METHOD FOR PRODUCING SEMICONDUCTOR WAFERS USING A WIRE SAW, WIRE SAW AND SEMICONDUCTOR WAFER FROM SINGLE CRYSTALLINE SILICON
Patent Information
- Application Number
- DE502019013339
- Authority / Receiving Office
- DE · DE
- Patent Type
- Patents
- Current Assignee / Owner
- Priority Date
- 2018-12-17
- Filing Date
- 2019-12-12
- Publication Date
- 2025-05-28
- Estimated Expiration
- 2039-12-12
AI Technical Summary
Existing procedures for producing semiconductor slices using a wire saw face challenges in accurately compensating for axial movement due to heat expansion, leading to suboptimal warp and nanotopography in the produced slices.
The procedure involves dividing wires into multiple wire groups, each with its own compensation movement, activated by drive elements, to correct for misalignments in the cutting gap, thereby improving the precision of semiconductor slice production.
This approach results in semiconductor slices with reduced warp and improved nanotopography, enhancing the quality of the slices and reducing the need for subsequent refining processes.
Description
[0001] The invention relates to a method for producing semiconductor wafers from a workpiece by machining the workpiece using a wire saw, a wire saw for carrying out the method, and a semiconductor wafer made of monocrystalline silicon that is obtained by the method. In particular, the invention relates to a semiconductor wafer according to the preamble of claim 1. Such a semiconductor wafer is disclosed in DE 11 2016 005417 T5. State of the art / problems
[0002] WO 2015 / 188 859 A1 describes in detail a wire saw and its operating principle.
[0003] From JP 09 109 143 A a method for producing wafers from a workpiece (ingot) by means of a wire saw is known, which comprises the following steps: detecting the position of wires of a wire field of the wire saw and, in the presence of a deflection of the wires, bringing about a compensating movement of a wire guide roller to correct the deflection.
[0004] A similar method is known from JP 11 165 251, which comprises the following steps: detecting the position of wires of a wire field of the wire saw and, in the presence of a deflection of the wires, causing a compensating movement of the workpiece.
[0005] A similar method is known from US 5 875 770, which comprises the following steps: detecting the warp of slices before machining the workpiece and causing a compensating movement of the workpiece along an axial direction of the workpiece to such an extent that slices with reduced warp are produced.
[0006] A similar method is known from JP 2009 61 527 A, which takes into account the fact that the workpiece can move axially due to thermal expansion by providing means to enable the wire guide rollers to move compensatingly.
[0007] Despite these available solutions, there is still a need to improve the process for producing semiconductor wafers from a workpiece using a wire saw. In particular, it must be considered that the axial movement of the workpiece due to thermal expansion can only be approximately compensated by a corresponding axial movement of the wire guide rollers. Furthermore, controlling the axial positions of the wires via the temperature of the wire guide rollers is comparatively slow. Furthermore, the measured warp of wafers from a previously machined workpiece can only approximately describe deviations for a workpiece yet to be machined, and the axial movement of an entire workpiece can only partially compensate for such deviations.
[0008] In particular, an improvement of the process is necessary to make semiconductor wafers accessible whose flatness, particularly with regard to warp and nanotopography, is better than that of wafers produced in a known manner.
[0009] The problem described above led to the object of the invention. The patent specification discloses a method for producing semiconductor wafers from a workpiece by machining the workpiece using a wire saw, comprising the feeding of the workpiece by an arrangement of wires which are arranged in wire groups and are stretched between wire guide rollers and move in a running direction; the generation of cutting gaps when the wires engage in the workpiece; for each of the wire groups, the determination of an incorrect position of the cutting gap of the wire group; and for each of the wire groups, the bringing about of compensating movements of the wires of the wire group depending on the determined incorrect position of the cutting gap of the wire group in a direction perpendicular to the running direction of the wires of the wire group during the feeding of the workpiece by the arrangement of wires by activating at least one drive element.
[0010] At least two wire groups are provided, preferably at least three wire groups, and particularly preferably four wire groups. However, it is also possible for each wire to be considered a separate wire group, i.e., the number of wire groups corresponds to the number of wires in the wire arrangement (wire grid). Wire groups preferably have the same axial width, and the spacing between adjacent wires in a wire group is the same.
[0011] For each wire group separately, compensating movements of the wires in the wire group are induced in a direction perpendicular to the running direction of the wires in the wire group, i.e., in the direction of the rotation axes of the wire guide rollers. For this purpose, at least one drive element is activated, which moves the wires in the specified direction and over the specified distance. The compensating movements of the wires in the wire group are induced independently of any simultaneous compensating movements of the wires in another wire group.
[0012] Electromagnetic, mechanical, hydraulic, pneumatic, magnetostrictive, and preferably piezoelectric actuators can be considered as drive elements. For example, if the wires are divided into four wire groups, the wire groups form two inner and two outer wire groups. If the workpiece expands due to heat, the misalignment of a cutting gap attributable to a wire in one of the inner wire groups is smaller than that attributable to a wire in one of the outer wire groups. Accordingly, the amounts of the compensating movements of the wires in the outer wire groups must be selected to be larger than the amounts of the compensating movements of the wires in the inner wire groups. If each wire is considered a separate wire group, each wire is subjected to a compensating movement with its own amount and direction.
[0013] A semiconductor wafer separated from the workpiece has an upper and a lower side surface and an edge running between them. Typically, the aim is for the upper and lower side surfaces to be as flat as possible after separation from the workpiece and to have as even a distance from each other as possible. The better the initial flatness of the side surfaces and the uniformity of the thickness of the semiconductor wafer, the easier it is and the less costly it is to refine the semiconductor wafer through subsequent steps such as lapping and / or grinding, etching, polishing, and, if necessary, coating into a target product that meets the strict requirements of the industry that further processes the semiconductor wafer into electronic components.The upper side surface is also referred to as the front side of the semiconductor wafer and is usually the surface on or in which the structures of electronic components are intended to be accommodated during further processing of the semiconductor wafer.
[0014] The aim of the process is to ensure that, when machining the workpiece using a wire saw, kerfs are created in the workpiece whose position deviates as little as possible from a position considered ideal. If semiconductor wafers with a uniform thickness and side surfaces that are as flat as possible are desired, an ideal kerf runs straight and at right angles to the longitudinal axis of the workpiece. In other words, the trajectory through the center of such a kerf runs along a straight line oriented perpendicular to the longitudinal axis of the workpiece. Such a trajectory is referred to below as the desired trajectory. Accordingly, a kerf is incorrectly positioned if the actual trajectory deviates from the desired trajectory. This is the case when a position vector pointing to the center of the kerf no longer ends on the desired trajectory.
[0015] A misalignment of a cutting gap occurs, for example, when a wire moves perpendicular to its running direction during engagement with the workpiece, i.e., in the direction of the rotational axes of the wire guide rollers between which it is tensioned, or when the workpiece expands axially during feeding due to the heat generated by the arrangement of wires. In the latter case, the misalignment of a cutting gap is greater the further the cutting gap is from the center of the workpiece. The center of the workpiece is the location between the two ends of the workpiece.
[0016] One aspect of the method is to determine the kerf misalignment for each wire group separately, regardless of the cause that led to relative movement between the workpiece and the wire group. Examples of such causes include movement of the wire group, movement of the workpiece, or thermal expansion of the workpiece. Another aspect of the method is to distinguish between kerf misalignment that occurs systematically when using a specific wire saw and kerf misalignment that occurs randomly and independently of the use of a specific wire saw.
[0017] It is expedient to set up at least one closed control loop separately for each of the wire groups, in which a control deviation, i.e. a determined incorrect position of the cutting gap of the wire group, is reacted to by changing the manipulated variable, i.e. by bringing about a compensating movement of the wires of the wire group.
[0018] According to a first embodiment of the method, the incorrect position of the cutting gap of the wire group is determined separately for each wire group during the feed of the workpiece by arranging the wires. According to a first alternative of the first embodiment, the position of each cutting gap is preferably measured relative to a fixed reference point and compared with a target position. The target position of a cutting gap is the position relative to the fixed reference point, which is a prerequisite for an ideal cutting gap to be created. The deviation of the measured position of the cutting gap from its target position corresponds to the incorrect position of the cutting gap. Since the deviation is fundamentally different for each cutting gap in the wire group, the deviations are averaged to form an incorrect position that represents the incorrect position of the cutting gap in the wire group.In other words, each cutting gap in the wire group is assigned the same averaged misalignment. Averaging can be performed without weighting, or misalignments of specific cutting gaps can be given special weighting. From the misalignment of the wire group's cutting gap, a correction profile can be derived that specifies the amount and direction by which the wires in the wire group must be moved during workpiece advancement in order to correct the misalignment. The correction profile, viewed across the penetration depth of the wires into the workpiece, has a profile that is complementary to the profile of the determined misalignment of the wire group's cutting gap.
[0019] The position of the wire group's cutting gap is preferably measured by irradiating the wire group's cutting gap with optical radiation, IR radiation, X-ray radiation, or γ-radiation. Mechanical scanning of the wire group's cutting gap or inductive or capacitive measurement of the wire group's cutting gap can also be considered. Such direct observation of the wire group's cutting gap reveals any relative movement between the workpiece and the wires of a wire group.
[0020] According to a second alternative of the first embodiment, the position of each wire in a wire group and of the workpiece relative to a fixed reference point is preferably measured simultaneously and compared with a target position in order to detect such a relative movement. Since the deviation is fundamentally different for each wire in the wire group, the deviations are averaged to form an incorrect position that represents the incorrect positions of the wires in the wire group. In other words, each wire in the wire group is assigned the same, averaged incorrect position. The averaging can be carried out without weighting, or incorrect positions of certain cutting gaps can be given special weighting. The target position of the wire is the position of the wire relative to the fixed reference point, which is a prerequisite for creating an ideal cutting gap. The same applies to the target position of the workpiece.The deviation of the actual trajectory from the desired trajectory is approximately determined from the sum of the measured deviations from the target positions of the wire and workpiece.
[0021] The position of the wires in the wire group or the position of the workpiece is measured by irradiating the wires in the wire group or the workpiece with optical radiation, IR radiation, X-ray radiation, or γ-radiation, or by capacitive or inductive measurement. Mechanical scanning of the wires in the wire group or the workpiece, or inductive or capacitive measurement of the wires in the wire group or the workpiece, is also possible. The position of the workpiece can be determined relative to the end faces of the workpiece and, preferably, relative to reference points marked on the workpiece.
[0022] According to a second embodiment of the method, the misalignment of the cutting gap of the wire group is determined separately for each wire group before the workpiece is fed by arranging the wires. This procedure determines a misalignment of the cutting gap that systematically occurs when using a specific wire saw. To determine the misalignment of the cutting gap, the local geometry of semiconductor wafers that were previously produced using a specific wire saw is measured. These semiconductor wafers originate from one or more workpieces that were produced using this wire saw, specifically using wires from the wire group for which the misalignment of the cutting gap is determined. The local geometry of a semiconductor wafer approximately maps the trajectory of the cutting gap adjacent to the semiconductor wafer.Preferably, the local geometry is derived from the median surface of a warp measurement according to SEMI MF 1390-0218 as follows: a contour line (line scan, LS) is generated by selecting those measured values of the median surface that are located on a line passing through the center of the semiconductor wafer. The measured values lie on a line that follows a diameter of the semiconductor wafer, preferably in the direction of the workpiece feed during the separation of the semiconductor wafer, or at least deviates from such a direction by no more than ± 20°.
[0023] To detect misalignment of the cutting gap of the wire group, which systematically occurs when using a specific wire saw, the local geometry of the semiconductor wafers originating from one or more workpieces produced using this wire saw and produced using wires from the wire group is averaged to form a single local geometry. The averaging can be performed without weighting, or the local geometry of certain semiconductor wafers can be given special weighting based on their relative position in the workpiece. Based on the averaged local geometry, it is then concluded what the trajectory of the cutting gap will be if the specific wire saw is used and other influences that affect the trajectory are disregarded. Such a trajectory is referred to below as the expected trajectory.The misalignment of the wire group's cutting gap, which can be expected during workpiece advancement, is determined by comparing the expected trajectory with the desired trajectory. This comparison yields a wire saw-specific correction profile that specifies the direction and magnitude of the compensatory movements of the wires in the wire group as a function of the wires' penetration depth into the workpiece during workpiece advancement through the arrangement of wires. The course of the wire saw-specific correction profile is essentially complementary to the course of the averaged local geometry.
[0024] The wire saw-specific correction profile is preferably used additionally to detect and respond to changes in the wire saw's performance at an early stage. Changes in the wire saw-specific correction profile that occur during workpiece processing indicate wear of the wire and / or the lining of the wire guide rollers or another component of the wire saw subject to wear. A threshold for changes in the wire saw-specific correction profile can therefore be defined; upon reaching this threshold, preventive maintenance measures (predictive maintenance) are initiated. Even before such a threshold is reached, changes in the wire saw-specific correction profile can be used as an opportunity to implement adjustments that counteract wear-related deterioration of the work result.Such adjustment measures may include, for example, changing the composition and / or temperature of a cutting fluid suspension or changing the temperature of a coolant, as well as changing the wire speed or other process-specific parameters.
[0025] A third embodiment of the method provides for the combination of the first and second embodiments. A first part of the compensating movements of the wires in the wire group is brought about on the basis of a correction profile which, in accordance with the first embodiment of the method, is determined in real time during the advancement of the workpiece as a function of the penetration depth of the wires in the wire group. A further part of the compensating movements of the wires in the wire group is brought about on the basis of a wire saw-specific correction profile which, in accordance with the second embodiment of the method, was determined before the advancement of the workpiece by arranging the wires for the respective wire group. In this way, influences on the incorrect position of the cutting gap which occur randomly and are therefore unpredictable and those which occur systematically due to the use of a specific wire saw are taken into account in a decoupled manner.
[0026] A wire saw-specific correction profile can of course also be obtained by recording the correction profile derived according to the first embodiment of the method.
[0027] The disclosed methods can be used in conjunction with wires that have abrasive grit bonded to the wire, or in conjunction with wires that are free of such grit and exert their effect in combination with an abrasive suspension. Diamond, in particular, is considered as bonded abrasive grit. The wires referred to here are sections of a wire wound around the wire guide rollers of the wire saw. The number of wire guide rollers of the wire saw is not essential for the use of the method. For example, the wire saw can comprise two, three, four, or an even higher number of wire guide rollers.
[0028] The workpiece preferably consists of a semiconductor material such as silicon, which can be in a multicrystalline or monocrystalline state. The circumference of the workpiece is square, rectangular, or circular. The process is particularly suitable for producing round semiconductor wafers made of monocrystalline silicon with a diameter of at least 200 mm, in particular at least 300 mm.
[0029] Furthermore, a wire saw for producing semiconductor wafers by machining a workpiece is disclosed, comprising wire guide rollers between which wires are stretched to form an arrangement of wires which are divided into wire groups and move in a running direction; a unit for advancing the workpiece by arranging wires, creating cutting gaps when the wires engage the workpiece; drive elements, at least one of which is assigned to each of the wire groups, for moving the wires of the assigned wire group; and a control unit for activating the drive elements, wherein, in the event of an incorrect position of the cutting gaps of the wire groups, the control unit activates the drive element assigned to the wire group, whereby the wires of this wire group perform compensating movements perpendicular to the running direction.
[0030] The wire saw comprises a unit for feeding a workpiece through an arrangement of wires stretched between wire guide rollers.
[0031] The wire saw can comprise two or more wire guide rollers around which saw wire is wound. In the wire saw, the wires are divided into wire groups, and at least one drive element is assigned to each of the wire groups. When a drive element is activated, the wires of the wire group assigned to it are moved perpendicular to the direction of travel of the wires, i.e. in the direction of the rotation axes of the wire guide rollers. The activated drive element moves the wires of the wire group assigned to it simultaneously and by the same amount in the same direction. Because of the assignment of the drive elements, the amount and direction of movement of the wires in one wire group are independent of the amount and direction of movement of the wires in another wire group. The wire guide rollers, between which the wires arranged in wire groups are stretched to form the arrangement of wires, are provided with the drive elements.The drive elements are arranged between the wire groups of at least one of the wire guide rollers, between which the workpiece is fed by the arrangement of wires. It is not necessary, but also not impossible, that the other wire guide roller spanning the wire frame is also equipped with such drive elements.
[0032] The wire saw also includes a control unit for activating the drive elements. If the cutting gap of a wire group is misaligned, the control unit activates the drive element assigned to the wire group, whereupon the wires of this wire group perform a compensating movement that reduces or eliminates the detected cutting gap misalignment.
[0033] The control unit accesses either data on the misalignment of the cutting gap, which are provided during the feeding of the workpiece by means of a measuring device according to the first embodiment of the method, or wire saw-specific data, which are stored in a data memory and are made available before the feeding of the workpiece according to the second embodiment of the method, or both the one and the other data.
[0034] The measuring device is used, separately for each wire group, to determine the misalignment of the cutting gap of the respective wire group during the feed of the workpiece through the arrangement of wires. The information on the misalignment of the cutting gap of the wire group is transmitted to the control unit, where it is further processed into a signal (manipulated variable) for activating the drive element assigned to the corresponding wire group. The measuring device and the control unit are components of a first closed control loop for minimizing the misalignment of the cutting gap. By means of this control loop, the drive element assigned to the wire group is activated separately for each wire group, regardless of the possible simultaneous activation of a drive element assigned to another of the wire groups.
[0035] The data memory is used to store data for activating the drive elements. The data represents a wire saw-specific correction profile, one for each wire group. Such a wire saw-specific correction profile specifies the direction and magnitude of the compensating movements of the wires in the wire group depending on the penetration depth of the wires into the workpiece. The control unit accesses this data during the workpiece feed by arranging the wires and activates the drive element assigned to the respective wire group according to the specifications of the wire saw-specific correction profile applicable to the wires in this wire group. Preferably, the data memory and the control unit are components of another closed control loop for minimizing misalignments in the cutting gap.
[0036] If the wire saw is equipped with a data storage device, a processing unit is preferably also provided to track changes in a wire saw-specific correction profile during the processing of multiple workpieces. When a specified threshold of changes is reached, the processing unit issues a signal to initiate preventive maintenance.
[0037] The object of the invention is achieved by a semiconductor wafer, according to claim 1, made of single-crystal silicon with an upper side surface and a lower side surface, comprising a warp of less than 1.2 µm; a nanotopography of the upper side surface, expressed as THA25 10%, of less than 5 nm; and a sub-area-related nanotopography of the upper side surface of less than 6 nm, expressed as the maximum peak-to-valley distance on a sub-area and related to sub-areas with an area of 25 mm x 25 mm each.
[0038] The semiconductor wafer has a diameter of at least 200 mm, particularly preferably 300 mm. The above-mentioned properties of the semiconductor wafer regarding warp, nanotopography, and area-related nanotopography refer to a semiconductor wafer with a diameter of 300 mm.
[0039] The warp of the semiconductor wafer is determined according to the standard SEMI MF 1390-0218.
[0040] An interferometer, such as a WaferSight™ from KLA-Tencor Corp., is used to examine nanotopography. This type of interferometer is suitable for measuring the topography on the upper side surface of a semiconductor wafer. The device creates a height map of the upper side surface of the semiconductor wafer, which is filtered and over which an analysis window with a defined analysis area is moved. The height differences in the analysis window are evaluated using THA (threshold height analysis) according to the procedures specified in SEMI M43-0418 and SEMI M78-0618. THAXX 10% < 5 nm means that no more than 10% of the analyzed area of the upper side surface of the semiconductor wafer in the analysis window with the analysis area specified by XX may have a maximum PV distance (peak-to-valley metric) of 5 nm or greater.In a semiconductor wafer according to the invention, the nanotopography of the upper side surface, expressed as THA25 10%, is less than 5 nm. A circular analysis window with a diameter of 25 mm is used, and the unfiltered topography signal is filtered with a single Gaussian high-pass filter with a 20 mm cutoff wavelength. The cutoff wavelength shrinks to 1 mm at the edge of the semiconductor wafer. An edge exclusion of 5 mm is observed before filtering and an edge exclusion of 15 mm after filtering.
[0041] When measuring the nanotopography of the upper side surface, the semiconductor wafer can be in the state after separation from the workpiece or in a state following separation from subsequent processing steps, such as etching and polishing. Preferably, the upper side surface of the semiconductor wafer is in a polished state.
[0042] The nanotopography is evaluated on a site-by-site basis, i.e., based on user-specific sites on the upper side surface of the semiconductor wafer. The PV spacing is determined as described above (analysis window with a circular circumference and a diameter of 25 mm, simple Gaussian high-pass filter, cutoff wavelength of 20 mm, shrinking to 1 mm towards the edge of the semiconductor wafer). However, the edge exclusion is 2 mm before filtering and 3 mm after filtering. The upper side surface of the semiconductor wafer is divided into sites grouped around a site whose lower left corner is located in the center of the upper side surface of the semiconductor wafer. In a semiconductor wafer according to the invention, the site-by-site nanotopography of the upper side surface is less than 6 nm, expressed as the maximum PV spacing on a site and based on sites with an area of 25 mm x 25 mm.
[0043] The invention is further explained below with reference to drawings. Short description of the characters
[0044] Fig. 1 shows main features of a state-of-the-art wire saw. Fig. 2 shows features of a wire guide roller of the wire saw. Fig.3 shows schematically the process flow. Fig.4 shows how an actual trajectory through the center of a cutting gap can differ from a desired trajectory. Fig.5 and Fig.6 show the direction and amounts of possible compensating movements of wire groups using two examples. Fig.7 shows a typical correction profile on the basis of which compensatory movements of the wires of a wire group are brought about. Fig.8 to Fig.10 and Fig.11 to Fig.13 show contour lines of three semiconductor wafers each derived from median areas of a warp measurement. Fig.14 to Fig.16 are equivalent to Fig.8 to Fig.10with the difference of a higher resolution scaling of the ordinate. Fig.17, 18 and 19 show how a wire saw-specific correction profile can change during the machining of several workpieces. List of reference symbols used
[0045] 1 saw wire 2 grooves 3 left wire guide roller 4 right wire guide roller 5 axis 6 axis 7 rotation 8 Direction of rotation 9 Wire longitudinal movement 10 Wire longitudinal movement 11 Wire gate 12 Delivery device 13 Cutting gap 14 axis 15 workpiece 16 saw bar 17 Glue 18 Arrow direction 19 nozzle comb 20 nozzle comb 21 nozzles 22 beam 23 beam 24a-e drive element 25a-d Wire group 26 Fixed bearing 27 warehouse 28 Control unit29 Data specification 30 measuring device 31 Data storage 32 desired trajectory
[0046] Fig.1 shows main features of a state-of-the-art wire saw and serves to explain the basics of a method for producing semiconductor wafers from a workpiece by machining the workpiece using a wire saw.
[0047] A suitable wire saw comprises saw wire 1, which is wrapped several times in a spiral shape around a left wire guide roller 3 and a right wire guide roller 4 and is guided by grooves 2 such that the wire sections running on the upper side of the wire guide rollers, which are referred to as wires for the description of the exemplary embodiment, run parallel and form a wire frame 11. A workpiece 15 is fastened to a saw bar 16, for example, glued by means of an adhesive 17. The saw bar 16, with the workpiece 15, is advanced perpendicularly against the wire frame 11 in the direction of arrow 18 by a feed device 12 (illustrated only briefly) and brought into engagement with the wires of the wire frame 11.Optionally, the wire saw comprises left nozzle combs 19 and right nozzle combs 20 with nozzles 21 for supplying a cutting agent suspension in the form of a left elongated jet 22 and a right elongated jet 23 onto the left wire guide roller 3 and the right wire guide roller 4.
[0048] The wire guide rollers are mounted for rotation about axes 5 and 6. Their axes and the axis 14 of the workpiece 15 – in the example shown, a cylindrical rod – are aligned parallel to each other. To initiate the cutting process, one wire guide roller, for example, the left wire guide roller 3, is driven to rotate 7 ("master"). The other wire guide roller ("slave"), in the example, the right wire guide roller 4, rotates in the same direction, pulled by wire 1, in the direction of rotation 8. When the wires engage the workpiece 15, cutting gaps 13 are formed.
[0049] Typically, the direction of the wire longitudinal movement 9, 10 is reversed several times during a complete cut through the workpiece 15, wherein in each of these pairs of direction reversals, called "pilgrimage steps," the wire is moved by a greater length in one direction and a smaller length in the opposite direction.
[0050] A wire saw according to the embodiment has at least one wire guide roller comprising drive elements. An example of such a wire guide roller is shown in Fig.2shown. When a drive element is activated, an activated wire group is moved in a direction perpendicular to the running direction of the wires of the wire group. In the example shown, five drive elements 24a-e and four wire groups 25a-d are provided between the fixed bearing 26 and the bearing 27 of the wire guide roller. The bearing 27 can be designed as a floating bearing or as a fixed bearing. The drive element 24e, which is adjacent to the bearing 27, can be omitted if the bearing 27 is a floating bearing. Regardless of this, the bearing 27 is drawn as a floating bearing in the example shown. The double arrows indicate possible directions of movement in which a wire group moves when its associated drive element is activated. The wire guide roller can be divided into segments provided with grooves for wires, which are arranged on a shaft, for example into segments as described in JP 11 123 649 A2.A drive element is then positioned between each end of the segments and adjacent to fixed bearings. With such a design, the drive element shown in . Fig.2 The wire guide roller shown has a number of segments corresponding to the number of wire groups.
[0051] Fig.3 schematically shows the process by which a control unit 28, based on data relating to an incorrect position of the cutting gap of a wire group, acts on drive elements 24 with a correction signal 32 based on a correction profile and / or a wire saw-specific correction profile, causing the wires of the wire groups 25 to perform compensating movements. The control unit 28 receives the necessary input from a measuring device 30 and / or a data memory 31 in which the wire saw-specific correction profile is stored.
[0052] Fig.4shows a cross-sectional image that can be obtained either by observing the cutting gap or by observing the wires and the workpiece while the wires engage with the workpiece. It shows a part of the workpiece 15 and a cutting gap 13 that extends through the workpiece. The actual trajectory, which runs through the center of the cutting gap 13, deviates more or less significantly from a desired trajectory 32 while the cutting gap is being created. The difference represents the determined misalignment of the cutting gap 13. As mentioned, it is intended to bring about compensating movements of the wires in the wire group for each wire group, depending on the determined misalignment of the cutting gap of the wire group, in a direction perpendicular to the running direction of the wires in the wire group, by activating a drive element assigned to the respective wire group.
[0053] Fig.5 and Fig.6show two examples of the direction and magnitude of possible compensating movements of wire groups. The drive elements 24a-e are, for example, piezoelectric actuators. In the case of Fig.5 causes the same activation of the Fig.2 shown drive elements 24a-d, that the piezoelectric actuators each expand by the same amount, so that the wire groups 25a-d and the bearing 27 of the wire guide roller are accordingly moved away from the fixed bearing 26. The amount of the compensating movement increases, as indicated by the arrows, evenly from the wire group 25a to the wire group 25d. However, it may also be necessary, for example, that, as shown in Fig.6 shown, compensating movements are required, during which wire groups move against each other. Fig.6 The result can be seen when the Fig.2The drive elements shown are activated in such a way that the drive element 24a expands by an amount a and the drive element 24d by an amount d (amount d > amount a) towards the bearing 27, and the drive element 24b shrinks by an amount b and the drive element 24c by an amount c (amount c > amount d) towards the fixed bearing 26. The total then results in the compensating movements of the wires of the wire groups and the bearing 27 indicated by arrows: The wires of the wire group 25a are moved by the greatest amount towards the bearing 27, the wires of the wire group 25b are not moved, the wires of the wire group 25c are moved by an amount towards the fixed bearing 26, and the wires of the wire group 25d are moved by the smallest amount towards the bearing 27.
[0054] The comparison of the actual trajectory and the desired trajectory according to the first embodiment or the comparison of the expected trajectory and the desired trajectory according to the second embodiment leads to a description of the course of the incorrect position of the cutting gap of a wire group as a function of the penetration depth of the wires of the wire group in the workpiece and to a correction profile (first embodiment) or to a wire saw-specific correction profile (second embodiment) which is complementary to the course of the incorrect position of the cutting gap.
[0055] Fig.7shows a correction profile in which the deviation Δ of the actual trajectory from the desired trajectory is plotted as a function of the penetration depth P of the wires in the wire group. Compensatory movements of the wires in the wire group with a direction and a magnitude corresponding to the deviation Δ are induced by activating the drive element assigned to the wire group. Only if there is no misalignment of the cutting gap (Δ = 0), which in the example shown is initially the case at a penetration depth of approximately -100 mm, are such compensatory movements omitted.
[0056] Fig.8 to Fig.13 show contour lines LS of three semiconductor wafers each, which were separated from a workpiece by wires of a wire group, whereby during the separation of the semiconductor wafers, the wires of the wire group were subjected to compensating movements specified by the correction profile ( Fig.8 to Fig.10) or the initiation of such compensatory movements was waived ( Fig.11 to Fig.13 ). The contour lines are each derived from the median area of a warp measurement, whereby measured values of the median area were selected that lie on a line that follows the diameter of the respective semiconductor wafer in the direction of the workpiece during the separation of the semiconductor wafer. The position of the semiconductor wafers in the workpiece was such that, during the separation of the semiconductor wafers, 50 additional semiconductor wafers were created between each of the three semiconductor wafers. As the comparison of the contour lines reveals, semiconductor wafers become significantly flatter through the application of the disclosed method, without any particular influence of their position in the workpiece. This is also confirmed by Fig.14 to Fig.16 , which differ from Fig.8 to Fig.10 only differ in that the ordinate scaling has a higher resolution.
[0057] The Fig.17, 18 and 19show how a wire saw-specific correction profile can constantly change during the processing of several workpieces. It is therefore advantageous to define a threshold for the deviation Δ, which, if exceeded, triggers preventive maintenance measures. The threshold can, for example, be defined in such a way that only a wire saw-specific correction profile with a maximum deviation Δ max , as defined in Fig.19 shown, results in preventive maintenance measures being initiated.
[0058] The features specified with regard to the above-mentioned embodiments of the method can be transferred accordingly to the device. Conversely, the features specified with regard to the above-mentioned embodiments of the device can be transferred accordingly to the method. These and other features of the embodiments are explained in the description of the figures. The individual features can be implemented as embodiments either separately or in combination. The above description of exemplary embodiments is to be understood as exemplary. The disclosure thus made enables the person skilled in the art, on the one hand, to understand the exemplary embodiments and the associated advantages, and on the other hand, also includes obvious variations and modifications of the described structures and methods within the understanding of the person skilled in the art.
[0059] Therefore, all such alterations and modifications as well as equivalents are intended to be covered by the scope of the claim.
Claims
1. Semiconductor wafer composed of monocrystalline silicon having an upper side surface and a lower side surface and a diameter of at least 200 mm, characterized by a warp of less than 1.2 µm, the warp being determined in accordance with the SEMI MF 1390-0218 standard; a nanotopography of the upper side surface, expressed as THA25 10%, of less than 5 nm; and a site-referenced nanotopography of the upper side surface of less than 6 nm, expressed as a maximum peak-to-valley distance on a site and referenced to sites with a surface area of in each case 25 mm x 25 mm; wherein in order to examine the nanotopography by means of an interferometer, a height map of the upper side surface of the semiconductor wafer is formed, which is filtered and over which an analysis window with a defined analysis area is moved; the analysis window has a circular circumference and a diameter of 25 mm and the unfiltered topography signal is filtered with a single Gaussian high-pass filter with a cut-off wavelength of 20 mm; the evaluation of the height differences in the analysis window is carried out in accordance with the method protocols defined in the SEMI M43-0418 and SEMI M78-0618 standards; a THA25 10% of less than 5 nm means that at most 10% of the analysed area of the upper side surface of the semiconductor wafer in the analysis window with a circular circumference and a diameter of 25 mm has a maximum height difference of 5 nm or greater; and in order to determine the site-referenced nanotopography, the upper side surface of the semiconductor wafer is subdivided into sites with a surface area of 25 mm x 25 mm, which are grouped around a site whose lower left corner is arranged in the centre of the upper side surface of the semiconductor wafer.