness variations (i.e. thickness variation transverse to the direction of material movement). A further type of defect may be periodic deposits or variations in thickness. A further type of defect may be agglomerations (in either direction), for example where a lump of crosslinked material is present. Such defects may cause quality issues at downstream processing locations. Further types of defect include coating gaps or voids (e.g. where there is an absence of slurry on a supporting substrate, or a three dimensional void within a thicker solid material) and streaks (e.g. where slurry is absent in a film). Identification of such defects may enable suitable remedial action to be taken. Different ones of the defects identified above may have different required actions. For example, lateral variation issues may be improved by adjustment of web tension, whereas longitudinal variation may indicate slurry homogeneity problems.

Further, it may be determined that a plurality of defects exist (e.g. based on the processing described above), and additionally determined that a defect density (or count) is, or is not, within an acceptable range.

Further still, data indicative of a characteristic of the material (e.g. permittivity, thickness, solvent content, porosity, density, etc.) generated by processes described herein may be used to generate data indicative of a characteristic of the material or device being monitored (e.g. battery electrode, or polymer foam component, etc.). For example, data indicative of a characteristic of the battery electrode may comprise electrode performance data (e.g. that an electrode will be defective, or that an electrode will provide a certain storage capacity).

Moreover, while the processing described above has generally been described in the context of the manufacture of battery electrodes, it will be appreciated that alternative uses and applications for the ECT characterisation techniques described herein are possible.

For example, while the described embodiments relate to coating, drying, and calendaring processes, the techniques are generally applicable to any form of processing apparatus in which a layer of material that is supported by a substrate (e.g. the collector 14) can be processed, and in which real-time information regarding material characteristics is of use. Further, in some circumstances, rather than the material of interest being deposited on or supported by a substrate, characteristics of the substrate itself may be of interest. Moreover, in some circumstances, a web of material may be characterised by the above described techniques, without any need for a layer to be deposited thereon. By “web” it is meant a long, thin, and flexible material (e.g. a film or foil).

Further still, in some circumstances, the material of interest may have a periodic property. For example, the techniques described herein may be used to characterise a series of separate articles that are processed or manufactured, with an “image” of each article being captured by the ECT sensor as that article passes the monitoring location (e.g. on a conveyor belt).

Generally speaking, layers, webs, films or articles may be suitable for monitoring by the techniques described herein when they are relatively constant in size. That is, if there is a high degree of variation in the expected position of the monitored surface with reference to the sensor, it may be difficult to obtain useful measurement data. Similarly, if a process environment is highly unstable (e.g. in terms of temperature or vibration), it may also be difficult to obtain useful measurement data. On the other hand, in relatively stable environments, and where variations in surface position are relatively small and/or are due to the variations in process or product quality to be monitored, then the planar ECT sensing technique described herein can be particularly advantageous. In such circumstances, planar ECT sensing techniques can be used to assess product quality and/or consistency.

The techniques described herein may also provide particular advantages when applied to an apparatus which is required to remove a material from a substrate, when it is desirable to determine the extent to which a material layer or deposit has been removed. For example, the monitoring apparatus may be configured to verify that cleaning has been effective in removing material deposits, or to verify that a patterning process has been effective in removing parts of a material layer.

Moreover, if prior knowledge exists of the permittivity of materials being processed (or components thereof), such reference data can be used to improve the accuracy of the determination, for example to provide a quantitative measure of the thickness or density of material present (rather than a simple indication that some material is present). However, as described above, such reference data is not essential.

The techniques described herein may be incorporated into a processing apparatus or system. Alternatively, a monitoring apparatus may be provided which can be configured to characterise materials applied by (or otherwise processed by) one or more different processes, subject to suitable calibration.

It will be understood that where a plurality of sensors is used (e.g. in a dual plane arrangement, as shown in Fig. 10), each of the sensors may be addressed using a single set of processing electronics (i.e. a single controller, a single energisation source and a single detector) using a set of switches or multiplexers to allow sequential addressing of each array. Alternatively, multiple sets of processing electronics may be provided. Similarly, separately described monitoring apparatus 20a, 20b, 20c may share some components (e.g. controller 24) in certain implementations.

Further still, the location of the one or more of the monitoring apparatuses may be selected so as to coincide with areas of the processing apparatus which were particularly susceptible to deviations in material characteristics, or which are critical to the performance of the final part.

While it is described above that an array of electrodes are provided in a single or dualplane arrangement adjacent to the material under consideration, it will be understood additional, or further electrodes may also be provided. For example, a conductive material substrate may act as an electrode in some measurements. In this way, an energisation signal could be applied to one of the sensor electrodes, and a response signal measured from the substrate, or vice versa. Similarly, an electrode could be provided on a surface of the substrate (e.g. a back surface, opposite to the surface on which the material is deposited.)

While it is described above that a two dimensional tomogram may be generated based upon the sensed permittivity data, in some instances a three-dimensional tomogram may be preferred. For example, where a sensor has a complex geometry, or the material of interest does not constitute a relatively thin film or coating, it may be preferred to visualise the permittivity in three dimensions rather than two, where such a representation is considered to be more suitable to represent the permittivity distribution of the material layer 12. For example, where a thick layer of material (e.g. a polymer foam) is monitored, a three-dimensional tomogram may be preferred, with the third-dimension providing information regarding the depth of the material, in addition to two dimensions which extend in a plane of the material surface. Where such processing is performed, each element within the tomogram will be represented by a ‘voxel’ (i.e. an element of volume that represents a notional three-dimensional space) rather than a pixel.

It will, of course, be appreciated that the particular characteristics of each application can be taken into account when determining the number, type and location of sensors. Similarly, the particular characteristics of each application can be taken into account when determining the reliance placed upon the data generated based upon the detected data. That is, while in some applications a single sensor as described above may provide sufficient confidence that a material or material layer is of sufficient quality, in some applications it may be preferable to use several sensors in key locations to provide increased confidence in any determination that the layer and/or material is as expected. Similarly, in some applications it may be desirable to use a sensor (or several sensors) as described above in combination with an alternative sensing or inspection technique.

It will be appreciated by one of ordinary skill in the art that the invention has been described by way of example only, and that the invention itself is defined by the claims. Numerous modifications and variations may be made to the exemplary design described above without departing from the scope of the invention as defined in the claims. For example, the precise shape, configuration, and number, of the various components may be varied.