METHOD AND DEVICE FOR SORTING SILICON BREAKS

Description

[0001] The invention relates to a method for sorting silicon fragments.

[0002] Polycrystalline silicon (polysilicon) is typically produced using the Siemens process (chemical vapor deposition). In this process, silicon filament rods (thin rods) are heated in a reactor by direct electric current, and a reaction gas containing a silicon-containing component (e.g., monosilane or halosilane) and hydrogen is introduced. The surface temperature of the filament rods is typically above 1000°C. At these temperatures, the silicon-containing component of the reaction gas decomposes, and elemental silicon is deposited from the gas phase as polysilicon onto the rod surface, increasing the rod diameter. Once a predetermined diameter is reached, the deposition process is stopped, and the resulting polysilicon rods are removed.

[0003] Polysilicon is the starting material for the production of single-crystal silicon, which is manufactured, for example, using the Czochralski process. It is also required for the production of multicrystalline silicon, for example, using ingot casting. Both processes require the polysilicon rods to be broken into fragments. These fragments are then typically sorted by size using separators.

[0004] The separation devices can be multi-stage screening machines that mechanically classify the polysilicon granules into different size classes. For example, US 6,375,011 B1 discloses a vibratory conveyor that enables classification into three size classes.

[0005] Improved separation, or even separation based on optical criteria, can be achieved using optoneumatic sorting systems. US 2007 / 0235574 A1 discloses such a system, which is connected downstream of a polysilicon comminution device. For size separation, the shadow area of ​​the comminution-prone polysilicon pieces projected onto a plane is used.

[0006] Furthermore, US 6,265,683 B1 describes an optopneumatic device for classifying semiconductor materials, wherein size separation is camera-assisted via the detection of a projection surface of the materials to be sorted. Optionally, the materials can also be sorted according to their surface properties.

[0007] Document US-A-2009 / 021751 discloses a method for sorting fragments, comprising the steps of singulating the fragments in a singulation area, capturing the projection surface of a fragment in a 2D profile plane with at least one first measuring device, capturing at least one height information with respect to a 2D profile plane with the measuring device, calculating the size of the fragment from the information, and controlling at least one deflection device depending on the calculated size, as well as the corresponding device.

[0008] Known methods based on two-dimensional (2D) projection of fragments using transmitted light have the disadvantage that the fragments are only captured from one side. Fragments can differ significantly in shape. It can happen that a fragment is oriented so unfavorably at the time its projection surface is captured that its longest dimension is located in front of or behind the projection surface (i.e., perpendicular to the projection surface). This results in a miscalculation of the fragment size (missorting).

[0009] This problem led to the objective underlying the invention, namely the provision of an improved sorting method in which the orientation of the broken material to be sorted plays only a subordinate role.

[0010] This task is solved by a method for sorting fragments, especially silicon fragments, comprising the following steps. Singling the fragments in a singling area, capturing the projection surface of a fragment in a 2D profile plane with at least one first measuring device, capturing at least one height information above and / or below the 2D profile plane with at least one further measuring device, calculating the size of the fragment from the projection surface and the height information, controlling at least one deflection device depending on the calculated size.

[0011] The additional measuring device makes it possible to supplement the projection surface of the fragment obtained with the first measuring device with height information. The measuring devices are preferably arranged in different positions around the fragment, thus viewing it from different angles. This allows information about the extent of the fragment above and / or below the profile plane (in the direction of the application axis (z-axis)) to be obtained, thereby preventing sorting errors. In this way, the separation quality is improved.

[0012] The fragments are preferably crushed polysilicon, e.g. crushed polysilicon rods from the Siemens process.

[0013] The separation of the fragments means, in particular, that the fragments are spaced apart from one another, i.e., no longer arranged one above the other or partially overlapping. This can be achieved, for example, by a vibrating motion on a conveyor belt. It is not absolutely necessary for the fragments to be arranged in a row (behind or next to each other) for detection by the measuring device.

[0014] Preferably, the first measuring device and at least one of the further measuring devices are photoelectric transmitted light or reflected light measuring systems with a detection range through which the fragment passes.

[0015] Preferably, the fragment passes through the detection area in free fall.

[0016] The incident light measuring system preferably includes a light section sensor and / or at least a camera system.

[0017] The light sectioning method used with the light section sensor is based on optical triangulation and requires relative movement between the sensor and the fragment. The fragment is illuminated linearly with a suitable light source, and the resulting light stripe is captured by an area scan camera (part of the sensor). The surface normals of the light source and the camera are tilted relative to each other at a triangulation angle.

[0018] The camera system can, in principle, consist of just one camera that uses ambient light as a light source to capture the projection surface of the book fragment. Preferably, however, the camera is supplemented by an external light source.

[0019] Furthermore, it could be a camera system for photometric stereo analysis. This is a method for analyzing the projection surfaces and reflections of an object's (fragment's) surface in three-dimensional (3D) space. Typically, an external light source is directed at the object and moved to obtain multiple images of the resulting lighting scenarios. For moving objects, such as a fragment in free fall, multiple cameras with different viewing angles are used instead of a moving light source. Alternatively, several light sources, each with a different illumination direction, and a camera, or a combination of several light sources and cameras, can be used.

[0020] The transmitted light measurement system preferably comprises a light barrier, a light curtain, or a light grid. Light beams (e.g., infrared) are transmitted from an optical transmitter (light barrier) or several spaced-apart transmitters (light grid / curtain) to the corresponding receivers. If one or more beams are interrupted, a signal can be sent to a control device, which can then trigger, for example, a deflection device. The resolution of light grids and curtains can be determined by the spacing of the beams. Typical examples of such measurement systems are optical micrometers, light band micrometers, profile projectors, CCD laser micrometers, and laser sensors with light barrier functionality.

[0021] According to a preferred embodiment, the first measuring device is a camera system, in particular comprising a light source and a camera for capturing the projection surface.

[0022] The further measuring device is preferably a light curtain or grating, especially if a camera system was chosen as the first measuring device.

[0023] Preferably, the acquisition of the projection surface and the acquisition of the height information are performed with a time interval of 0 to 100 ms, more preferably 0 to 50 ms, and most preferably simultaneously. A minimal time interval ensures that the position of the fragment, especially during free fall, does not change significantly.

[0024] The deflection device responsible for sorting the fragments can be a pneumatic and / or mechanical deflection device.

[0025] The pneumatic deflection device preferably comprises at least one nozzle from which gases (e.g., air, inert gas) or liquids (e.g., high-purity water) are expelled at pressures of 3 to 20 bar. Reference may be made in this regard to US 6,265,683 B1.

[0026] Preferably, the method according to the invention is an optopneumatic classification method.

[0027] The size of the fragment can be calculated using an evaluation device (e.g., software-supported process control system, e.g., MATLAB (MathWorks)), which is connected to both the measuring devices and the deflection device.

[0028] The fundamental measurement information obtained with the first measuring device is the projection surface of the fragment, which contains information about the fragment's shape, albeit only in 2D as a contour. From this projection surface, the evaluation device can derive various length measurements, particularly diameters, which allow conclusions to be drawn about the size of the fragment. Numerous methods are generally known for determining these length measurements or diameters. Examples include determining the diameter derived from an equivalence circle (perimeter equivalent diameter) or determining a Feret diameter. These are a whole group of parameters, all defined by the distance between two tangents to the contour of the projection surface in a specific measurement direction.

[0029] The height information obtained by the additional measuring device, e.g., a light barrier, can be, in particular, a height value (unit of length). The combination of the obtained measurement information is then, for example, a 3D point cloud consisting of the contour obtained with the first measuring device and a contour shifted by the height value.

[0030] The maximum extent can be calculated from the length of a vector between the points furthest apart from each other: Punktewolke als Vektor X = x 1 , y 1 , z 1 , x 2 , y 2 , z 2 , … , xn yn zn

[0031] Combination of all points relative to each other and calculation of the distance in 3D: L i j = xi − xj 2 + yi − yj 2 + zi − zj 2 0 , 5

[0032] Determining the maximum value for L(i,j).

[0033] The more detailed the height information from the additional measuring device, the smaller the measurement error will be.

[0034] Another aspect of the invention relates to a device for sorting fragments, in particular silicon fragments, comprising a singulation area for singulating the fragments, at least one first measuring device for capturing the projection surface of a fragment in a 2D profile plane, at least one further measuring device for capturing height information above and / or below the 2D profile plane, at least one deflection device, an evaluation device that calculates a size of the fragment from the projection surface and the height information and controls the deflection device depending on this size.

[0035] Preferably, it is a device for carrying out the described method.

[0036] The singulation area preferably comprises at least one vibrating conveyor trough and / or a conveyor belt. Optionally, the singulation area may also include a screen plate or a vibrating screen for removing fines. Generally, it can also be a section of a conveyor belt onto which the fragments are separated. Dust extraction may also be provided. Preferably, it is a singulation area as described in EP 0 983 804 A1.

[0037] Preferably, the device comprises a plurality of first and / or further measuring devices to enable high-throughput sorting. By combining a plurality of measuring devices with the high computing power of the evaluation device, the distance between individual fragments can be reduced to a minimum.

[0038] Preferably, it is a pneumatic or mechanical deflection device, in particular a pneumatic deflection device comprising a series or a matrix of individual nozzles.

[0039] Regarding the measuring devices and the evaluation device, reference can be made to the above statements and EP 0 983 804 A1. Fig . 1: Device according to the invention with light grid (transmitted light measuring system). Fig. 2: Device according to the invention with light section sensor (incident light measuring system). List of reference symbols used

[0040] 10 Vibrating conveyor belt 12 Movement arrow 20 Sliding surface 22 End of sliding surface 30 Fragment 32 Projection surface 40 Camera 42 Light source 44 Detection area 50 Light grid 52 Transmitter strip 54 IR beam path 56 Receiver strip 60 Light section sensor 62 Laser scanner 63 Static laser line 64 Receiving optics 65 Reflected light 70 Deflection device 80 First collection container 81 Separating element 82 Second collection container 90 Evaluation device 100 Sorting device 200 Sorting device

[0041] The Figure 1Figure 1 shows a sorting device 100 comprising a vibrating conveyor belt 10 as a singulation area and an inclined sliding surface 20. The feed direction of singulated fragments 30 is indicated by arrows 12. A camera 40 with an external light source 42 is arranged below the sliding surface 20 as a first measuring device. The camera 40 is, for example, a CCD camera with an optical resolution of 0.05 to 2.0 mm. The light source 42 is, for example, an LED with diffuse area illumination. A detection area 44 of the first measuring device, relative to the fragment 30, is indicated by a star. A light grid 50 is attached to a lower end 22 of the sliding surface 20 as a further measuring device.This consists of a transmitter strip 52 with a total of five infrared light sources (laser or LED light sources or light points in the visible range are also conceivable), the radiation of which is indicated by a dashed line 54, and a receiver strip 56, which accordingly has five sensors. Below the first measuring device is a pneumatic deflector 70, and below this are a first and a second collection container 80, 82. The collection containers 80, 82 are connected to each other by a separating element 81 with a triangular cross-section. Furthermore, the sensor strip 56 of the first measuring device, the camera 40 and light source 42 of the second measuring device, and the deflector 70 are all connected to an evaluation device 90. The evaluation device is a computer with image processing software, e.g., MATLAB.

[0042] When a fragment 30, for example pyramid-shaped, which has been isolated by the vibrating motion of the vibrating conveyor belt 10, reaches the inclined sliding surface 20, it aligns itself so that its center of gravity is as low as possible. This can generally be adjusted to the size of the fragments by means of an angle-adjustable sliding surface 20. After passing the end 22 of the sliding surface, the fragment 30 passes the light grid 50 such that its elongated side points in the z-direction and is thus captured along its entire length by the light grid 50. Subsequently, the fragment 30 falls freely through the detection area 44 of the camera 40, which captures only a projection surface 32 corresponding to the base area of ​​the fragment 30.From both pieces of information, the projection surface 32 and the height information obtained through the light grid 50, the evaluation device 90 calculates the size of the fragment 30 and transmits this information to the pneumatic deflection device 70, whereby fragments 30 that are not deflected are collected in the second collection container 82 and fragments 30 that are deflected by a pneumatic impulse are collected in the first collection container 82. The separating element 81 facilitates this separation.

[0043] The Figure 2 shows another sorting device 200 according to the invention, which is essentially the same as that shown in the Figure 1Corresponding elements have corresponding reference symbols; the evaluation device, deflection device, and collection containers are not shown. The sorting device 200 includes a light section sensor 60 as a further measuring device. This consists of a laser projector 62 and a receiving optic 64. The light section sensor 60 uses the triangulation principle to determine the height information (3D profile acquisition). A laser beam is expanded into a laser line 63 (represented by the fine dashed lines) via special optics and projected onto the surface of the passing fragment 30. The receiving optic 64 images the light 65 reflected from the surface (represented by the coarse dashed lines) onto a sensor matrix.An evaluation device (not shown) connected to the light section sensor 60 can calculate the height information of the passing fragment (z-axis) from the matrix image obtained along the laser line (x-axis). This information can then be output in a sensor-fixed two-dimensional coordinate system. For moving objects or when the sensor is traversed, 3D measurements can also be obtained. Comparative example 1

[0044] Classifying crushed (broken) polysilicon with a fracture size (BG) of 2.

[0045] The size class of polysilicon fragments is defined as the longest distance between two points on the surface of a silicon fragment (corresponding to the maximum length): BG00,1 to 5 mm BG13 to 15 mm BG210 to 40 mm BG320 to 60 mm BG445 to 120 mm BG5100 to 250 mm

[0046] The polysilicon sample material used for the test was prepared from a mixture of 9000 fragments with a length range of 10 to 40 mm (BG2) and 1000 fragments with a length range of > 40 to 65 mm, the fraction to be separated. A mechanical sieving method (analytical sieve according to DIN ISO 3310-2 with a nominal hole size W = 4 mm (square perforation)) was used to prepare the sample material and separate the 0 to 10 mm fragment fractions. The maximum length of both the 10 to 40 mm fragments and the 40 to 65 mm fragments of the fraction to be separated was determined manually (using calipers), and the polysilicon sample material was then blended.

[0047] This polysilicon sample material was subjected to conventional optoneumatic sorting of the > 40 mm fraction.

[0048] The optopneumatic sorting device used was equipped with a first 2D measuring device (CCD camera and light source according to Fig. 1 The device was equipped to detect a projection surface of the fragments. The fragments passed through the detection area of ​​the measuring device individually (typically with a distance of 0.5–10 mm between fragments) via an inclined sliding plane. The measuring device was coupled to an evaluation unit at a process control station. The deflection device, comprising a nozzle bar approximately 50 cm wide with two rows of 100 nozzles each, was also connected to this control station. A cut-off width of 40 mm was set at the control station.

[0049] Following sorting, the slippage of the separation process was determined by manual analysis (calipers) in conjunction with a manual counting of the fragments. This amounted to 0.1%. Example 1

[0050] The polysilicon sample material described in the comparative example was separated using an essentially identical opto-neumatic sorting device (with a cut-off width of 40 mm). However, unlike comparative example 1, this device had a second 2D measuring device. This was a light grating similar to that used in... Fig. 1 described.

[0051] One way of evaluating the information obtained from the measuring devices is shown below and was carried out using MATLAB (MathWorks). Image acquisition with the first measuring device is implemented as a sensor array and provides the shadow cast by the fragments as a vector x with grayscale values ​​[0..255] per unit of time. Simultaneously, a value for height information is acquired with the second measuring device: scalar z in [mm]. The individual measurements are combined over time t to form an image: image [x,y] with grayscale values ​​[0..255, 0..255]. The height information is combined over time t to form a vector and assigned to the ordinate: vector [z] with values ​​in [mm]. The resulting images are then repeatedly evaluated. Using an adjustable threshold (e.g., grayscale value 128), the grayscale values ​​[0..255] are converted into binary values ​​[0 = dark or 1 = light]. By finding contiguous areas with [0 = dark] in the image, a list of fragments is obtained. A repeated calculation is then performed for each fragment.Calculation of the points at the edge of the area as a list and conversion to [mm] : List [. xi, yi ] in [mm]. Assignment of height values ​​to the list: [ xi,yi,zi ] in [mm]. Calculation of the maximum distance in 3D for all points in the list [ xi,yi,zi ] relative to each other: Scalar for the maximum extent in [mm] for the fragment. If an adjustable limit value in [mm] for the maximum extent of the fragment is exceeded, a command is issued to blow out the fragment in the area of ​​the nozzle bar where the fragment is located [min(xi), max( xi )].

[0052] The slippage after this sorting was determined manually as described and was 0.0%. Comparative example 2

[0053] Classification of crushed polysilicon of BG3.

[0054] Initially, polysilicon sample material consisting of 9000 fragments in a length range of 20 to 60 mm (BG3) and 1000 fragments in a length range > 60 to 85 mm (fraction to be separated) was produced.

[0055] The fragment lengths were checked manually. The polysilicon sample material was then finalized by mixing.

[0056] The polysilicon sample material was sorted using the optopneumatic sorting device described in comparative example 1, but with a cut-off width of 60 mm.

[0057] The slippage after this sorting was determined manually as described and amounted to 1.0%. Example 2

[0058] The polysilicon sample material described in comparative example 2 was sorted here using an optoneumatic sorting device according to the Fig. 2sorted. This therefore had a second measuring device in the form of a light section sensor, consisting of a laser projector and a receiving optic.

[0059] The slippage after this sorting was determined manually as described and was 0.0%.

[0060] It was demonstrated that the sorting result, particularly for pyramidal objects, can be significantly improved by providing additional height information determined using a second measuring device. The fundamental goal of any sorting process is a clean, uninterrupted cut resulting in a single fragment size.

Claims

1. Method for sorting chunks, comprising the steps - singulating the chunks in a singulating region, - recording the projected area of a chunk in a 2D profile plane with at least one first measuring device, - recording at least one height information item above and / or below the 2D profile plane with at least one further measuring device, - calculating the size of the chunk from the projected area and the height information item, - controlling at least one deflecting device as a function of the calculated size.

2. Method according to Claim 1, characterized in that the first measuring device and at least one of the further measuring devices are photoelectric transmitted-light or reflected-light measuring systems having a detection region through which the chunk passes.

3. Method according to Claim 2, characterized in that the silicon chunk passes through the detection region in freefall.

4. Method according to Claim 2 or 3, characterized in that the reflected-light measuring system comprises a light section sensor and / or at least one camera system.

5. Method according to Claim 4, characterized in that the camera system is a camera system for photometric stereo analysis.

6. Method according to Claim 2 or 3, characterized in that the transmitted-light measuring system comprises a photoelectric barrier, a light curtain or a light grid.

7. Method according to at least one of the preceding claims, characterized in that the first measuring device is a camera system.

8. Method according to at least one of the preceding claims, characterized in that the further measuring device is a light curtain or a light grid.

9. Method according to one of the preceding claims, characterized in that the recording of the projected area and the recording of the height information item take place with a time spacing of from 0 to 100 ms, preferentially from 0 to 50 ms, particularly preferentially simultaneously.

10. Method according to one of the preceding claims, characterized in that the deflecting device is a pneumatic or mechanical deflecting device.

11. Device for sorting chunks, comprising - a singulating region for singulating the chunks, - at least one first measuring device for recording the projected area of a chunk in a 2D profile plane, - at least one further measuring device for recording a height information item above and / or below the 2D profile plane, - at least one deflecting device, - a software-aided controller which calculates a size of the chunk from the projected area and the height information item and controls the deflecting device as a function of this size.

12. Device according to Claim 11, characterized in that the singulating region comprises at least one vibrating conveyor trough and / or conveyor belt.

13. Device according to Claim 11 or 12, characterized in that the deflecting device is a pneumatic or mechanical deflecting device.

14. Device according to Claim 13, characterized in that the pneumatic deflecting device comprises a row or a matrix of individual nozzles.