Method for in-situ measurement and regulation of wafer warping change in liquid phase epitaxy process

Abstract

The invention discloses a method for in-situ measurement and regulation of wafer warping change in a liquid phase epitaxy process, and the method comprises the steps: starting a wafer warping in-situ measurement unit at a stage of liquid phase epitaxy growth of a rare earth iron garnet single crystal film, and measuring the warping degree of a wafer in a furnace body in real time; dynamically adjusting the rotation speed of the wafer according to the position change of the light spot on the wafer warping in-situ measurement unit until the growth is finished; the dynamic adjustment comprises a high-speed stage change and a low-speed stage change. According to the technical scheme, in-situ monitoring of the crystal growth process is achieved, the warping condition of crystal growth can be adjusted by adjusting the rotating speed, and fluctuation of the wafer cutting yield caused by the difference between a substrate and equipment is effectively improved. According to the warping change of the wafer, the rotation speed is continuously adjusted in the growth process, so that the warping is controlled within a certain range, the problem of wafer cracking is reduced, the product yield is improved, and the growth of a large-size Faraday optical rotation sheet wafer is facilitated.

Description

Technical Field

[0001] The present invention relates to the field of artificial crystal growth equipment, and in particular to a method for in-situ measuring and regulating wafer warpage changes in a liquid phase epitaxial process. Background Art

[0002] Faraday rotators, that is, single crystal materials of rare earth iron garnet, play an important role in optical communications, especially in the construction of data centers. An optical isolator made with Faraday rotators as the core uses the Faraday effect to change the polarization state of light through a magnetic field. This is used in optical communication systems to control signal polarization and ensure the accuracy and effectiveness of data transmission. With the rapid development of big data computing, artificial intelligence, and 5G communications in recent years, the demand for Faraday rotators has also skyrocketed.

[0003] According to the characteristics of rare earth iron garnet single crystal materials, it is mainly obtained by the crystal growth method of liquid phase epitaxy. However, in the process of growing rare earth iron garnet wafers by liquid phase epitaxy, due to the difference in lattice constant and thermal expansion coefficient between the wafer and the substrate, the wafer is prone to different degrees of warping and bending at different stages of crystal growth, resulting in cracking, dislocation and other problems. For example, publication number CN105702564A uses stress film and ion implantation process to make targeted adjustments to the warpage of the wafer in a specific direction. Publication number CN107170696A solves the problem of wafer warpage by placing a ring heating wire at the warping position.

[0004] The existing methods for controlling the warpage often require waiting until the crystal growth is completed before observation can be made, which hinders the further improvement of the yield of liquid phase epitaxial growth of rare earth iron garnet single crystal materials. Therefore, it is necessary to find a method to observe the warpage of the wafer in real time as the crystal grows, and to control the growth parameters in situ according to the warpage to control the warpage of the wafer within a certain range.

[0005] The existing technology does not fully consider the influence of wafer warpage on the rare earth iron garnet film, and cannot directly observe it during the growth process, which reduces the yield rate of rare earth iron garnet single crystal material. Summary of the invention

[0006] In view of the above-mentioned deficiencies in the prior art, the present invention constructs a quantifiable method to in-situ monitor the changes of the wafer during the growth process, achieves in-situ regulation, improves the yield rate, and enriches the means of regulating the composition of rare earth iron garnet.

[0007] To achieve the above object, the technical solution adopted by the present invention is as follows: a method for in-situ measuring and regulating the change of wafer warpage during liquid phase epitaxy comprises the following steps:

[0008] Step 1, preparing a melt; placing the required raw materials in a crucible, heating the crucible to completely melt the raw materials, and keeping the crucible warm for at least 12 hours to homogenize the raw materials;

[0009] Step 2, liquid phase epitaxial growth of rare earth iron garnet single crystal film; specifically, slowly cooling the melt to the growth temperature at a certain speed, for example, 843°C ± 10°C; slowly inserting the tooling with the substrate into the furnace body and contacting it with the liquid surface of the melt; turning on the wafer warpage in-situ measurement unit to measure the warpage of the wafer in the furnace body; dynamically adjusting the rotation speed of the wafer until the growth is completed according to the position change of the light spot on the wafer warpage in-situ measurement unit; the wafer warpage in-situ measurement unit includes a laser transmitter, a light reflector, and a light receiver; the dynamic adjustment includes two stages of fast and slow changes; the first stage is a fast speed regulation stage.

[0010] Furthermore, the rotational acceleration of the rapid speed regulation is 3-5 rpm / h 2 .

[0011] Furthermore, the rotational acceleration of the slow speed regulation is 0.4 to 2 rpm / h 2 .

[0012] Preferably, the rotation acceleration of the slow speed regulation is 0.6 to 1.2 rpm / h 2 .

[0013] Furthermore, the implementation time of the slow speed regulation is greater than 1.0 rpm / h when the change value of the position of the light spot from the previous moment is greater than 0.1 cm. 2 When the change is less than 0.1cm, the speed is adjusted to less than 1.0rpm / h. 2 The acceleration adjusts the speed.

[0014] Furthermore, the implementation time of the first stage is the first 4 hours of crystal growth period.

[0015] Furthermore, before and after the wafer growth, the position change value of the light spot is controlled within -0.6 cm to +0.1 cm. Preferably, the position change value of the light spot is controlled within -0.3 cm to +0.1 cm.

[0016] Preferably, the initial rotation speed of the substrate is 70 rpm.

[0017] Furthermore, the substrate has a size of 3 to 4 inches.

[0018] Beneficial effects: In view of the problem that the wafer is prone to warping and cracking, which reduces the yield rate, the present invention proposes a technical solution to monitor the warping of the wafer in situ and scientifically regulate the rotation speed during the growth process, so as to control the warping within a certain range, reduce the problem of wafer cracking, and improve the yield rate of the product. The controllability of the growth process is enhanced, which greatly helps the composition design of rare earth iron garnet single crystals and helps to grow large-size Faraday rotator wafers.

[0019] The present invention finds that the change value of the spot position X before and after growth should be controlled between -0.6cm and +0.1cm, at which time the warping degree of the wafer can be suppressed within 35%, and the cutting rate exceeds 90%. Preferably, the change value of X should be controlled between -0.3cm and +0.1cm. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] Figure 1 It is a schematic diagram of the wafer warpage status and a schematic diagram of the relationship between the receiver display spot displacement and the wafer warpage degree;

[0021] Figure 2 It is the measurement and control result of wafer warpage of Comparative Example 1;

[0022] Figure 3 The measurement and control results of the wafer warpage in Example 1;

[0023] Figure 4 The measurement and control results of wafer warpage in Example 2;

[0024] Figure 5 These are the measurement and control results of wafer warpage in Comparative Example 2.

[0025] Explanation of reference numerals: 1. first reflector; 2. second reflector; 3. rotating rod; 4. furnace body; 5. substrate; 6. receiver; 7. rare earth iron garnet film. DETAILED DESCRIPTION

[0026] In order to better understand the technical solution of the present invention, the embodiments of the present invention are described in detail below with reference to the accompanying drawings.

[0027] It should be clear that the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0028] like Figure 1As shown, a laser is used as a light source to illuminate the wafer surface. The light passes through the first reflector 1 and irradiates the substrate 5, and then reflects to the second reflector 2, and finally is displayed on the receiver 6. The position of the light spot is recorded as X1. During the crystal growth process, a rare earth iron garnet film 7 will be deposited below the substrate 2, or on both sides (the unilateral growth state shown in the figure). Due to the difference in thermal expansion coefficient and lattice constant between the substrate 2 and the rare earth iron garnet film 7, the wafer surface is partially warped. The degree and direction of the warping are affected by the lattice structure of the rare earth iron garnet single crystal. The change in wafer warping has a crucial impact on the quality of the crystal.

[0029] The laser is reflected in a direction different from the original direction on the warped wafer surface, which eventually causes the position X of the light spot received on the light screen of the receiver 6 to change. Figure 1 In a warping state shown in B, the edge of the wafer is higher than its center and bulges toward the melt side, so that a gap is generated between the edge of the wafer and the horizontal plane. At this time, the position of the light spot received by the receiver 6 moves to the right, which is represented by X2 in the figure. Figure 1 C is another warping state, in which the center of the wafer is higher than its edge and bulges toward the non-melt side. At this time, the light spot received by the receiver 6 moves in the opposite direction, which is represented by X3 in the figure. The change in the X value displayed by the receiver 6 can provide feedback on the change in the warping of the wafer. Furthermore, the value of X can reflect the degree of wafer warping, and the rate of change of the X value can provide feedback on the rate of change of wafer warping, thereby playing a guiding role in improving wafer warping and designing the lattice structure of rare earth iron garnet single crystals. It should be noted that the warping of the wafer includes curvature, which can be measured by the radius of curvature. When the reflection angle of the laser and the optical path from the wafer surface to the reflector are constant, the distance the light spot on the receiver 6 moves can reflect the distance the wafer is warped away from the initial horizontal plane.

[0030] At the same time, the present inventors also found in their research work that during the epitaxial growth of rare earth iron garnet, if the wafer rotates faster, the wafer tends to bulge toward the non-melt side, that is, Figure 1 C; on the contrary, the wafer tends to bulge toward the melt, that is, Figure 1 B. Therefore, by adjusting the rotation speed during the epitaxy process, the warpage of the wafer can be changed from convex to the non-melt side to convex to the melt side, or from convex to the melt side to convex to the non-melt side, so as to achieve dynamic adjustment and improvement of the wafer warpage during the crystal growth process. Crucially, when the warpage exceeds a certain value, the wafer will crack, affecting the yield rate. Therefore, monitoring and controlling the warpage of the wafer during the crystal growth process is the core of obtaining high-quality crystal materials. Figure 2 The position changes of the light spot on the receiver 6 at different times during the crystal growth process of Comparative Example 1 were recorded, and the leftmost side of the light screen of the receiver 6 was taken as the starting end, which was recorded as zero scale. Figure 2 The first X value data recorded in is not zero, which is mainly caused by the adjustment of the optical path. In fact, the initial position can also be regarded as zero. Then the X value approaches the starting end in a certain undulating manner, indicating that the wafer may gradually bulge toward the non-melt side. The wafer is taken out of the furnace body 4 and observed and compared. The direction and degree of warping reflected by the light spot are consistent, that is, the wafer bulges toward the non-melt side.

[0031] During the crystal growth process, the rotating rod 3 drives the substrate 2 to rotate. The change in the rotation speed will affect the temperature field distribution on the wafer surface, and then affect the deposition rate of the rare earth iron garnet film 7 and the growth direction of the crystal. If the rotation speed is too fast or too slow, it may cause uneven material deposition and stress concentration, which will cause wafer warping. In addition, the growth rate of the crystal is approximately proportional to the square root of the rotation speed. The appropriate rotation speed plays a positive role in improving stress distribution and suppressing wafer warping, thereby controlling the overall warping within a reasonable range.

[0032] The present invention is described below by way of examples and comparative examples, but the scope thereof is not limited thereto.

[0033] Comparative Example 1: The required raw materials are placed in a platinum crucible, and the powder raw materials are completely melted by heating the crucible to about 1200°C, and the raw materials are homogenized by keeping the temperature for at least 12 hours. The melt is slowly cooled to the growth temperature, about 843°C, at a rate of 180°C / min. Subsequently, the tooling containing the SGGG substrate material is slowly introduced into the furnace body and contacts the liquid surface of the melt. The substrate size used is 4 inches.

[0034] Adjust the direction of the laser (not shown in the figure) so that the laser beam is reflected according to the optical path through the first reflector 1 outside the furnace body 4 and through the second reflector 2 to the receiver 6 screen display and read as the initial position. . Start the function of the epitaxial equipment rotation, so that the substrate 2 rotates forward and reverse at a speed of 70 revolutions per minute and a period of 20 seconds. As the crystal grows, observe the moving direction and distance of the light spot. Record the position of the light spot on the receiver 6 every 36 minutes for a total of 24 hours to complete the crystal growth. After the growth is completed, slowly cool the wafer to room temperature at a rate of 0.4℃ / min and then take it out. The recording interval can be adjusted as needed, or it can be recorded continuously without interruption.

[0035] The position change of the light spot is as follows Figure 2 As shown in the figure, the initial position is 1.9 cm from the starting end, at which time the wafer has already produced a certain degree of warpage. Within 24 hours, the value X of the spot position shows an overall decreasing trend, indicating that the center of the wafer is higher than the edge and bulges toward the non-melt side. 1 / 2 Represents the square root of the rotational speed, and its value is 8.367.

[0036] The value of X changed by -1.1cm before and after growth (the negative sign indicates that the light spot moves toward the starting end of the receiver 6 light screen, and the positive sign indicates that the light spot moves away from the starting end), which means that the warping of the wafer has increased, which has changed by about 57.9% compared with the initial growth. The wafer taken out was observed, and about 6 curved cracks were observed between the substrate and the growing crystal, and the number of cracks exceeding 2cm was 4. The wafer was cut, and a total of 31 crack-free Faraday rotators with a size of 11mm*11mm were obtained. The cutting yield was 70.5%.

[0037] Example 1, the same operating steps as Comparative Example 1, the difference is that the growth process is controlled by speed control. The specific process is as follows: the required raw materials are placed in a platinum crucible, the powder raw materials are completely melted by heating the crucible to about 1200°C, and the raw materials are homogenized by keeping the temperature for at least 12 hours. The melt is slowly cooled to the growth temperature, about 843°C, at a rate of 180°C / min. Subsequently, the tooling containing the SGGG substrate material is slowly introduced into the furnace body and contacts the liquid surface of the melt. The substrate size used is 4 inches.

[0038] Adjust the direction of the laser (not shown in the figure) so that the laser beam is reflected along the optical path to the receiver 6 screen display and read as the initial position. Start the function of the epitaxial device rotation, and make the substrate 2 rotate at a speed of 70 revolutions per minute (ω 1 / 2 =8.367), with a period of 20 seconds of forward and reverse switching rotation. Record the initial spot position of substrate 2, whose X value is 1.9 cm. As the crystal grows, observe the moving direction and distance of the spot, and record the position of the spot on receiver 6 every 36 minutes. The crystal growth is completed after 24 hours in total. After the growth is completed, the wafer is slowly cooled to room temperature at a rate of 0.4℃ / min and then taken out.

[0039] The specific operation of dynamically adjusting the rotation speed of the rotating rod 3 in situ during the growth process: in the first stage of crystal growth, more specifically within 4 hours, at 3 to 5 rpm / h 2 Then, according to the change of wafer warpage, the acceleration is 0.4~2rpm / h. 2 Slowly reduce the speed, preferably the speed reduction acceleration is 0.6-1.2rpm / h 2 The change in wafer rotation speed and warpage is as follows: Figure 3 shown.

[0040] Example 1 Figure 3 As shown, with a cycle of 36 minutes, the first stage is 4 rpm / h 2 The acceleration was reduced to 62 rpm / min for two cycles. After the light spot position shifted, it returned to the initial position. Then the rotation speed was increased to 1.0 rpm / h. 2Slowly reduce the speed for two cycles to 60rpm / min and hold for a while, then increase the speed to 2.0rpm / h 2 Slowly reduce the speed to 56rpm / min for a while, then increase it to 0.6rpm / h 2 After reducing the speed to 51.2rpm / min, that is, ω in the figure 1 / 2 =7.197, maintain the current speed for the next 12 hours until the crystal growth is completed. When the change in the position of the light spot from the previous moment is greater than 0.1cm, use a speed greater than 1.0rpm / h 2 If the speed is slightly larger, the speed will be reduced slowly. When the speed is less than 0.1cm, the speed will be less than 1.0rpm / h. 2 Slow down the speed a little bit.

[0041] It can be observed that within the last 12 hours of crystal growth, the position of the light spot keeps moving slightly toward the starting end of the receiver 6. Before and after the final crystal growth, the change in the light spot position is -0.6 cm, which is about 31.6% more than the initial growth stage, much lower than that of Comparative Example 1. The wafer taken out was observed, and there was a total of 1 straight crack between the substrate 2 and the growing crystal, with a length of less than 2 cm. The wafer was cut to obtain a total of 41 crack-free Faraday rotator plates with a size of 11 mm*11 mm. The cutting yield was 93.2% (calculated with 44 pieces as the upper limit of the yield).

[0042] Example 2: Place the required raw materials in a platinum crucible, heat the crucible to about 1200°C to completely melt the powder raw materials, and keep the temperature for at least 12 hours to homogenize the raw materials. Slowly cool the melt to the growth temperature, about 843°C, at a rate of 180°C / min. Then, slowly place the tooling containing the SGGG substrate material into the furnace and contact the liquid surface of the melt. The substrate size used is 4 inches.

[0043] The same steps as in Example 1, adjust the laser to make a light spot appear on the screen of the receiver 6 and read it as the initial position. Start the function of the epitaxial device rotation, and make the substrate 2 rotate at a speed of 70 rpm and a period of 20 forward and reverse switching. The initial position of the light spot is 1.7 cm. As the crystal grows, observe the moving direction and distance of the light spot. Record the position information of the light spot on the screen of the receiver 6 every 36 minutes, and complete the crystal growth after a total of 24 hours. During the process, the rotation speed change and warping of the wafer are as follows Figure 4 shown.

[0044] Specific operations for dynamically controlling the rotation speed of the rotating rod 3 in situ during the growth process: Figure 4 As shown, with a cycle of 36 minutes, the first stage is 5 rpm / h 2 Reduce the speed once with an acceleration of 2.0rpm / h, observe the position shift of the light spot, and then 2Slowly reduce the speed to 62rpm / min. After a while, it is observed that the position of the light spot is shifting in the opposite direction. At this time, the speed is increased to 63rpm / min. Similarly, the light spot decelerates when the receiver 6 moves to the right and accelerates when it moves to the left. When the change in the position of the light spot from the previous moment does not exceed 0.1cm, increase the speed to 0.8rpm / h. 2 Adjust the speed. When the change of the light spot position from the previous moment is greater than 0.1cm, adjust the speed to 1-3rpm / h 2 Adjust the rotation speed. When the spot position is offset, continue to adjust the rotation speed until the crystal growth is completed. After the growth is completed, slowly cool the wafer to room temperature at a rate of 0.4℃ / min and then take it out.

[0045] The X change value before and after growth is +0.1cm, which is about 5.9% higher than that in the initial growth stage, much lower than that in Comparative Example 1 and Example 1. The wafer taken out was observed, and no cracks were observed between the substrate 2 and the growing crystal. The wafer was cut to obtain a total of 44 crack-free Faraday rotators with a size of 11mm*11mm. The cutting yield was 100.0%.

[0046] Comparative Example 2 has the same operating steps as Comparative Example 1, except that the rotation speed is slowly reduced with a constant acceleration during the growth process. Figure 5 As shown, the acceleration was reduced from 70 rpm / min to 37 rpm / min every 36 min (the crystal growth process lasted for about 15 hours at this time), and then the rotation speed was maintained until the end of the growth.

[0047] The average distance between the light spot on the receiving screen and the initial position on the screen of receiver 6 was recorded every 36 minutes, and the crystal growth was completed after 24 hours in total. After the growth was completed, the wafer was slowly cooled to room temperature at a rate of 0.4℃ / min and then taken out.

[0048] The X change value before and after growth is +0.8cm, which is about 50% different from the initial growth. The wafer taken out was observed, and 6 cracks were observed between the substrate 2 and the growing crystal, of which 3 were less than 2cm in length. The wafer was cut to obtain a total of 33 crack-free Faraday rotators with a size of 11mm*11mm. The cutting yield was 75.0%.

[0049] On the one hand, the substrate material used to grow the Faraday rotator has certain differences in properties such as thermodynamics and lattice constants, which causes certain changes in the wafer warpage during the growth process; on the other hand, there are certain fluctuations in the consistency between equipment furnaces, which affects the cutting yield of the Faraday rotator wafer. Through the technical solution proposed in the present invention, in-situ monitoring of the crystal growth process is achieved, and the warpage of the crystal growth can be adjusted by adjusting the rotation speed. Through the above technology, the fluctuation of the wafer cutting yield caused by the difference between the substrate 2 and the equipment is effectively improved. According to the change of the wafer warpage, the rotation speed is continuously adjusted during the growth process, so that the warpage is controlled within a certain range, and the wafer yield is improved.

[0050] The present invention found that the change value of the spot position X before and after growth should be controlled between -0.6cm and +0.1cm, at which time the wafer has a higher cutting rate. Preferably, the change value X should be controlled between -0.3cm and +0.1cm.

[0051] Due to different melt designs, there are differences in the setting of growth parameters, which makes the growth process difficult to monitor and modulate. Through the technical solution proposed by the present invention, in-situ monitoring of the crystal growth process is achieved, and it is also found that the warping of the crystal growth can be adjusted by adjusting the rotation speed. By reducing or increasing the rotation speed as the crystal grows, the warping is controlled within a certain range, effectively improving the yield rate of the wafer.

[0052] The above description is only a preferred embodiment of the present invention and does not limit the technical scope of the present invention. Therefore, any slight modification, equivalent changes and modifications made to the above embodiments based on the technical essence of the present invention still fall within the protection scope of the present invention.

Claims

1. A method for in-situ measurement and control of wafer warpage change during liquid phase epitaxy, characterized in that: The following steps are involved: Step 1, preparing a melt; placing the required raw materials in a crucible, heating the crucible to completely melt the raw materials, and keeping the crucible warm for at least 12 hours to homogenize the raw materials; Step 2, liquid phase epitaxial growth of rare earth iron garnet single crystal thin film; specifically, slowly cooling the melt to the growth temperature at a certain speed; slowly inserting the tooling with the substrate into the furnace body and contacting it with the liquid surface of the melt; turning on the wafer warpage in-situ measurement unit to measure the warpage of the wafer in the furnace body; dynamically adjusting the rotation speed of the wafer until the growth is completed according to the position change of the light spot on the wafer warpage in-situ measurement unit; the wafer warpage in-situ measurement unit includes a laser transmitter, a light reflector, and a light receiver; The dynamic adjustment includes two stages of changes: fast and slow; the first stage is the fast speed regulation stage.

2. The method for in-situ measuring and controlling wafer warpage during liquid phase epitaxy according to claim 1, characterized in that: The rotational acceleration of the rapid speed regulation is 3-5 rpm / h 2 .

3. The method for in-situ measuring and controlling wafer warpage during liquid phase epitaxy according to claim 1, characterized in that: The rotation acceleration of the slow speed regulation is 0.4-2rpm / h 2 .

4. The method for in-situ measuring and controlling wafer warpage change during liquid phase epitaxy according to claim 3, characterized in that: The rotation acceleration of the slow speed regulation is 0.6 to 1.2 rpm / h 2 .

5. The method for in-situ measuring and controlling wafer warpage change during liquid phase epitaxy according to claim 3, characterized in that: The implementation time of the slow speed regulation is greater than 1.0 rpm / h when the change value of the position of the light spot from the previous moment is greater than 0.1 cm 2 When the change is less than 0.1cm, the speed is adjusted to less than 1.0rpm / h. 2 The acceleration adjusts the speed.

6. The method for in-situ measuring and controlling wafer warpage during liquid phase epitaxy according to claim 1, characterized in that: The first stage is implemented during the first 4 hours of crystal growth.

7. The method for in-situ measuring and controlling wafer warpage during liquid phase epitaxy according to claim 1, characterized in that: Before and after the wafer growth, the position change value of the light spot is controlled within -0.6cm to +0.1cm.

8. The method for in-situ measuring and controlling wafer warpage during liquid phase epitaxy according to claim 7, characterized in that: Before and after the wafer growth, the position change value of the light spot is controlled within -0.3cm to +0.1cm.

9. The method for in-situ measuring and controlling wafer warpage during liquid phase epitaxy according to claim 1, characterized in that: The initial rotation speed of the substrate was 70 rpm.

Images