Method for calibrating low-temperature thermal field of epitaxial machine

Abstract

The invention provides a method for calibrating a low-temperature thermal field of an epitaxy machine, and is suitable for the technical field of reduced-pressure Ge and GeSi epitaxy. The method comprises the following steps: firstly, measuring the thickness of a Ge epitaxial layer at each position of a wafer in a static state through an ellipsometer, adjusting the offset value of each temperature zone, and ensuring that the temperature distribution difference of each temperature zone is controlled within 15%; and then, the offset is finely adjusted in the rotating state of the base, so that the thickness uniformity of the Ge epitaxial layer is within 2.5%. And finally, the TC compensation values of the center and edge temperature zones are adjusted, so that the thickness mean value of the Ge epitaxial layer reaches the range of + / -5% of the thickness of the standard sheet, and accurate calibration of the low-temperature thermal field is realized. The method does not need additional consumables or temperature monitoring instruments, is compatible with an existing epitaxial process, is rapid and accurate, remarkably improves the accuracy of low-temperature field control, and optimizes the thickness uniformity and lattice quality of an epitaxial layer.

Description

Technical Field

[0001] The invention belongs to the technical field of reduced-pressure Ge and GeSi epitaxy, and relates to a method for calibrating a low-temperature temperature field of an epitaxial machine. Background Art

[0002] In the current reduced pressure epitaxial growth process, the accuracy and uniformity of the temperature field will have a significant impact on the epitaxial growth of pure Ge and GeSi, and directly affect the thickness uniformity and lattice quality of the corresponding epitaxial layer.

[0003] At present, there are two main technical routes for temperature control of single-wafer reduced pressure epitaxial furnace: 1. The reaction chamber of the ASM Epsilon 2000 series epitaxial equipment often uses a multi-temperature zone temperature control method to control the temperature field, that is, multiple thermocouples are used to measure the temperature at different points in the reaction chamber (the specific structure is as follows Figure 1 As shown in the figure), PID is used for feedback control to achieve different temperatures set in the recipe, and temperature difference control between different temperature ranges can be achieved. However, due to the differences in the reaction chamber structure and the Tc measurement points, in order to achieve the accuracy and uniformity of the wafer surface temperature, it is often necessary to set an offset value for the temperature of the Tc measurement point, and it is necessary to set the offset value of other temperature zones relative to the center temperature. The existing technology achieves the uniformity of the temperature field on the wafer surface by setting temperature offset values ​​and offset values ​​for each temperature zone and monitoring the different temperatures of each temperature zone.

[0004] 2. AMAT C5200 and NAURA E320 series machines usually use Pyro temperature control mode, which controls the wafer surface temperature by PID controlling the temperature measured by the upper / lower Center Pyro and the Up / Lw Inner Ratio, so that the temperature and uniformity during the process reach the target level.

[0005] The main temperature control measurement methods currently used in mass production include ion implantation wafer surface resistance measurement method, TC Wafer test method, etc.

[0006] In the prior art, in the low-temperature epitaxial temperature range (340℃~400℃), the temperatures of each temperature zone in the five-zone TC temperature control system will affect each other, and an independent PID control system is adopted for each temperature zone. There is compensation and interference in the power between the temperature zones, and the TC is located under the graphite base and at a certain distance from the wafer surface, which makes the actual temperature and energy obtained in each area of ​​the wafer surface differ greatly from the actual measurement of the TC.

[0007] Since the Ge epitaxial layer is extremely sensitive to temperature in the low temperature range, and within this temperature range, the ion implantation temperature control sheet has low temperature sensitivity or even fails, and accurate temperature monitoring cannot be completed; TC Wafer is expensive, and a cavity needs to be opened when measuring temperature data. Under reduced pressure, this temperature measurement method seriously affects the airtightness of the machine, and temperature monitoring during epitaxial layer growth cannot be performed. After the test, the flange needs to be reinstalled and leak detection performed. The process is cumbersome, affects machine time, and poses a major safety hazard. Summary of the invention

[0008] The purpose of the present invention is to provide a method for calibrating the low-temperature temperature field of an epitaxial machine, which solves the problems of inaccurate temperature control and large difference between the actual temperature of the wafer surface and the measured temperature in a five-zone TC temperature control system.

[0009] The technical solution adopted by the present invention is: A method for calibrating a low temperature field of an epitaxial machine comprises the following steps: S1: Select the Ge epitaxial recipe that needs to be calibrated, set the base not to rotate, and fix the MonitorWafer Notch orientation position for a test run; S2: The thickness of the pure Ge epitaxial layer at the corresponding positions of Center / Front / Side / Rear on the wafer is measured by an ellipsometer, and then the corresponding temperature of each position of the wafer in the static state is determined, and the temperature distribution of each temperature zone in the cavity can be obtained; S3: Adjust the offset of the corresponding temperature zones of Front / Side / Rear and conduct multiple experiments until the thickness difference between the temperature zones meets the following conditions: (Max Thk-Min Thk) / (Max Thk+Min Thk) ≤ 15%; S4: Use the Ge epitaxial recipe at the same temperature, set the susceptor to rotate, and continue the test run; S5: Measure the thickness of the Ge epitaxial layer under the rotating state of the base by using an ellipsometer, and calculate the thickness mean and uniformity, compare the test thickness mean with the Ge epitaxial thickness at the corresponding standard temperature, and calculate the corresponding deviation; S6: Fine-tune the offset values ​​of the corresponding temperature zones of Front / Side / Rear, and conduct multiple experiments until the thickness uniformity of the Ge epitaxial layer is ≤2.5%, completing the temperature field distribution uniformity debugging; S7: According to the difference between the Ge Thk average and the standard chip, adjust the TC compensation value of the Center / Front / Side1 / Side2 / Rear temperature zone until the Ge thickness average reaches the range of ±5% of the standard chip thickness, and the absolute temperature calibration of the cavity is completed.

[0010] In S1, the Monitor Wafer Notch is oriented so that the Notch faces the rear flange.

[0011] The specific steps of S1 are as follows: (1) Heating the reaction chamber of the epitaxial furnace to remove the reactants remaining in the chamber, thereby ensuring the cleanliness of the internal environment of the chamber; (2) Cool the reaction chamber temperature and load the B-Blank wafer into the reaction chamber; (3) Raise the temperature of the reaction chamber, set the offset value of each temperature zone, reduce the chamber pressure, remove the native oxide layer on the wafer surface, and the base speed is 0 r / min; (4) Lower the chamber temperature to the calibration temperature and introduce GeH4 to grow the Ge epitaxial layer.

[0012] The specific steps of S2 are as follows: Use an ellipsometer to measure the thickness of the Ge epitaxial layer on the wafer surface and calculate the uniformity of the Ge thickness. The uniformity calculation method is (Max Thk-Min Thk) / (Max Thk+Min Thk).

[0013] The specific steps of S4 are as follows: (1) Heating the reaction chamber of the epitaxial furnace to remove the reactants remaining in the chamber, thereby ensuring the cleanliness of the internal environment of the chamber; (2) Cool the reaction chamber temperature and load the B-Blank wafer into the reaction chamber; (3) Raise the temperature of the reaction chamber, set the offset value of each temperature zone, the offset value is determined according to the S3 result, reduce the chamber pressure, remove the native oxide layer on the wafer surface, and rotate the base; (4) Lower the chamber temperature to the calibration temperature and introduce GeH4 to grow the Ge epitaxial layer.

[0014] The specific steps of S6 are as follows: (1) Heating the reaction chamber of the epitaxial furnace to remove the reactants remaining in the chamber, thereby ensuring the cleanliness of the internal environment of the chamber; (2) Cool the reaction chamber temperature and load the B-Blank wafer into the reaction chamber; (3) Raise the temperature of the reaction chamber and set the offset value of each temperature zone. The offset value is determined according to the result of S5. Reduce the chamber pressure, remove the native oxide layer on the wafer surface, and rotate the base at the same speed as S4. (4) Lower the chamber temperature to the calibration temperature and introduce GeH4 to grow the Ge epitaxial layer.

[0015] The beneficial effects of the present invention are: The calibration method disclosed in the present invention can rely on the most basic epitaxial equipment and consumables to calibrate the low-temperature temperature field required by the process, and is fully compatible with the existing epitaxial process flow, without the need to increase the cost of additional consumables or temperature monitoring instruments (only epitaxial machines and monitor wafers are required). The method is accurate, fast and convenient, and only requires observing the thickness and uniformity of the Ge epitaxial layer deposited on the wafer to obtain the actual temperature in the current cavity and the temperature distribution of each temperature zone.

[0016] This calibration method is basically applicable to epitaxial machines of any technical route, and its main effects are as follows: 1. The average temperature in the cavity during the low-temperature process is detected based on the temperature-sensitive characteristics of the Ge epitaxial layer thickness at 340~400℃; 2. The absolute temperature range between each temperature zone of the cavity temperature field is determined based on the different thickness distributions of the Ge epitaxial layer in the stationary wafer; 3. The temperature distribution during the cavity process can be simulated based on the uniformity and mean value of the thickness of the Ge epitaxial layer in the rotating wafer. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 It is 350℃ mapping.

[0018] Figure 2 It is the mapping after 350℃ chamber calibration. DETAILED DESCRIPTION

[0019] The present invention is described in detail below with regard to specific implementation modes.

[0020] A method for calibrating the low temperature field of an epitaxial machine, using a common monitoring sheet in production as a material, and calculating the accurate actual temperature by monitoring the GR of Ge, specifically comprising the following steps: S1: Select the Ge epitaxial recipe that needs to be calibrated, set the base not to rotate, and fix the MonitorWafer Notch direction (usually Notch toward the rear flange direction) for a test run; S2: The thickness of the pure Ge epitaxial layer at the corresponding positions of Center / Front / Side / Rear on the wafer is measured by an ellipsometer, and then the corresponding temperature of each position of the wafer in the static state is determined, and the temperature distribution of each temperature zone in the cavity can be obtained; S3: Adjust the offset of the corresponding temperature zones of Front / Side / Rear and conduct multiple experiments until the thickness difference between the temperature zones meets the following conditions: (Max Thk-Min Thk) / (Max Thk+Min Thk) ≤ 15%; S4: Use the Ge epitaxial recipe at the same temperature, set the susceptor to rotate, and continue the test run; S5: Measure the thickness of the Ge epitaxial layer under the rotating state of the base by using an ellipsometer, and calculate the thickness mean and uniformity, compare the test thickness mean with the Ge epitaxial thickness at the corresponding standard temperature, and calculate the corresponding deviation; S6: Fine-tune the offset values ​​of the corresponding temperature zones of Front / Side / Rear until the thickness uniformity of 49 points of the Ge epitaxial layer is ≤2.5% ((Max Thk-Min Thk) / (Max Thk+Min Thk) ≤2.5%), completing the temperature field distribution uniformity debugging; S7: According to the difference between the Ge Thk average and the standard chip, adjust the TC compensation value of the Center / Front / Side1 / Side2 / Rear temperature zone until the Ge thickness average reaches the range of ±5% of the standard chip thickness, and the absolute temperature calibration of the cavity is completed.

[0021] According to the Ge epitaxial layer thickness of xxxA grown at 350℃ for 2700s on the BSL machine (the Ge thickness Monitor data after the BSL machine is calibrated to 350℃ for TC Wafer), the subsequent machine A is calibrated by the method described in the first part. At present, the Ge Thk is xxxA under the same conditions, and the 49P Full Map Uni is 1.86%. The Mapping matches the BSL machine, and it is believed that the temperature of the A machine and the BSL cavity are consistent. (At 300~400℃, the Ge epitaxial layer GR is extremely sensitive to temperature, showing an exponential relationship) Example

[0022] The specific steps of S1-S3 are as follows: (1) The temperature of the epitaxial furnace reaction chamber was raised to 1050°C, and the reaction chamber and graphite components were etched and baked for 150 seconds using high-flow 20 SLM HCl and 10 SLM H2 to remove the residual reactants in the chamber, thereby ensuring the cleanliness of the internal environment of the chamber; (2) Cool the reaction chamber temperature to 750 °C and load the B-Blank wafer into the reaction chamber; (3) Raise the temperature from 750 °C to 1050 °C, set the offset value of each temperature zone (front, back, side) to 11 / 4 / 30, reduce the chamber pressure to 60 Torr, bake for 180s in a 20 SLM H2 atmosphere to remove the native oxide layer on the wafer surface, and the base speed is 0 r / min; (4) Lower the chamber temperature to 350°C and introduce 100 sccm GeH4 for 2700 s to grow the Ge epitaxial layer; (5) The thickness of the Ge epitaxial layer on the wafer surface was measured using an ellipsometer, and the uniformity of the Ge thickness was determined to be 11.38%. The temperature difference between each temperature zone met the requirements.

[0023] Steps S4-S5 are as follows: (1) Based on the experimental results of implementation case 1, the debugging offset is tentatively set as 11 / 4 / 30; (2) The temperature of the epitaxial furnace reaction chamber was raised to 1050°C, and the reaction chamber and related base components were etched and cleaned with high-flow 20 SLM HCl and 10 SLM H2 for 90 seconds. The H2 flow rate was then increased to 80 SLM for 60 seconds to further remove the residual reactants in the chamber, thereby ensuring the cleanliness of the internal environment of the chamber. (3) Cool the reaction chamber temperature to 750 °C and load the B-Blank wafer into the reaction chamber; (4) Raise the temperature from 750 °C to 1050 °C, set the offset value of each temperature zone (front, back, side) to 11 / 4 / 30, reduce the chamber pressure to 60 Torr, bake for 180s in a 20 SLM H2 atmosphere to remove the native oxide layer on the wafer surface, and adjust the base speed to 35 r / min; (5) Lower the chamber temperature to 350°C and introduce 100 sccm GeH4 for 2700 s to grow the Ge epitaxial layer; (6) The thickness of the Ge epitaxial layer on the wafer surface was measured using an ellipsometer. The average thickness was 699.75A and the uniformity was 5.83%. Both the uniformity and the average did not meet the requirements; (the standard thickness at 350°C is 600±30A and the uniformity requirement is ≤2.5%) The steps of S6 are as follows: (1) Based on the experimental results of implementation case 2, the offset is adjusted to 13 / 6 / 32; (2) The temperature of the epitaxial furnace reaction chamber was raised to 1050°C, and the reaction chamber and related base components were etched and cleaned with high-flow 20 SLM HCl and 10 SLM H2 for 90 seconds. The H2 flow rate was then increased to 80 SLM for 60 seconds to further remove the residual reactants in the chamber, thereby ensuring the cleanliness of the internal environment of the chamber. (3) Cool the reaction chamber temperature to 750 °C and load the B-Blank wafer into the reaction chamber; (4) Raise the temperature from 750 °C to 1050 °C, set the offset value of each temperature zone (front, back, side) to 11 / 4 / 30, reduce the chamber pressure to 60 Torr, bake for 180 s in a 20 SLM H2 atmosphere to remove the native oxide layer on the wafer surface, and the base speed is 35 r / min; (5) Lower the chamber temperature to 350°C and introduce 100 sccm GeH4 for 2700 s to grow the Ge epitaxial layer; (6) The thickness of the Ge epitaxial layer on the wafer surface was measured using an ellipsometer, and the average thickness was determined to be 744.22A, and the uniformity was 2.34%. The average was higher than the standard conditions, but the uniformity met the requirements; S7 specific steps are as follows: (1) According to the experimental results of Implementation Case 3, the TC compensation temperature of the machine was adjusted to reduce the overall temperature of the cavity. The five-zone compensation of C / F / S1 / S2 / R was changed from -10 / -10 / -10 / -10 / -10 to -16 / -17 / -17 / -17 / -17; (2) The temperature of the epitaxial furnace reaction chamber was raised to 1050°C, and the reaction chamber and related base components were etched and cleaned with high-flow 20 SLM HCl and 10 SLM H2 for 90 seconds. The H2 flow rate was then increased to 80 SLM for 60 seconds to further remove the residual reactants in the chamber, thereby ensuring the cleanliness of the internal environment of the chamber. (3) Cool the reaction chamber temperature to 750 °C and load the B-Blank wafer into the reaction chamber; (4) Raise the temperature from 750 °C to 1050 °C, set the offset value of each temperature zone (front, back, side) to 13 / 6 / 32, reduce the chamber pressure to 60 Torr, bake for 180s in a 20 SLM H2 atmosphere to remove the native oxide layer on the wafer surface, and the base speed is 35 r / min; (5) Lower the chamber temperature to 350°C and introduce 100 sccm GeH4 for 2700 s to grow the Ge epitaxial layer; (6) The thickness of the Ge epitaxial layer on the wafer surface was measured using an ellipsometer, and the average thickness was determined to be 588.89A and the 49P uniformity was 1.86%, which met the standard temperature field requirements at 350°C. The cavity temperature field calibration was completed.

[0024] The final 350℃ cavity calibration is mapped as follows Figure 2 As shown, compared Figure 1 From the 350℃ standard wafer mapping, we can see that the average thickness of the calibrated wafer Ge epitaxial layer is 568.98A, and the uniformity is 2.31% (matching the standard wafer average of 574.1A and uniformity of 1.85%). The average temperature in the wafer is determined to be xx, and the temperature difference in the wafer is <xx, which fully meets the process temperature accuracy requirements.

Claims

1. A method for calibrating the low temperature field of an epitaxial machine, characterized in that: The specific steps include: S1: Select the Ge epitaxial recipe that needs to be calibrated, set the base not to rotate, and fix the Monitor WaferNotch orientation position for a test run; S2: The thickness of the pure Ge epitaxial layer at the corresponding positions of Center / Front / Side / Rear on the wafer is measured by an ellipsometer, and then the corresponding temperature of each position of the wafer in the static state is determined, and the temperature distribution of each temperature zone in the cavity can be obtained; S3: Adjust the offset of the corresponding temperature zones of Front / Side / Rear and conduct multiple experiments until the thickness difference between the temperature zones meets the following conditions: (Max Thk-Min Thk) / (Max Thk+Min Thk) ≤ 15%; S4: Use the Ge epitaxial recipe at the same temperature, set the susceptor to rotate, and continue the test run; S5: Measure the thickness of the Ge epitaxial layer under the rotating state of the base by using an ellipsometer, and calculate the thickness mean and uniformity, compare the test thickness mean with the Ge epitaxial thickness at the corresponding standard temperature, and calculate the corresponding deviation; S6: Fine-tune the offset values ​​of the corresponding temperature zones of Front / Side / Rear, and conduct multiple experiments until the thickness uniformity of the Ge epitaxial layer is ≤2.5%, completing the temperature field distribution uniformity debugging; S7: According to the difference between the Ge Thk average and the standard chip, adjust the TC compensation value of the Center / Front / Side1 / Side2 / Rear temperature zone until the Ge thickness average reaches the range of ±5% of the standard chip thickness, and the absolute temperature calibration of the cavity is completed.

2. A method for calibrating the low temperature field of an epitaxial machine according to claim 1, characterized in that: In S1, the Monitor Wafer Notch is oriented so that the Notch faces the rear flange.

3. A method for calibrating the low temperature field of an epitaxial machine according to claim 2, characterized in that: The specific steps of S1 are as follows: (1) Heating the reaction chamber of the epitaxial furnace to remove the reactants remaining in the chamber, thereby ensuring the cleanliness of the internal environment of the chamber; (2) Cool the reaction chamber temperature and load the B-Blank wafer into the reaction chamber; (3) Raise the temperature of the reaction chamber, set the offset value of each temperature zone, reduce the chamber pressure, remove the native oxide layer on the wafer surface, and the base speed is 0 r / min; (4) Lower the chamber temperature to the calibration temperature and introduce GeH4 to grow the Ge epitaxial layer.

4. The method for calibrating the low temperature field of an epitaxial machine according to claim 1, characterized in that: The specific steps of S2 are as follows: Use an ellipsometer to measure the thickness of the Ge epitaxial layer on the wafer surface and calculate the uniformity of the Ge thickness. The uniformity calculation method is (Max Thk-Min Thk) / (Max Thk+Min Thk).

5. The method for calibrating the low temperature field of an epitaxial machine according to claim 1, characterized in that: The specific steps of S4 are as follows: (1) Heating the reaction chamber of the epitaxial furnace to remove the reactants remaining in the chamber, thereby ensuring the cleanliness of the internal environment of the chamber; (2) Cool the reaction chamber temperature and load the B-Blank wafer into the reaction chamber; (3) Raise the temperature of the reaction chamber, set the offset value of each temperature zone, the offset value is determined according to the S3 result, reduce the chamber pressure, remove the native oxide layer on the wafer surface, and rotate the base; (4) Lower the chamber temperature to the calibration temperature and introduce GeH4 to grow the Ge epitaxial layer.

6. The method for calibrating the low temperature field of an epitaxial machine according to claim 1, characterized in that: S6 specific steps are as follows: (1) Heating the reaction chamber of the epitaxial furnace to remove the reactants remaining in the chamber, thereby ensuring the cleanliness of the internal environment of the chamber; (2) Cool the reaction chamber temperature and load the B-Blank wafer into the reaction chamber; (3) Raise the temperature of the reaction chamber and set the offset value of each temperature zone. The offset value is determined according to the result of S5. Reduce the chamber pressure, remove the native oxide layer on the wafer surface, and rotate the base at the same speed as S4. (4) Lower the chamber temperature to the calibration temperature and introduce GeH4 to grow the Ge epitaxial layer.

Images