JP5109095B2 - Manufacturing method of glass substrate for magnetic disk - Google Patents

Description

  The present invention relates to a method for manufacturing a glass substrate for a magnetic disk.

  A glass substrate is used as a substrate of a magnetic disk used in a magnetic disk storage device or the like (for example, Patent Document 1). This glass substrate has a disk shape with a circular hole in the center, and a Φ65 mm magnetic disk is required to have a high level difference (Peak to Valley) of 1 μm between 65 mm lengths. A general method for producing a glass substrate will be briefly described by taking the production of a glass substrate having a diameter of 65 mm as an example. After preparing and melting glass raw materials, the molten glass is formed into a plate shape and cooled. A glass plate having a relatively large size, for example, about 1000 mm × 1000 mm (hereinafter, referred to as a 1000 mm square) is obtained by cutting the cooled plate glass into a size of about 6000 mm × 1000 mm (FIG. 1A). Further cut into a plate (FIG. 1B). Subsequently, this glass base plate is cut into a glass plate that is slightly larger than the magnetic disk substrate, for example, about 75 mm × 75 mm (hereinafter referred to as 75 mm square) for a Φ65 mm magnetic disk (FIG. 1C). Then, a desired glass substrate is manufactured through processes such as shape processing, lapping, and polishing.

  If the difference in height of a glass plate of about 75 mm square manufactured in such a manufacturing process is bad, the quality of the final product may be deteriorated. Therefore, in the conventional glass substrate manufacturing process, after cutting into a 75 mm square glass plate that is slightly larger than the magnetic disk substrate, the difference in height is measured and the measurement result is fed back to the operator who manages the temperature during glass molding. By doing so, the height difference of the glass base plate is adjusted.

  The height difference of the glass plate is represented by a parameter called Mesa flatness, for example. The measurement is performed using a Mesa interferometer manufactured by ZYGO. As shown in FIG. 2, in the Mesa interferometer 10, a laser diode light source 11 irradiates a sample with two beams having different incident angles through two diffraction gratings 12, and the same grating reflects two beams reflected from the sample. By recombining to generate an interference pattern and directing it toward the camera array 13, the diffraction grating 12 efficiently generates a wavelength about 20 times the wavelength of the light source 11, and thereby a surface having a large roughness and deformation. It is possible to perform the measurement.

  In the Mesa interferometer 10, one of the diffraction gratings is accurately moved by the piezoelectric transducer (PZT) 14 during measurement of the sample, and as a result, a constant phase shift occurs between the reference wavefront and the measurement wavefront, While the diffraction grating 12 is moving, data is acquired when the phase shift between the mutual interference wavefronts becomes 90 degrees. Then, by analyzing the phase data with predetermined software, the surface shape of the sample becomes clear.

JP 2008-103061 A

  However, the measurement using this Mesa interferometer is performed after two cutting steps, that is, a step of cutting into a 1000 mm square glass base plate and a step of cutting into a 75 mm square glass plate, There is a problem that it takes time to feed back to temperature management or the like at the time of glass forming, and the measurement result cannot be reflected on the glass base plate manufactured during that time.

  Usually, the process of cutting into a 1000 mm square glass base plate is performed by a series of processes using a manufacturing apparatus that cools the plate glass formed into a plate shape after preparing and melting the raw materials. The process of cutting into a glass plate is generally performed once by taking a 1000 mm square glass base plate and using another manufacturing apparatus or manufacturing facility. The measurement result of Mesa flatness is the temperature at the time of glass molding. It took a lot of time to give feedback to management.

  Then, an object of this invention is to provide the manufacturing method of a glass substrate with high productivity.

The present invention provides the following aspects.
(1) Forming molten glass into sheet glass, cooling it, first cutting step of cutting the cooled sheet glass into a glass base, and cutting the glass base A method for producing a glass substrate for a magnetic disk, comprising: a second cutting step for forming a glass plate; and a shape processing step for processing the glass plate into a desired shape,
The manufacturing method of the glass substrate for magnetic discs which has a sampling inspection process which measures the bending of the said glass base plate between a 1st cutting process and a 2nd cutting process.
(2) In the sampling inspection process, after measuring the distance between one main surface and the sensor while supporting two opposing sides of the glass base plate, the distance between the other main surface and the sensor is measured, and the glass The method for producing a glass substrate for a magnetic disk according to (1), wherein the deflection of the glass base plate is obtained by offsetting the deflection due to the weight of the base plate.
(3) The method for manufacturing a magnetic disk glass substrate according to (1) or (2), wherein the deflection of the glass base plate in the sampling inspection step is fed back to the forming step.
(4) Preliminarily measure the correlation between the deflection of the glass base plate and the difference in height of the glass plate,
In the sampling inspection step, a sampling inspection is performed based on the correlation, and the method of manufacturing a magnetic disk glass substrate according to any one of (1) to (3).

  In the present invention, there is a correlation between the difference in height of a glass substrate for a magnetic disk in units of several microns, which has been considered to be completely unrelated, and the deflection of a glass base plate in units of hundreds of microns to several millimeters. It was made based on the discovery and hypothesis that there is.

  According to the method for manufacturing a glass substrate for a magnetic disk described in (1) of the present invention, since there is a sampling inspection step for measuring the deflection of the glass base plate between the first cutting step and the second cutting step, It is possible to shorten the time until the result of the sampling inspection is fed back and the measurement result is reflected in the forming process of the glass base plate. Thereby, productivity of a glass substrate can be improved.

  According to the method for manufacturing a glass substrate for a magnetic disk described in (2) of the present invention, the influence of the deflection due to the weight of the glass base plate is eliminated by a simple method, and the glass base plate originally possessed by the glass base plate is eliminated. The deflection of the board can be measured.

  According to the method for producing a glass substrate for a magnetic disk described in (3) of the present invention, the glass base plate can be used for forming a glass base plate with less deflection by feeding back the glass base plate deflection to the forming step.

  According to the method for producing a glass substrate for a magnetic disk described in (4) of the present invention, it is applied to various types of glass by measuring in advance the correlation between the deflection of the glass base plate and the height difference of the glass plate. can do.

(A) is the schematic diagram of what cut | disconnected the plate-like glass which flowed from the glass manufacturing apparatus to about 6000 mm x 1000 mm, (b) is the model of the glass base plate which cut | disconnected the plate-like glass of (a) into 1000 mm square. It is a figure, (c) is a schematic diagram of the glass plate which cut | disconnected the glass base plate of (b) at 75 mm square. It is a schematic diagram explaining the principle of a Mesa interferometer. It is a flowchart explaining the manufacture flow of glass. It is a schematic diagram of a glass manufacturing apparatus. It is a graph which shows the relationship between the Mesa flatness of a 75 mm glass plate, and the deflection | deviation of a 1000 mm square glass base plate. It is a graph which shows the relationship between the Mesa flatness of a 75 mm glass plate, and the deflection | deviation of a 1000 mm square glass base plate. It is the graph which plotted the relationship between the average value of the Mesa flatness of the five 75 mm square glass plates obtained from the 1000 mm square glass base plate, and the deflection | deviation of the 1000 mm square glass base plate. (A) (b) is explanatory drawing explaining the process which calculates the theoretical value of FIG.

  Hereinafter, an embodiment of a method for producing a glass substrate for a magnetic disk of the present invention will be described in detail with reference to the drawings. FIG. 3 is a flowchart showing a method of manufacturing a magnetic disk glass substrate according to an embodiment of the present invention.

  As shown in FIG. 3, the method for producing a magnetic disk glass substrate (hereinafter referred to as a glass substrate) of this embodiment typically comprises a raw material preparation step (S1), a melting step (S2), and a forming step. (S3), first cutting step (S4), sampling inspection step (S5), second cutting step (S6), shape processing step (S7), lapping step (S8), and polishing step (S9) And).

  The raw material preparation step (S1) is a step of preparing a predetermined glass raw material, and thereafter, the melting step (S2) and the forming step (S3) are performed by the glass manufacturing apparatus 20 shown in FIG. The melting step (S2) is a step of melting the glass raw material charged from the raw material charging port in the melting tank 21. In the forming step (S3), a glass ribbon is floated on tin with a float bath 22 to form a fixed width and thickness, and slowly cooled in a cooling zone 23 so as not to cause distortion inside the plate glass. Is a step of forming.

  In the float bath 22, the upper space of the glass is divided into a plurality of regions, and the operator manages the temperature of the heater 24 provided in each region, so that the width, thickness, deflection, etc. of the glass ribbon (sheet glass) can be reduced. Be controlled. In the first cutting step (S4), the sheet glass thus formed is first cut to about 6000 mm × 1000 mm, and further cut into a 1000 mm square glass base plate. The plate glass of about 6000 mm × 1000 mm is likely to be uneven as viewed from the flow direction. Here, as a typical example, as shown in FIG. 1A, a description will be given using a case where a plate glass of about 6000 mm × 1000 mm has a deflection that is convex upward when viewed from the flow direction. Note that the shape of the deflection is not limited to this, and it is needless to say that there may be almost no deflection. In the present invention, “deflection” refers to the shape (warp) of a glass base plate obtained by cooling and cutting, and excludes the influence of gravity.

  The glass base plate (FIG. 1 (b)) cut to 1000 mm square in the first cutting step (S4) also has a deflection that is convex upward. In the following description, for convenience, the convex side is referred to as the front surface, and the concave side is referred to as the back surface.

  Subsequently, the glass cut in the first cutting step (S4) is subjected to a sampling inspection in a sampling inspection step (S5) at a rate of, for example, 100 to 1000 sheets. In the sampling inspection step (S5), the deflection of the 1000 mm square glass base plate is measured.

  The second cutting step (S6) is a step of cutting a 1000 mm square glass base plate into a 75 mm square glass plate.

  In the shape processing step (S7), a through-hole (inner hole) is formed in the center of a 75 mm square glass plate and processed into a circular glass of φ65 mm, and the edge of the cut-out circular glass is chamfered. A chamfering step and an end surface polishing step of polishing the inner periphery and outer periphery to a mirror surface.

In the lapping step (S8), for example, grinding is performed while supplying a slurry containing water as a main dispersion medium containing any one of alumina abrasive grains, zirconia abrasive grains, silicon carbide abrasive grains, and diamond abrasive grains with a double-sided processing apparatus. In the step (S9), the main surface of the glass substrate is polished on the main surface of the sheet glass in the lapping step by supplying a ceria slurry in which, for example, CeO ₂ abrasive grains are dispersed in water with a double-sided processing device. It is possible to obtain a mirror-finished main surface by reducing the fine uneven shape. In addition, a lapping process (S8) and a grinding | polishing process (S9) are not limited to this, A well-known method can be employ | adopted.

  As the glass substrate for magnetic disk, lithium silicate glass, aluminosilicate glass, aluminolithium silicate glass, aluminoborosilicate glass, soda time glass, borosilicate glass, etc. are used, and preferably aluminosilicate glass is used.

Next, the sampling inspection process (S5), which is a feature of the present invention, will be described in detail.
The deflection measurement performed in the sampling inspection step (S5) is performed by supporting two sides on opposite sides of a 1000 mm square glass base plate, and using a laser displacement sensor disposed substantially in the center of the glass base plate, The distance to the back of the base plate is measured. Subsequently, the glass base plate is turned over, and the distance from the displacement sensor to the surface of the glass base plate is measured. At this time, the glass base plate of 1000 mm square has a deflection due to its own weight in addition to the deflection of the glass base plate, and the deflection of the glass base plate is canceled by offsetting the deflection due to its own weight during the front surface measurement and the back surface measurement. Is required. That is, the calculated deflection W, the distance A from the displacement sensor to the back surface when the dead weight does not act, the distance B from the displacement sensor to the front surface when the dead weight does not act, the distance a from the displacement sensor to the back surface when the dead weight is considered When the distance b from the displacement sensor to the surface in consideration of its own weight is taken, the calculated deflection W is expressed by the following formula 1.

  Here, the relationship between the Mesa flatness of a 75 mm square glass plate using a conventional Mesa interferometer and the deflection of the 1000 mm square glass base plate in the sampling inspection step (S5) will be examined.

  5 and 6 are graphs showing the relationship between the Mesa flatness of a 75 mm square glass plate using a Mesa interferometer and the deflection of a 1000 mm square glass base plate, and FIG. 5 is a glass with a relatively good height difference. FIG. 6 shows a case of a glass base plate having a relatively low height difference. The Mesa flatness is measured using a Mesa interferometer (ZYGO Mesa product number: 99-32-66055), and the deflection is measured by the above-described deflection measurement using a displacement sensor LK-G85 manufactured by Keyence Corporation. It is. The glass plate is alkali-alumino-silicate glass, and its thickness is about 0.1 mm.

  In the figure, A, B, C, D, E, and F are numbered sequentially from the left to the glass base plate of about 6000 mm × 1000 mm shown in FIG. Is. In addition, A1, A2, A3, A4, and A5 are obtained by randomly extracting five from the left from a 1000 mm square glass base plate A cut into a 75 mm square glass base plate and numbering them sequentially from the left. is there. The Mesa flatness and the deflection of the glass base plate at the center of each 1000 mm square glass base plate A are shown in the figure. The same applies to B, C, D, E, and F.

  As can be seen from FIGS. 5 and 6, the Mesa flatness of the 75 mm square glass plate and the deflection of the 1000 mm square glass base plate were almost the same. Further, comparing FIG. 5 and FIG. 6, it was confirmed that when the deflection of the 1000 mm square glass base plate is increased (deteriorated), the Mesa flatness is also deteriorated. This result proves that the hypothesis that there is a certain correlation between the height difference of the glass substrate for magnetic disks in units of several microns and the deflection of the glass base plate in units of hundreds of microns to several millimeters is correct.

  Conventionally, there is a difference in the order of several hundred times between the difference in height of a glass substrate for magnetic disks in units of several microns and the deflection of a glass base plate in units of hundreds of microns to several millimeters. Since it was thought that the deflection of the glass base plate was released when the base plate was cut into a 75 mm square glass plate, there was a correlation between the height difference of the glass substrate for magnetic disks and the deflection of the glass base plate. I thought it was impossible. However, as a result of inventor's diligent study, as can be seen from the measurement results of FIGS. 5 and 6, the knowledge that there is a correlation between the height difference of the glass substrate for magnetic disks and the deflection of the glass base plate was obtained. Invented.

  FIG. 7 is a graph plotting the relationship between the Mesa flatness of a 75 mm square glass plate and the deflection of a 1000 mm square glass base plate. The Mesa flatness is an average value of the Mesa flatness of five 75 mm square glass plates obtained from a 1000 mm square glass base plate. In the figure, ○ (actual measurement 1) and □ (actual measurement 2) have different measurement dates and times. Moreover, the dotted line in a figure shows the theoretical value of the Mesa flatness of a 75 mm square glass plate, and the bending of a 1000 mm square glass base plate.

Here, a theoretical value calculation method will be described with reference to FIGS.
As shown in FIG. 8A, assuming that a 1000 mm square glass base plate is warped in an arc shape with a radius r, and assuming that the plate width of the glass base plate is T and the deflection is W, the radius r is three squares. From the theorem, it can be expressed by Equation 2.

  On the other hand, as shown in FIG. 8B, when a 1000 mm square glass base plate is cut into a 75 mm glass plate, assuming that the shape is maintained, the radius of curvature r is the same. When the plate width of the glass plate is t and the Mesa flatness is F, the radius r can be expressed by Equation 3 from the three-square theorem.

  Equation 4 is obtained by Equation 2 and Equation 3.

Here, if {W ² + (T / 2) ² } / 2W is B and (t / 2) ² is C, Formula 4 is Formula 5
When this is solved for the Mesa flatness F, it is expressed by Equation 6.

F = B− (B ² −C) ^{1/2 according to the} subject matter, when the deflection W is 0.1 mm, the plate width T of the glass base plate is 1000 mm, and the plate width t of the glass plate is 75 mm, the Mesa flatness F is 0.56 μm.

  Comparing the theoretical value calculated in this way with the actual measurement value, the difference in height of the glass substrate for magnetic disks in units of several micrometers and the deflection of the glass base plate in units of hundreds of micrometers to several millimeters is several hundred times. Although there is a difference, it was confirmed that there was no big difference between the theoretical value and the actual measurement value compared to the difference of several hundred times. Thus, it was theoretically confirmed that there is a correlation between the deflection of the 1000 mm square glass base plate and the 75 mm square Mesa flatness. Note that the difference between the calculated theoretical value and the actually measured value is assumed to be that part of the deflection is restored due to the stress released by cutting.

  Accordingly, the correlation between the deflection of the 1000 mm square glass base plate and the 75 mm square Mesa flatness is obtained in advance, and after the first cutting step (S4), before the second cutting step (S6), the 1000 mm square is obtained. By performing a sampling inspection to measure the deflection of the cut glass base plate, it is possible to quickly feed back to the temperature control at the time of glass forming. Thereby, productivity can be improved compared with the case where it passed through two cutting processes like the past.

  For example, if there is a correlation as shown in the graph of FIG. 7 between the deflection of a 1000 mm square glass base plate and the 75 mm square Mesa flatness, in order to reduce the 75 mm square Mesa flatness to 6 μm or less, The deflection of the 1000 mm square glass base plate may be 0.3 mm or less. Therefore, if the deflection of the 1000 mm square glass base plate exceeds 0.3 mm as a result of the sampling inspection, the glass base plate in the float bath is notified by notifying the operator of the molding process of the deflection of the 1000 mm square glass base plate. The temperature of the heater around the region that has passed through can be controlled in accordance with empirical rules, etc., and can be used for forming a glass base plate with less deflection.

  Also, from the graph shown in FIG. 7, the relationship between the Mesa flatness of the 75 mm square glass plate and the deflection of the 1000 mm square glass base plate is, for example, W = 25F, and the allowable width is set based on W = 25F. When the thickness is out of the range, the operator of the forming process may be notified of the deflection of the 1000 mm square glass base plate, which may be used for forming the glass base plate with less deflection.

  As described above, according to the manufacturing method of the present embodiment, the first cutting step (S4) for cutting the plate glass to form a 1000 mm square glass base plate, and the 1000 mm square glass base plate with a φ65 mm magnetic plate. By performing a sampling inspection step (S5) for measuring the deflection of a 1000 mm square glass base plate before the second cutting step (S6) for cutting into a plurality of 75 mm square glass plates slightly larger than the disk glass substrate, The result of the sampling inspection step (S5) can be immediately fed back to the temperature management during glass forming.

  In the sampling inspection step (S5), the opposing two sides of the glass base plate are supported and the distance between one main surface and the sensor is measured, and then the distance between the other main surface and the sensor is measured. By calculating the deflection of the glass base plate by offsetting the deflection due to the weight of the plate, the influence of the deflection due to the self weight of the glass base plate is eliminated by a simple method. Can be measured. Note that the deflection measurement is not limited to the measurement method of the present embodiment, and any method can be employed. For example, in the said embodiment, although both surfaces of the glass base plate were measured, only the distance of one main surface may be measured, and the deflection amount by the dead weight calculated beforehand may be subtracted from the obtained measured value. Alternatively, the amount of deflection may be measured by optically capturing a three-dimensional image.

  Further, by feeding back the deflection of the glass base plate in the sampling inspection step (S5) to the molding step, it can be used for molding of the glass base plate with less deflection.

  In addition, the correlation between the deflection of the glass base plate and the height difference of the glass plate is measured in advance, and the sampling inspection step (S5) is applied to various types of glass by performing a sampling inspection based on this correlation. can do.

The present invention is not limited to the embodiment described above, and can be implemented in various forms without departing from the gist of the present invention.
For example, in the embodiment described above, the manufacture of a Φ65 mm glass substrate used for, for example, a notebook personal computer has been described as an example. However, the present invention is not limited to this. It can be applied to a size glass substrate.

  Moreover, although this embodiment demonstrated the glass shaping | molding by the float method to an example, it can apply not only to this but to a downdraw method.

DESCRIPTION OF SYMBOLS 10 Mesa interferometer 11 Laser diode light source 12 Diffraction grating 13 Camera array 14 Piezoelectric transducer 20 Glass manufacturing apparatus 21 Dissolution tank 22 Float bath 23 Cooling zone

Claims (3)

Forming a molten glass into a sheet glass and cooling the glass; a first cutting step of cutting the cooled sheet glass to form a glass base; and cutting the glass base to a glass plate A method for producing a glass substrate for a magnetic disk having a second cutting step and a shape processing step of processing the glass plate into a desired shape,
Between the first cutting step and the second cutting step, it has a sampling inspection step for measuring the deflection of the glass base plate,
The sampling inspection step supports two opposing sides of the glass base plate and measures the distance between one main surface and the sensor, then measures the distance between the other main surface and the sensor, A method of manufacturing a glass substrate for a magnetic disk, wherein the deflection of the glass base plate is obtained by offsetting the deflection due to its own weight.   The method of manufacturing a magnetic disk glass substrate according to claim 1, wherein the deflection of the glass base plate in the sampling inspection process is fed back to the forming process. Measure in advance the correlation between the deflection of the glass base plate and the difference in height of the glass plate,
3. The method of manufacturing a magnetic disk glass substrate according to claim 1, wherein the sampling inspection step performs a sampling inspection based on the correlation.

Images