Classifications

• • G—PHYSICS
  • G01—MEASURING; TESTING
  • G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
  • G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
  • G01N21/62—Systems in which the material investigated is excited whereby it emits light or causes a change in wavelength of the incident light
  • G01N21/63—Systems in which the material investigated is excited whereby it emits light or causes a change in wavelength of the incident light optically excited
  • G01N21/64—Fluorescence; Phosphorescence
  • G01N21/6489—Photoluminescence of semiconductors
• • G—PHYSICS
  • G06—COMPUTING; CALCULATING OR COUNTING
  • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
  • G06T7/00—Image analysis
  • G06T7/0002—Inspection of images, e.g. flaw detection
  • G06T7/0004—Industrial image inspection
• • G—PHYSICS
  • G01—MEASURING; TESTING
  • G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
  • G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
  • G01N21/62—Systems in which the material investigated is excited whereby it emits light or causes a change in wavelength of the incident light
  • G01N21/66—Systems in which the material investigated is excited whereby it emits light or causes a change in wavelength of the incident light electrically excited, e.g. electroluminescence
• • G—PHYSICS
  • G01—MEASURING; TESTING
  • G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
  • G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
  • G01N21/84—Systems specially adapted for particular applications
  • G01N21/88—Investigating the presence of flaws or contamination
  • G01N21/95—Investigating the presence of flaws or contamination characterised by the material or shape of the object to be examined
  • G01N21/9501—Semiconductor wafers
• • H—ELECTRICITY
  • H01—ELECTRIC ELEMENTS
  • H01L—SEMICONDUCTOR DEVICES NOT COVERED BY CLASS H10
  • H01L27/00—Devices consisting of a plurality of semiconductor or other solid-state components formed in or on a common substrate
  • H01L27/14—Devices consisting of a plurality of semiconductor or other solid-state components formed in or on a common substrate including semiconductor components sensitive to infrared radiation, light, electromagnetic radiation of shorter wavelength or corpuscular radiation and specially adapted either for the conversion of the energy of such radiation into electrical energy or for the control of electrical energy by such radiation
  • H01L27/144—Devices controlled by radiation
  • H01L27/146—Imager structures
  • H01L27/14601—Structural or functional details thereof
  • H01L27/14603—Special geometry or disposition of pixel-elements, address-lines or gate-electrodes

Definitions

• the invention relates to a method and a device for electroluminescence inspection and / or for photoluminescence inspection of a luminescent object, for example.
• a PN semiconductor in particular a solar cell or a solar module, wherein the object by applying a voltage or more generally by a electrical action, to which, for example, the action by an electric field belongs, and / or by irradiation of light to emit electromagnetic radiation, for example.
• an optical recording device in particular a camera, for example, an area camera and / or a line camera, in particular a digital camera, recorded and output as an image.
• the image is subjected to an image evaluation, in particular in a computing unit connected to the recording device.
• Possible errors of the object are determined in the image evaluation, in particular by checking the preferably digital, i. formed by individual pixels, image on typical error structures, which may be defined in particular by a spatial extent and / or intensity.
• the arrangement for receiving the luminescence-excited object is arranged in particular in a darkened receiving chamber, since the luminescence effects have only a low light intensity.
• the method proposed according to the invention and the corresponding apparatus for luminescence inspection can be described in electroluminescent and / or photoluminescence applications both in manufacturing processes, also in-line in a production line, and / or offline, for example.
• Particularly useful fields of use according to the invention are semiconductors, thin-film technologies and all types of substrates.
• a concrete example are photovoltaic structures, such as, for example, solar cells or solar modules.
• the invention is not limited to this specific example.
• Luminescent imaging inspection uses the structure and properties of luminescent-capable objects, for example, solar cells based on a PN semiconductor junction, and the behavior of minority carriers on the P (positive) and N ( negative) side of the semiconductor exploits.
• the holes (positive carriers) are in the minority.
• the P side of the luminescent object electrons are the minority carriers. If, by applying a transition voltage, the minority carriers diffuse through the luminescence-capable object, a current is generated in the luminescence-capable object, which leads to the emission of electromagnetic radiation from the luminescence-capable object.
• Images of this radiation provide information on the quality and / or the properties of the luminescence-capable object, which has been explained using the example of a semiconductor element, even if the invention is not limited to such objects and, for example, also for the investigation of luminescence-capable other Objects, for example.
• General lattice structures is suitable.
• a particularly preferred application relates to a luminescence inspection of solar cells, or more generally photovoltaic substrates, which are monocrystalline, quasi-monocrystalline or multicrystalline Si solar cells and / or solar photovoltaic cells. modules, thin films or concentrated solar cells (CPV - concentrating photovoltaics) may be formed.
• the photovoltaic substrate hereinafter also referred to as a solar cell
• the photovoltaic substrate is brought to emit electromagnetic radiation by means of an external excitation, for example the application of a voltage, due to diffusion of the minority charge carriers, the electromagnetic radiation emitted by the solar cell can be received by optical sensors or recording devices which have a sensitivity to electromagnetic radiation of a wavelength above 800 nm.
• optical sensors or recording devices which have a sensitivity to electromagnetic radiation of a wavelength above 800 nm.
• the intensities of the luminescent images which are preferably recorded with low-noise and high-sensitivity cameras in the near infrared range (ie wavelengths between about 800 nm to 2500 nm), are proportional to the number of minority carriers in each region of the solar cell or of the luminescence-capable object and therefore allows conclusions to be drawn on the quality of or error within the luminescence-capable object.
• Known methods with conventional Si-CCD cameras are not suitable for high production rates due to slow image acquisition since they are only able to inspect a maximum of about 1400 solar cells per hour. This is due to the long exposure and readout times of the cameras.
• the evaluation algorithms are often also limited to a particular material, i. for example, on mono-, multi- or quasi-monocrystalline structures or thin-film layers.
• the object of the invention is therefore to propose a method which is preferably able to recognize and distinguish the individual types of errors more reliably at an increased processing speed. This object is achieved by a method having the features of claim 1 and a device having the features of claim 10.
• the detection of the electromagnetic radiation by the recording device ie the recording of the object after or during the luminescence excitation, in at least two images in different spectral ranges, ie in different recorded wavelength ranges occurs.
• at least one (multispectral) double image of the luminescence-excited object with different wavelengths is provided, so that certain errors assigned to certain wavelengths can be identified and identified better and more reliably. This also enables a more secure classification of the error candidates, and in particular distinguishing intrinsic and extrinsic errors of the object.
• the method can also be used in the same way for electroluminescence methods and photoluminescence methods, it being possible, if necessary, for the spectral ranges to be adapted.
• spectral ranges filters for different wavelengths or wavelength ranges (spectral ranges) are used, which are synchronized by a filter changer with the detection of the electromagnetic radiation, ie with the recording of the images, before Receptacle changed or replaced.
• any spectral range in the entire sensor-active region of the recording device can be selected flexibly, whereby the bandwidth of the wavelength range can also be selected selectively by means of suitable filters.
• the filter changer may have a filter guide, in which the various filters are accommodated and which are displaceable relative to the receiving device or the optics of the receiving device such that the optics of the receiving device views and receives the luminescence-excited object by a respective other filter ,
• a wave range between 800 nm and 1800 nm can be covered by the inspection, ie by selecting the suitable filters, this wavelength range, which is also characterized as a near infrared range, can be detected in principle.
• This area is, for example, particularly suitable for the inspection of photovoltaic substrates.
• suitable filters for the at least two recordings according to the invention can cover a wavelength range around approximately 150 nm and a wavelength range around 1500 nm. It is therefore possible to use a low-pass filter at 150 nm and a high-pass filter at 1500 nm.
• the bandwidth of the filters may, for example, be such that wavelength ranges of about 900 to 150 nm in the one filter and wavelength ranges of about 1300 to 1600 nm are transmitted in the other filter.
• the corresponding images complement each other in their information regarding the behavior in two different wavelength ranges. As a result, the types of errors occurring precisely in photovoltaic substrates can be distinguished well and thus classified.
• the method proposed according to the invention is used in-line during a production process, it is usually necessary to reduce the inspection period for an inspection to such an extent that the inspection of the production process is not hindered.
• the recording of only two different images in different wavelength ranges already allows a significant improvement in quality over the prior art, usually with only one shot a wavelength range of 800 nm to 1 100 nm with less sensitive sensors in the peak range of the luminescence of solar cells covers.
• an InGaAs camera indium gallium arsenite camera
• InGaAs cameras have significantly improved sensitivity, especially in the short-wave infrared wavelength range (SWIR - Shortwave-IR), for example between 800 nm and 2000 nm.
• SWIR - Shortwave-IR short-wave infrared wavelength range
• the significantly better sensitivity leads to a shorter image acquisition time and an improved signal Ratio, so that on the one hand the inspection time for a recording significantly shortened and on the other hand, a better quality of recording is achieved.
• identified error candidates are evaluated in the further picture taken in a spectral range different from the spectral range of the first evaluated picture.
• several further images with different further spectral ranges can be used instead of the one further image.
• Such a procedure accelerates the evaluation in other areas, since preferably those image areas are to be evaluated, which have already shown abnormalities in the first image.
• error candidates and information about the individual error candidates can be collected via the various images and, according to the invention, passed on to an error classification, which, like the image processing, can be performed in a computing unit connected to the recording device.
• error candidates determined from the first image are passed on to the second image. If, in the second image, error candidates are added that have not yet appeared in the first image, these are passed on to a possibly existing further image or the error classification, which represents a kind of final error processing. If, on the other hand, an error candidate of the preceding (first) image can be safely excluded in a subsequent (second) image, then this error candidate no longer needs to be examined further in the further images.
• This procedure can be applied analogously to any number of images. It is furthermore particularly advantageous to collect and represent the occurrence of errors, in particular of classified errors, possibly resolved with the error class, in a spatial error distribution over time (ie an error frequency distribution). This allows for early detection of process errors.
• the classification of individual errors can be carried out using methods of artificial intelligence, based on learned and classified examples of error types as data basis.
• the classification classifies the final categorization of the defect types from the candidates identified in the images as a crack or column, shunt, finger interruption, series resistance, dark region, inactive region, firing deffections, dislocation, hot spot , Scratches or contour defects are classified. These collected quality criteria can also be used to predict the electrical classification of the solar cell into different quality classes.
• This calibration can be carried out in particular program-controlled, automatically and / or with the cooperation of the user.
• the contrast and standard deviation of a plurality of rectangular image regions are used to describe the sharpness.
• the program user can ask the user to change the sharpness settings of the object. manually or automatically slowly from one extreme position to the other extreme position. The program saves the optimum focus value and outputs an error message if the sharpness adjustment is too fast. In a second pass, the user is then asked to repeat this process while the application compares the current sharpness obtained with the stored best sharpening value and indicates that the best focus point has been reached, or automatically stops further adjustment. Alternatively, this can also be done by a controller without user intervention.
• a configuration file is modified to adjust the resolution information. This can be done automatically or by the user.
• the average reproduction rate (resolution) is calculated in the context of the calibration by means of a calibration target, whereby certain parameters are automatically adjusted or possible sources of error, for example a dirty target or a wrong distance between the object plane and the camera, are given as advice if the determined reproduction rate does not meet the desired requirements.
• shading images can be calculated as part of the calibration, the visible shading being caused mainly by the shading of the camera lenses.
• a simple solution for generating shading images is the use of the illumination curves presented by the lens manufacturer which describe the percentage reduction in image brightness from the image center to the edges. To align the optical axis to the center of the image, appropriate offsets can be used. If such lens information is not available, images of spatial cells can be recorded according to the invention. Subsequently, a model with a training data set of several random points, which specifies the location (X, Y) and the intensity, is trained, whereby the training sentence consists only of picture elements of the object. This training set may be additionally filtered to take into account luminescence-specific effects, such as lower light emission at the edges of the cells.
• the proposed method according to the invention is particularly suitable for the inspection of monocrystalline, quasi-monocrystalline or multicrystalline Si solar cells or solar modules, thin-film solar cells or layers and / or concentrated solar cells (CPY - concentrating photovoltaics, ie solar cells with a concentration of incident light (Sunbeams) through lenses, such as. Fresnel lenses).
• the invention also relates to a device for electroluminescence inspection and / or photoluminescence inspection of a luminescent-capable object, for example a PN semiconductor, in particular a solar cell or a solar module, with a device for exciting an electroluminescent device - Neszenz and / or a photoluminescence in the object.
• a device for exciting an electroluminescence may in particular be a current or voltage supply or a device for generating an electric field or the like.
• a device for exciting a photoluminescence can, in particular, be a lighting device.
• the device has a receiving device and a particularly movable object carrier for holding and possibly transporting the object.
• the slide may, for example, be a conveyor belt.
• a computing unit for controlling the device and for the evaluation of provided by the receiving device taken pictures of the luminescence excited object.
• a filter changer with at least two filters of different spectral ranges is arranged between the object and the recording device, wherein the different filters can be positioned in front of the recording device so that the object can be picked up by the recording device respectively by one of the different filters, ie can be recorded.
• the arithmetic unit is preferably set up to carry out the above-described method or parts thereof, in particular by suitable program code means which, when executed on the arithmetic unit, execute the method described according to the invention.
• a particularly preferred recording device may be an InGaAs camera with a particularly sensitive sensitivity range in the short-wave infrared range (shortwave infrared), i. especially at wavelengths between about 800 and 2000 nm.
• one filter selects a wavelength range around 1 150 nm and another filter a wavelength range around 1500 nm, ie has a corresponding spectral range around these wavelengths in which the electromagnetic radiation can pass.
• the proposed method thus proposes multispectral dual luminescence imaging and inspection of objects, in particular solar cells or solar modules, in the context of the transition energy of photovoltaic substrates in different wavelengths between 800 nm and 1800 nm.
• an automatic filter changer is described, which contains several, at least two, filters for high-speed applications that can be synchronized with the recording device.
• FIG. 1 shows the schematic structure of an apparatus for electroluminescent inspection according to a preferred embodiment.
• FIG. 2 shows schematically the image acquisition with a device according to FIG. 1;
• FIG. 3b shows images obtained with a device according to FIG. 1 in a second spectral range
• FIG. 4 a shows an image obtained with the device according to FIG. 1 in a first spectral range
• FIG. 4b shows an image obtained with a device according to FIG. 1 in a second spectral range
• FIG. 5a shows a picture taken in the device according to FIG. 1;
• FIG. FIG. 5b shows a defect-free luminescence image reconstructed from the image according to FIG. 4a;
• FIG. and FIG. 6 shows an error distribution (frequency distribution) formed according to the invention.
• a preferred embodiment of the present invention is shown.
• This shows a device 1 for electroluminescent inspection of a luminescence-capable object 2, which is a solar cell in the illustrated example.
• the object 2 is introduced on a movable slide 3, which is designed as a conveyor belt and arranged in a production line, in a darkened receiving chamber 4, in which the solar cell 2 is positioned and connected to a device for electroluminescent excitation 5.
• contact elements 6 are provided in the darkened receiving chamber 4, which surrounds the solar cell 2 from two sides, i. contact at its P-layer or its N-layer, and excite the transition voltage of the PN junction on the solar cell 2 via a power supply 7 so that it emits electromagnetic radiation 8 in reversal of the actual photovoltaic effect.
• the electromagnetic radiation 8 represents infrared light in the wavelength range between approximately 800 nm and 1800 nm.
• This electromagnetic radiation 8 is picked up by a recording device 9, which is particularly preferably an InGaAs area camera and has a particularly high sensitivity for wavelengths between about 800 nm and 2000 nm.
• the images recorded by the recording device 9 are performed by a computing unit 10, which controls the entire recording process and evaluates the recorded images in the manner described in more detail below.
• two filters 11, 12 are arranged in front of the recording device 9, which filters through an automatic filter changer 13 which is synchronized with the recording device 9
• the computing unit 10 controls the one filter 1 1 or the other filter 12 positioned in front of the receiving device 9.
• the automatic filter changer 13 on a bspw. Also the foreclosure against scattered light serving Ge Anthony 14 synchronized with the receiving device 9 are moved.
• the filter 11 may be a low-pass filter in a spectral range around 1150 nm and the filter 12 a high-pass filter in the spectral range around 1500 nm in order to record low and high-spectrum images in different wavelengths for luminescence measurement at different energy transitions.
• the solar cell 2 shows the structure of the solar cell 2 and the process of image acquisition in a schematic flowchart.
• the solar cell 2 has a PN-type semiconductor 15, in which the minority charge carriers are respectively arranged on one side of the charge carrier and shown in Fig. 2 by small circles.
• On the semiconductor 15 is an anti-reflection and SiO 2 layer 16, which is interrupted by a front contact 17 forming tracks.
• On the backside of the PN semiconductor 15 is a back contact 18.
• the minority carriers in the semiconductor 15 diffuse and generate a current, so that the circuit 19 is closed and a current flows .
• a voltage is applied to the front contact 17 and the rear contact 18 so that the charge carriers diffuse and, as shown in FIG. 1, emit electromagnetic radiation 8.
• This process is shown in Fig. 2 by an outlined box. As a result, a captured luminescence image 20 is obtained.
• FIG. 3 a shows luminescence images 21 of two different solar cells 2 recorded with a low-pass filter, in each of which dark pixels marked by a white arrow are shown, which represent possible defects in the solar cell 2.
• FIG. 3 b shown on the right side of FIG. 3 a the same solar cells 2 are shown as images 22 recorded with a hachpass filter, which overall have a substantially darker structure and at the points marked with the arrow check the error candidates from the images 21 according to FIG 3a allow.
• Bright areas in the images 22 according to FIG. 3b indicate strong electron collectors (deep traps).
• FIGS. 4a and b show two further examples of images taken by a solar cell 2 in a first spectral range (FIG. 21) and in a second spectral range (FIG. 22).
• a multiplicity of Defect candidates as dark dots or dashes. These areas are identified, for example, as described below, and examined in more detail in the second image 22.
• the second image 22 was taken in the spectral range between 1350nm to 1600nm.
• material defects eg, dislocation
• due to material and not due to process errors appear brighter even though they appear darker in the spectral region of the first image 21 and have lower intensities as well as defect regions.
• the method according to the invention with at least two images in different spectral ranges. Due to the additional use of the InGaAs camera, the exposure time can be reduced to 5 ms (compared to 500 ms for conventional, known systems). The readout time is relatively long in conventional systems due to a necessary multiple triggering with 250 ms. In the device according to the invention, the read-out time for the recording device 9 can be reduced by the use of the InGaAs camera in the order of 33 ms.
• the recovery time of a camera is typically less than 40 ms, compared to more than 750 ms per image in conventional devices.
• the device according to the invention allows about 3600 solar cells to be inspected per hour, although two images per solar cell are taken. Conventional devices achieve (with only one shot per solar cell) only about 1400 cells per hour.
• the spring detection rate can be suppressed to below 0.2% compared to conventional systems with an error rate of more than 2%.
• the shorter recording time in the device according to the invention also leads to less stress of the solar cells 2 during recording, since they must be excited only for a shorter period of time for electroluminescence.
• FIGS. 5a and b show a particularly preferred option for locating possible error candidates in the images 20, 21, 22.
• Fig. 5a shows a captured image 20 in which a dark, elongated structure marked by a white arrow can be seen, forming an error candidate.
• the invention proposes to calculate from the recorded image 20 a defect-free luminescence image 23, as it should be.
• the recorded luminescence image 20 is then subtracted from the error-free, reconstructed luminescence image 23 as part of an image evaluation, whereby possible defect defect regions are automatically marked.
• These possible error ranges of a first recording are then further evaluated in the second and further pictures taken with other filters 11, 12 for an exact decision as to whether an error and possibly which error exists.
• a simple optical correction takes place by means of image processing, in which, in particular, a shading correction, a distortion correction and / or various digital image processing filters can be used.
• the second step then provides for the removal of the image areas which are not to be examined, which may in particular be the image background and printed circuit boards (busbars).
• the third step possible error candidates are determined and assigned from the first image 21 (FIGS. 3a, 4a).
• the first image may be a so-called near-infrared (about 900 to 150 nm wavelength) photograph in which most of the irregularities may be identified as darker spots.
• an error-free luminescence image 23 is reconstructed from the recorded image.
• a spectral image is generated from the recorded image 20 by means of a Fourier transformation, in which the error candidates can be assigned to specific frequencies. These supposed errors are removed from the spectral image by removing these frequency components. Subsequently, the spectral image is converted by a back-Fourier transform, i. the inverse Fourier transform, transformed back into an optical image, which then represents the reconstructed defect-free luminescence image. From this luminance image 23, the recorded image 20 is subtracted pixel by pixel. Areas with gray value differences that exceed a predetermined threshold are classified as error areas or error candidates.
• these error areas or error candidates are then forwarded to images of other spectral ranges and checked there, wherein the spectral regions of the second and the further images preferably have a greater wavelength than the first, examined in the third step image.
• some error candidates may be identified as dislocations showing brighter pixels in the second image.
• these images may then be deleted as the list of error candidates.
• the remaining i. not eliminating or defrauding error candidates to an error classification, which classifies the error based on the information collected in the various images and defines the error therewith.
• FIG. 6 shows a frequency distribution of errors (error distribution) in a specific region of a solar cell 2 by a spatial, over time accumulated distribution of errors.
• the dark areas show no or few errors, the bright area a medium error rate, and the darkening area a particularly high error rate close to 100%.
• the ambiguous gray coloration of the frequency distribution is due to the black and white representation of the drawings. In reality, different colors can be used here, so that a clear error rate can be recognized in the drawings.
• the frequency distribution 24 shows the relatively largest error frequency in the range of the white arrow added later in the frequency distribution 24. This indicates that a systematic process error might be present.

Landscapes

• Health & Medical Sciences (AREA)
• General Physics & Mathematics (AREA)
• Physics & Mathematics (AREA)
• Chemical & Material Sciences (AREA)
• Life Sciences & Earth Sciences (AREA)
• Analytical Chemistry (AREA)
• Biochemistry (AREA)
• General Health & Medical Sciences (AREA)
• Immunology (AREA)
• Pathology (AREA)
• Nuclear Medicine, Radiotherapy & Molecular Imaging (AREA)
• Engineering & Computer Science (AREA)
• Quality & Reliability (AREA)
• Computer Vision & Pattern Recognition (AREA)
• Theoretical Computer Science (AREA)
• Investigating, Analyzing Materials By Fluorescence Or Luminescence (AREA)

Applications Claiming Priority (2)

Publications (1)

Family

ID=48190525

Family Applications (1)

Country Status (6)

Families Citing this family (21)

Family Cites Families (4)

• 2012
  • 2012-05-09 DE DE102012104086A patent/DE102012104086A1/de not_active Withdrawn
• 2013
  • 2013-04-30 KR KR20147034487A patent/KR20150009576A/ko not_active Application Discontinuation
  • 2013-04-30 EP EP13718862.9A patent/EP2847577A1/de not_active Withdrawn
  • 2013-04-30 US US14/399,242 patent/US20150070487A1/en not_active Abandoned
  • 2013-04-30 WO PCT/EP2013/058999 patent/WO2013167428A1/de active Application Filing
  • 2013-04-30 CN CN201380024197.6A patent/CN104471383A/zh active Pending

Non-Patent Citations (1)

Also Published As

Similar Documents

Legal Events