JP2010264471A - Thermal stress cracking for brittle material by wide region non-uniform temperature distribution - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a method and a device which achieve both of high cracking speed and high cracking position precision without generating thermal damage on a workpiece by the heating of a material in high quality thermal stress cracking for a brittle material which does not generate cullets and microcracks which have been inevitable in the conventional mechanical scribing method. <P>SOLUTION: In a region wide as much as possible on a workpiece, a non-uniform heating temperature distribution of relatively low temperature which is slowly distributed is provided so as to reduce heating temperature for generating stress, and the thermal damage of the workpiece is prevented. Besides, heating energy concentrated on a relatively micro region as a cracking position decision factor is superimposed on the heating temperature distribution, also, the position is off-set, or negative feedback control is performed, and further, if required, positive feedback control is performed so as to improve cracking position precision. As the heating laser, a CO<SB>2</SB>laser, an Er:YAG laser capable of achieving full-cuts or an Fe<SP>+2</SP>:ZnSe laser capable of selectively achieving full-cuts and extremely deep scribes at various sheet thicknesses is used. <P>COPYRIGHT: (C)2011,JPO&INPIT

Description

The present invention relates to a thermal stress cleaving method for brittle materials, especially glass for flat panel displays and glass for solar cell substrates. Hereinafter, although the case of glass will be introduced as an example of a workpiece, the present invention can be applied to all brittle materials.

Recently, in the glass cleaving, the thermal stress cleaving method by laser light irradiation has been used in place of the diamond chip mechanical method which has been used for the past several centuries.

  According to this method, defects inherent in the mechanical method, that is, a decrease in glass strength due to the occurrence of microcracks, contamination due to the occurrence of cullet at the time of cleaving, existence of a lower limit value of the applied plate thickness, and the like can be eliminated.

As a result, according to the thermal stress cleaving method, polishing and cleaning, which are subsequent processes of mechanical cleaving, are not required, a mirror surface having a surface roughness of 1 μm or less is obtained, and the product external dimension accuracy is ± 25 μm or more. Furthermore, it can be used for glass plates having a thickness of up to 0.1 mm, and can be used for future flat panel display glass and solar cell substrate glass.

  The principle of this method is as follows. The glass is heated to the extent that cracks, melting, vaporization, etc. do not occur by laser irradiation locally. At this time, the glass heating section tries to expand thermally, but cannot sufficiently expand due to the reaction from the surrounding glass, and compressive stress is generated around the irradiation point. Even in the peripheral non-heated region, the peripheral portion is further distorted by the expansion from the heating portion, and as a result, compressive stress is generated. These compressive stresses are radial. By the way, when an object has a compressive stress, a tensile stress related to the Poisson's ratio is generated in the orthogonal direction. Here, the direction is a tangential direction. This is shown in FIG.

This figure shows the change of the radial stress component σ _x and the tangential stress component σ _y depending on the location when there is a temperature rise in the Gaussian distribution centered at the origin. The former is compressive stress from the beginning to the end (negative value in the figure), while the latter is compressive stress at the heating center, but changes to tensile stress (positive value in the figure) when leaving the center. Based on the stress distribution shown in the figure, a general temperature distribution can be handled as a linear superposition of the stress distribution. Even in that case, there is no difference that compressive stress and tensile stress are generated in the orthogonal direction at each point on the glass plate.

Of these stresses, the tensile stress is directly related to the cleaving. When the stress exceeds the fracture toughness value, which is a material specific value, fracture occurs everywhere and cannot be controlled. In the case of the thermal stress cleaving as in the present invention, since the tensile stress is selected to be equal to or less than the same value, no fracture occurs.

However, when there is a crack at the position where the tensile stress is present, stress expansion occurs at the tip, and the crack expands when the same force exceeds the fracture toughness value of the material. That is, controlled cleaving as processing occurs. By scanning the laser irradiation point, the crack can be extended. In this thermal stress cleaving, the fractured surface is similar to the cleaved surface of the crystal, so there is no microcracking or cullet generation, the disadvantages of the mechanical method described above can be eliminated, and extremely excellent properties as a glass processing method Will have.

  There are two types of thermal stress cleaving shown in FIG. FIG. 13 (a) shows that cracks 4 occur only in the surface layer (for example, 100 μm depth) of the work 1 as a result of heating 2 and cooling 3, and this is generally called surface scribe. 5 indicates the scribe direction.

On the other hand, what is shown in FIG. 13B is one in which cracks occur over the entire thickness of the workpiece and is generally referred to as full cut. 5 indicates a cleaving direction, and 6 indicates a cleaving line. In the case of the full cut shown in the figure, cooling is not performed. Of course, cooling is advantageous for the formation of tensile stress, but there is no cooling method that reaches the inside of the workpiece in bulk. Both have advantages and disadvantages. The former requires a break process after scribing, which is the biggest drawback. This process is not necessary for the latter, but the technique has a drawback of a so-called size effect such that the cleaving speed is lowered for a large workpiece and the cleaving position accuracy is lowered near the end of the workpiece.

This surface scribing technique by laser irradiation of glass has been vigorously developed by Kondratenco Bradymier Stepanovic, and a Japanese patent of Patent Document 1 has been established. This patent is also established as a European patent and a US patent in Patent Documents 2 and 3. He selected CO ₂ laser light as the laser light.

Similarly, a Japanese patent of Patent Document 4 has been established in which a surface sliver is performed using a CO ₂ laser, and a technique of converting a circularly symmetric laser beam into a plurality of beam spot arrays arranged in a straight line by a beam splitter.

In response to the above-mentioned surface scribe patent, the development of a full-cut technique shown in FIG. 13 (b) was pursued to eliminate the above-mentioned drawbacks, and Japanese Patent Application Laid-Open No. 2005-228688, which is a technique for irradiating a rare earth element doped glass with a semiconductor laser. A patent has been granted.

  Whether thermal stress cleaving is a full cut or surface scribe depends on whether the workpiece heating is a surface phenomenon or a bulk phenomenon, depending on whether the thermal stress occurs only in the surface layer or over the entire thickness of the workpiece. It depends on the difference in occurrence. As described above, since both have merits and demerits, they may be properly used according to the purpose. This patent is applicable to both technologies in common and is an improved patent that enables further performance improvement of brittle material thermal stress cleaving.

Kondratenko V.S., method for cutting brittle non-metallic materials, Japanese Patent No. 3027768 Kondratenko Vladimir S., Method of splitting non-metallic materials, EP0633867B1 Kondratenko Vladimir S., Method of splitting non-metallic materials, USP5609284 Tsutomu Teramoto, cutting device, Japanese Patent No. 3792639 Hiroshi Miura, Norio Kabebe, Cleaving method and apparatus for brittle materials, Japanese Patent No. 4179314

Vladimir V. Fedorov et al. 3.77-5.05 μm Tunable Solid-State Lasers Based on Fe2 + -Doped ZnSe Crystals Operating at Low and Room Temperatures, IEEE Journal of Quantum Electronics, Vol. 42, No. 9, pp 906-917, September (2006).

The thermal stress cleaving due to laser irradiation or the like is characterized by a higher quality processing method than a mechanical method. On the other hand, this processing method also has drawbacks. The first is that the processing point temperature is higher than room temperature. In contrast to the mechanical method that can be performed at room temperature, thermal stress cleaving must be heated above room temperature. In particular, when full cutting is performed with a large workpiece, it is necessary to increase the thermal stress in order to increase the cleaving speed. For this purpose, it is necessary to further increase the heating temperature. On the other hand, glass has a denaturation point temperature and melting point, and especially when there is a film on the glass surface, it is necessary to lower the heating temperature to prevent workpiece damage, which is the biggest problem in thermal stress cleaving. become.

  In order to solve this problem, in the present invention, the workpiece heating is not limited to one point on the workpiece, but is distributed over a wide area. Then, the generated thermal stress can be remarkably increased by dividing the thermal stress generated at each point of the heating region by the area in the entire heating region. However, thermal stress is not generated when the entire workpiece is uniformly heated. Rather, in order to remove the residual stress in the work, it is often performed that the work is put into a heating furnace and gradually heated and cooled. In order to generate thermal stress, a non-uniform temperature distribution is necessary. The distribution can be formed by heating or cooling in principle, but in practice heating is easier to carry out. It is also effective to use both in combination. By carrying out such distributed heating, we were able to cut the large non-alkali glass having a thickness of 0.7 mm at a low temperature of 50 ° C. to fully cut at a breaking speed of 300 mm / second.

This method is effective in reducing the heating temperature at the processing point. However, if the temperature distribution is gently wide, there is no determinant of the cleaving position, so that the same accuracy is lowered. In order to prevent this, a sharp peak may be superimposed on a gentle temperature distribution and used as a cleaving position determining factor. However, in general, the temperature distribution and the stress distribution that can be used for dissociation between atoms do not match. The latter actually contributes to the work cleaving, which depends not only on the temperature distribution but also on the overall structure of the workpiece. Therefore, a high cleaving position accuracy cannot be realized simply by matching the temperature distribution to the cleaving line. The accuracy can be improved by an offset technique when the deviation between the two can be accurately predicted by theory or experiment, and by negative feedback control or, if necessary, positive feedback control when the prediction is difficult.

According to the present invention, while realizing the high quality of thermal stress cleaving, the heating temperature is lowered to a sufficiently low value without damaging the glass itself or the film formation on the surface, and the positional accuracy is sufficiently increased. Can do. Laser glass breaking has many great technical advantages in processing quality, but has not been able to replace the diamond tip mechanical system that has been used over the past centuries. The present invention changes such a situation.

Advantages of the thermal stress cleaving of glass realized by the present invention, particularly the full cut technology, are as follows.
1) Full cut can be realized at a significantly higher speed than the conventional method.
2) The cleaving can be done in one process only with a full cut, and no break is required. In addition, post-processing such as polishing and cleaning is unnecessary.
3) There is no generation of micro cracks in the vicinity of the fractured surface, and the material strength of the workpiece becomes a high value.
4) There is no adhesion of cullet on the cut surface and it is clean.
5) The contour can be cleaved.
6) The cleaving position accuracy is increased.
7) The fractured surface is sufficiently perpendicular to the glass surface.
8) The fractured surface is a mirror surface and the surface roughness is good.
9) The selective cutting of the laminated glass can be performed by laser beam irradiation from one direction, and operations such as reversing the glass plate are not necessary.
10) The cleaving can be automated.

The schematic diagram which shows the thermal stress cleaving principle by wide area | region nonuniform temperature distribution. The generated thermal stress distribution diagram by Gaussian temperature distribution. An example of a relatively low temperature wide area non-uniform heating distribution for thermal stress generation. The stress intensity factor distribution map generated under an example of a wide area non-uniform heating distribution. Stress intensity factor distribution for laser spot scanning. Stress intensity factor distribution when scanning laser spot is superimposed on wide area non-uniform heating. Conversion of Gaussian laser beam to line beam by DOE. Rotation control of line beam direction when cleaving thermal stress along the contour. Control of line beam length and orientation during thermal stress cleaving along non-smooth contours. Deviation between the actual cleaving position and the target position when the thermal stress cleaving near the workpiece side edge. Infrared transmittance characteristics of a non-alkali glass plate having a thickness of 0.7 mm. An example in which a full cut direction and a very deep scribe direction are orthogonal to each other when cleaving a large number of workpieces. Schematic diagram of surface scribe (a) and full cut (b) by laser irradiation of glass.

  A schematic diagram of an embodiment of the present invention is shown in FIG. As described above, the longer the heating region is, the more advantageous it is that the heating temperature may be lowered. On the other hand, the cleaving position accuracy also decreases. In order to prevent this, a temperature distribution is used in which a scenic mountain range 7 is superimposed on the ridgeline of a mountain range having a wide temperature distribution, as shown at 11 in FIG. In this case, the sharp ridge line position 7 becomes a cleaving position determining factor. The thermal stress value as an integral value over a wide heating range becomes large despite the fact that the height of the mountain is relatively low, yet it does not reach the cleaving threshold. Set to exceed. In the figure, a region where the thermal stress is large enough to cause interatomic bond dissociation is indicated by 10. In the same region, the state in which the interatomic bond indicated by the line segment between adjacent atoms indicated by ◯ is dissociated is indicated by x. 8 indicates the cleaving by full cut, and 9 indicates the crack tip. 5 is a cleaving direction. In this case, the sharper the summit, the better the cleaving position accuracy.

According to linear elastic fracture mechanics, cleaving occurs due to the singularity of the thermal stress field near the crack tip on the workpiece. A thermal stress value that deviates from this singularity is set such that it cannot overcome the bonding force between adjacent atoms. Even if thermal stress is generated under the condition of the singularity near the crack, the value exceeding the value used for workpiece deformation can be used for atomic bond dissociation for the first time. The former mainly depends on the workpiece structure, and the larger the workpiece, the larger the value. The latter depends on the workpiece material constant and is independent of the workpiece shape. According to the linear elastic fracture mechanics theory, the physical quantity K ₁ (opening stress intensity factor) representing the latter is K ₁ > K _1c (where K _1c is the fracture toughness value of the material, and 0.73 MPa / In the case of m ² ), the cleaving proceeds as the crack tip progresses. In the present invention, the temperature distribution may be selected so that K ₁ > K _1c only at the required cleaving position. In that case, it is designed so that this condition can be realized at the lowest possible temperature. Since the method for calculating the K ₁ is known given by linear elastic fracture mechanics theory, explanation is omitted here.

  According to this embodiment, the heating temperature can be reduced as an effect of the non-uniform heating of the wide area, and the thermal damage of the workpiece can be avoided. Moreover, high cleaving position accuracy can be realized by superimposing a cleaving position determining factor on wide area heating. As a result, the inherent high quality cleaving characteristics of thermal stress cleaving can be utilized.

As a first embodiment, a first specific heating temperature distribution design example is introduced. It covers the case of linear full cut cleaving at the center of an alkali-free glass plate having a width of 580 mm and a thickness of 0.7 mm. This is a case where a full cut is quite difficult due to the inevitable size effect for large workpieces. As a gentle mountain range in FIG. 1, a Gaussian temperature distribution having a peak at a temperature of about 50 ° C. at the center of the glass width as shown in FIG. 3 is set. For the realization, a columnar infrared lamp with an electric input of 1 kW installed in parallel with the glass length at a position of 200 mm in height on the glass plate was used. FIG. 3 also shows the measured value of the glass surface temperature 3 minutes after power-on, which is the same time as the cleaving operation. Here, after waiting for 3 minutes from the start of the lamp, the lamp light was partially absorbed by the glass surface and reached the inside by heat conduction, and the condition for full cut was completed after a certain period of time. Because. If the infrared lamp light is irradiated through a 3-4 μm bandpass filter, the irradiated light reaches the inside of the glass and is directly absorbed, so this condition is achieved simultaneously with the lamp lighting and no waiting time is required.

FIG. 4 shows the glass length direction distribution of the K ₁ value at the center of the glass width calculated by the finite element method with respect to this temperature distribution. In the figure, the value of K _1c for non-alkali glass is also shown. Region K ₁ becomes larger than K _1c is an area from the cleaving start end portion of the glass to 310 mm, full cut Beyond this point would be stopped. In heating under this condition, full cutting automatically proceeds from the end to 310 mm and stops there.

Increasing the heating temperature will increase the full cut travel distance, and decreasing it will decrease it. In any case, the cleaving position accuracy cannot be expected with wide heating as shown in FIG. The merit under this condition is that the maximum heating temperature could be reduced to 50 ° C.

A single laser spot with an absorption laser output of 100 W, a spot diameter of 3 mm, and a scanning speed of 300 mm was applied for superposition heating as a cleaving position determining factor in the present invention. First it was K ₁ calculated by the finite element method with respect to only the laser beam heating. The actual experiment was performed using an LD laser beam having a wavelength of 976 nm. FIG. 5 shows the distribution in the glass length direction of the K ₁ value calculation result on the laser beam scanning line. This figure shows the K ₁ distribution and the temperature distribution when the scanning laser beam position reaches the position of 30, 100, 200, 250, 300, 400, 510 mm from the glass edge, respectively. The heating only by the laser, K ₁ is lower than the K _1c except in the vicinity of glass edges, fracture does not occur.

The gist of the present invention is to realize a heating temperature distribution as shown in FIG. 1 by superimposing heating by an infrared lamp and a laser beam. FIG. 6 shows the distribution in the glass length direction of the K ₁ value calculation result at the time of superimposition. This figure shows the K ₁ distribution and temperature distribution when the laser beam position is at the same point as in FIG. Here, the K ₁ value and the temperature at each point are the sum of the values shown in FIGS. The superposition principle is established for these physical quantities. As a result, K ₁ only in the planned cutting line may if to exceed the K _1c. The K ₁ value shown in FIG. 6 is larger than K _1c up to the vicinity of the laser beam, so that the cleaving proceeds there and stops there. The reason for stopping is that there is a strong compressive stress in the vicinity of the laser beam.

In the case of heating by the infrared lamp shown in FIG. 3, the cleaving proceeded very rapidly within the range of K ₁ > K _1c . This is because the propagation speed of the thermal stress distribution in the workpiece is almost close to the speed of sound. In this case, the fractured surface is roughened. On the other hand, in the case of the infrared lamp and laser beam heating superposition shown in FIG. 6, since the cleaving crack cannot exceed the laser beam as described above, the cleaving speed is the same as the laser beam scanning speed. This method is convenient because high cleaving quality can be maintained.

  Presented here is a non-alkali glass with a plate thickness of 0.7 mm and a width of 580 mm by superimposing scanning laser spot irradiation on heating at a considerably low temperature of 50 ° C. with a columnar infrared lamp, which is quite low temperature heating. An embodiment in which high-grade thermal stress cleaving can be performed at a speed of 300 mm at the center of the plate, which is an effect that none of the conventional techniques can achieve, and demonstrates the high functionality of the present invention.

The second configuration example of the heating temperature distribution shown in FIG. 7 uses a diffraction grating optical element (DOE) 14 that uses a Gaussian laser beam 13 emitted from the laser oscillator 12 as a line beam homogenizer as a cleaving position determining factor. The one converted into a line beam is used. FIG. 7 is a schematic diagram showing the state of this conversion. In the figure, reference numeral 15 denotes a laser beam after passing through the DOE, and a line beam 16 which is a design specification is realized at a predetermined position. About another structure, it is equal to a 1st Example. As this line beam, a beam having a width of about 0.1-1 mm and a length of about 25 mm is typically used. In this case, the intensity distribution is uniform in the length direction, but the width direction can be uniform or Gaussian. Also, the intensity distribution can be changed by adjusting the incident laser beam diameter, position, and the like.

Compared to the first embodiment, in the case of the second embodiment, the cleaving position accuracy is further increased, the required heating temperature can be further reduced, and in the case of surface scribe, the scribe depth can be increased. These are the effects of this configuration example.

A third configuration example of the present invention is shown in FIG. When contour cutting which is a free curve is performed, it is necessary that the line beam 16 as a cutting position determining factor be matched with the tangent of the contour. For this reason, it is necessary to constantly control the direction of the line beam. This can be done by the DOE rotation control 17.

As an effect of this configuration, since the major axis direction of the line beam always coincides with the contour line, there are an improvement in cleaving position accuracy, an increase in generated thermal stress, and a resulting cleaving speed.

A fourth configuration example of the present invention is shown in FIG. When the cutting curve is bent at random, it is not sufficient to control the direction of the line beam, and it is necessary to shorten the length or, depending on the case, to form a dot. Such control can be performed by using the wedge-shaped mask 19 that moves back and forth while simultaneously rotating with the DOE simultaneously with the rotation control of the DOE. In the figure, this back-and-forth movement is indicated by 20, and the length of the line beam 16 can be controlled to a required value as indicated by 21, 22, 23 by the control. In particular, 21 shows a point-like shape, which corresponds to an acute angle curve of the cleaving curve.

As an effect of this configuration, even when the contour curve is bent with a small curvature or bent at an acute angle, it is possible to prevent the cutting position accuracy from being lowered and the cutting speed from being lowered.

The thermal stress cleaving occurs at the maximum thermal stress position used for interatomic bond dissociation, not at the maximum temperature position. What we can give as input conditions is the temperature distribution, not this thermal stress distribution. The latter appears as a result depending on conditions such as the workpiece structure. Of course, there is a causal relationship between the two, but it is difficult to always grasp such a relationship at the actual production site. According to the principle of the present invention shown in FIG. 1, the actually generated thermal stress cleaving position is uniquely limited to any place near the determinant, but it is not necessarily the cleaving target position. Some kind of position control technology is necessary for high-precision control.

When the deviation between the temperature distribution and the position where the cleaving occurs can be theoretically specified, high cleaving position accuracy can be realized by setting the temperature distribution to be offset from the beginning to compensate for the deviation. FIG. 10 shows an example of the deviation calculated by the finite element method and proved in the experiment. This is a workpiece width of 500 mm, length of 1000 mm, thickness of 0.7 mm and width of 2000 mm, length of 2000 mm, thickness of 0.7 mm, and a diameter of 1.5 mm at a position 20 mm from the side edge. The actual cleaving position when the laser beam is scanned in the length direction is shown. Except for the cleaving start and the vicinity of the end, a maximum positional displacement of 0.7 mm occurs, and a high degree of reproducibility is observed. If a positional deviation of 0.7 mm is allowed, it can be left as it is. If it is not allowed, position control is required. If this deviation amount is calculated and recorded by the finite element method under various conditions, it is possible to set an offset of the thermal stress distribution and to realize high cleaving position accuracy. This is the fifth configuration example of the present invention.

With this configuration, it is possible to further increase the accuracy of the thermal cleavage position. In repetitive work under repetitive conditions in which conditions such as the workpiece shape and cleaving position are unchanged, this method may be appropriate if the allowable position error is about 50 μm. The device does not require a negative feedback control unit, and the device configuration can be simplified.

The negative feedback control of the laser beam position is the sixth configuration example of the present invention. Since the offset setting described above may be difficult at the production site, the actual cleave position is detected with a scribe monitor, and negative feedback control or positive feedback control is performed on the irradiation laser beam position corresponding to the cleaving position determination factor. It is possible to achieve a predetermined cleaving position accuracy. This control can be performed by applying a known control technique.

In contrast to the above-described high accuracy of the cleaving position by the laser beam position control, the cleaving position can be highly accurate by the line laser beam rotation angle control, which is the seventh configuration example of the present invention. The principle of this method is that the K ₁₁ (in-plane shear stress intensity factor) can be made zero or very small by controlling the rotation angle of the line laser beam, in which case the maximum temperature position matches the maximum thermal stress position. Take advantage of what you do. This fact could be confirmed by finite element calculation. In this case as well, there are two methods, offset setting and negative or positive feedback control, as with laser beam position control. However, since the calculation of the offset compensation amount is quite complicated, it is problematic for use in an actual production site, and negative or positive feedback control is more realistic. This method can also be performed by applying a known control technique.

According to the two types of negative feedback or positive feedback control, it is not necessary to evaluate in advance the amount of deviation between the temperature distribution and the cleaving position, and the high accuracy of the cleaving position can be realized by automated operation. There are immense benefits.

As the heating laser, a laser beam absorbed by glass can be used. A typical laser is a CO ₂ laser with a wavelength of 10.6 μm. This laser is convenient because a high-power laser can be easily obtained as a commercial laser. The feature of this laser is that the absorption coefficient by the glass is excessive, and 99% of the incident energy is absorbed by the surface layer having a depth of 3.7 μm, and the glass cannot be heated in bulk. Therefore, it is suitable for surface scribing, but cannot be used for full cut.

FIG. 11 shows infrared light transmission characteristics (%) of a non-alkali glass plate having a thickness of 0.7 mm. As the full-cut laser beam, it is ideal that about 50% is absorbed by the glass plate, so that a laser beam having an oscillation wavelength in the wavelength range of 2.75 to 4.5 μm can be used. Of course, it is sufficient that the shortest wavelength is 2.5 μm when the plate thickness is large, and the longest wavelength is within the range of 5 μm when the plate thickness is small. For this reason, an Er: YAG laser (2.94 μm) can be used as the full-cut laser.

The use of a mid-infrared laser that is tunable in the wavelength region of 3.77-5.05 μm can produce an ideal effect. This laser is a divalent iron ion-doped ZnSe crystal laser excited by a disk laser or a fiber laser. Oscillation efficiency is nearly 50%, and high output can be expected. The laser technology is reported in Non-Patent Document 1.

Currently, the thickness of the flat panel display glass tends to gradually decrease from the current value of 0.7 mm to 0.1 mm. The condition that 99% of the incident light energy is absorbed in these plate thicknesses is that the absorption coefficient α is in the range of 65-230 cm ⁻¹ from the relational expression I / I ₀ = e ^−αx . The wavelength range determined from the absorption characteristics of the alkali-free glass corresponding to this is well within the wavelength tuning range of this laser. Therefore, even if the glass plate thickness changes, this wavelength tuning allows almost all of the light energy to reach the back side of the glass plate and never penetrate the glass plate, making it possible to make the scribe depth close to the glass plate thickness. It is. If it becomes full cuts, cross processing at the intersection is difficult, but it is proved that cross processing can be performed if a non-cut layer remains on the back surface. This is shown in FIG. First, laser glass penetration by wavelength tuning is performed on the glass plate 1 to form deep scribes 24, 241, 242, 243, etc. that do not reach the back surface. Next, wavelength tuning is performed in the shorter wavelength direction, and the laser light is completely transmitted through the glass plate to realize full cuts 25, 251, 252, etc. in the direction orthogonal to the previous scribe. At the cross point of both lines, the full cut proceeds without any problem because the back surface is not yet separated at the scribe position. In this description, the initial crack provided at the end of the glass plate is omitted.

After these steps, a machine break is performed along the deep scribe line. This break is easy because the scribe depth is sufficient. During these processes, the workpiece is fixed to the electrostatic attraction plate 26 during the necessary processes so that the full-cut pieces do not fall apart. The suction plate may be structured so that voltage application at each position can be controlled independently, and suction may be released at a required timing at a required timing.

With this configuration, it is possible to realize a full cut for various glass thicknesses, or to achieve a very deep scribe that does not lead to a full cut. This technique is very useful in practice because it can realize cross cleaving when performing cross cleaving by thermal stress cleaving.

What has been described above are several embodiments for realizing the functions of the present invention, and it goes without saying that the spirit of the present invention can be realized in many other ways.

In this way, if the thermal stress cleaving of glass is introduced into the manufacturing process of flat panel displays and solar cells, the effects cannot be measured in improving the processing speed, processing quality, economics, etc., and overcoming the weaknesses of the conventional technology. Become a thing. These processes are currently performed with a diamond cutter, which presents problems such as the necessity of a cleaning step after cutting for generating cullet and a reduction in material strength due to the presence of microcracks. Such a problem can be solved by the thermal stress cleaving according to the present invention. In addition, since the heating temperature is sufficiently low, thermal damage is not caused to the glass substrate or the film formed on the surface thereof. Since the thermal stress cleaving position accuracy is sufficiently high and the heat-affected region can be sufficiently accommodated within the street width, the technology of the present invention can be used not only for substrate glass but also for cell processing.

1 Work
2 Heating 3 Cooling 4 Surface scribing 5 Scribing direction or cleaving direction 6 Cleavage line 7 Cleavage position determining factor 8 Full cut cleaving 9 Crack tip 10 Interatomic bond dissociation 11 based on thermal stress 11 Wide region non-uniform temperature distribution 12 Laser oscillator 13 Gaussian Laser beam 14 Diffraction grating type optical element (DOE)
15 Laser beam 16 after passing through DOE 16 Line beam cross-sectional shape 17 DOE rotation direction 18 Line beam rotation direction 19 Wedge mask 20 Wedge mask moving direction 21 Point laser spot 22 Short line beam 23 Long line beam 24 Very deep scribe 241 Same as 242 Same as 243 Same as 25 Full cut 251 Same 252 Same 26 Electrostatic suction plate

Claims (8)

A non-uniform temperature distribution that is lower than the thermal damage temperature of the workpiece and a temperature distribution concentrated in a minute region as a cleaving position determining factor are superimposed on a wide region on the workpiece, and the opening type stress intensity factor K is only in the vicinity of the minute region. ₁ is set to a value larger than the fracture toughness value K _1c of the material, and an offset is set in the temperature distribution of the cleaving position determining factor as necessary so that the cleaving position matches the target position, or In a thermal stress cleaving method for a brittle material that performs negative feedback control and, in some cases, positive feedback control, a line beam obtained by converting laser light by a diffraction grating type optical element is used as a cleaving position determining factor. 2. The method according to claim 1, wherein the direction or length of the line beam is controlled in accordance with the shape of the cleaving contour at the cleaving position. _{2. The} laser according to claim 1, wherein one of a CO ₂ laser, a Fe ²⁺ : ZnSe laser, and an Er: YAG laser is used as a laser according to a necessary cleaving type such as surface scribe, very deep scribe, and full cut. In Claim 1, it uses the electrostatic attraction | suction board which can switch voltage application independently for every place for workpiece | work fixation. A non-uniform temperature distribution that is lower than the thermal damage temperature of the workpiece and a temperature distribution concentrated in a minute region as a cleaving position determining factor are superimposed on a wide region on the workpiece, and the opening type stress intensity factor K is only in the vicinity of the minute region. ₁ is set to a value larger than the fracture toughness value K _1c of the material, and an offset is set in the temperature distribution of the cleaving position determining factor as necessary so that the cleaving position matches the target position, or A brittle material thermal stress cleaving apparatus that performs negative feedback control and, in some cases, positive feedback control, uses a line beam obtained by converting a laser beam by a diffraction grating type optical element for a cleaving position determining factor. 6. The method according to claim 5, wherein the direction or length of the line beam is controlled in accordance with the shape of the cleaving contour at the cleaving position. 6. The laser according to claim 5, wherein one of a CO ₂ laser, a Fe ²⁺ : ZnSe laser, and an Er: YAG laser is used as a laser according to a required cleaving type such as surface scribe, very deep scribe, and full cut. 6. The apparatus according to claim 5, wherein an electrostatic chucking plate capable of switching voltage application independently for each place for fixing the workpiece.

Images