JPH03134187A - Method and device for etching - Google Patents

Abstract

PURPOSE:To control a wafer temp. with high accuracy and to stably improve working characteristics in a low-temp. etching method by changing an etching atmosphere which is a heat generating source according to the wafer temp. in addition to the temp. control by a wafer cooler. CONSTITUTION:For example, a microwave plasma etching device is used and the wafer 14 is fixed onto a cooling electrode 9 connected to a machine 8 for supplying liquid nitrogen as a refrigerant. A cooling gas is packed on the rear surface thereof. An etching gas is introduced from an introducing system 6 while the wafer is cooled and the etching is started by impressing microwave electric power. The incident heat from the etching atmosphere and the temp. rise of the wafer by reaction heat are measured by a measuring instrument 10 at this time and the set value of etching conditions are so changed by a control system 11 as not to deviate from the desired wafer temp. control range in accordance with the measured temp. The temp. control of the high degree in excess of the cooling capacity is executed by this method.

Description

[Detailed description of the invention] [Industrial application field]

The present invention relates to a low temperature dry etching method. Particularly suppresses wafer temperature rise during etching. The present invention relates to an etching method and apparatus suitable for highly precisely temperature-controlled etching in an ultra-low temperature region.

[Conventional technology]

In dry etching technology that uses reactive gas plasma to form fine patterns on semiconductor integrated circuits, a low-temperature etching method that cools the wafer to be processed to below 0°C to process it with high precision was disclosed in Japanese Patent Application Laid-Open No. 158627/1984. has been done. In this method, the wafer is cooled by placing it on an electrode having a coolant reservoir, and the side etching reaction is suppressed by lowering the temperature of the reaction surface. (Side etching is mainly caused by a thermal reaction caused by neutral radicals,
Low temperature conditions with little thermal excitation are suppressed. On the other hand, the reaction at the etched bottom surface proceeds even if the wafer temperature becomes considerably low due to the irradiation energy of ions accelerated toward the bottom surface. ) Temperature control is important in such low-temperature etching methods, but during etching, the wafer temperature tends to fluctuate due to radiant heat from the plasma and etching reaction heat. Regarding this point, Japanese Patent Application Laid-Open No. 56-73 shows that if plasma is generated intermittently during room-temperature etching, this temperature rise can be reduced.
439.

Problem 1 to be Solved by the Invention Based on the principle, in the low temperature etching method, the processing accuracy improves as the wafer temperature is controlled to a lower temperature. However, the cooling temperature has a lower limit determined by the coagulation temperature of the reaction gas. If the wafer is cooled below the temperature at which agglomeration occurs, etching may be inhibited or prevented from progressing due to adhesion of reactants to the wafer surface. Therefore, temperature control for achieving the highest precision processing in low-temperature etching is to control the wafer surface to the lowest possible temperature within a range where etching is not inhibited by gas aggregation. However, in the above-mentioned prior art, a problem arises in that temperature control becomes insufficient when etching requires a large amount of reaction heat, such as etching of Si with F, or when plasma irradiation heat is large. In addition, with a method that simply performs etching intermittently, etching may stop due to excessive cooling, or
There has been a problem in that machining accuracy may deteriorate due to insufficient suppression of the upper limit temperature. [Means for Solving the Problems] The above object is achieved by controlling the temperature by a wafer cooling device and also by changing the etching atmosphere, which serves as two heat generation sources, in accordance with changes in the process temperature. The etching atmosphere was changed at this time. It must not impair the processing characteristics determined by the main etching atmosphere (for example, processing shape, etching mask material selectivity). Therefore, when the temperature tends to rise in the main etching atmosphere, there is little heat generation effect such as a low etching atmosphere, a trace deposition atmosphere, or an etching pause, and there is also little effect on the etching shape (or the etching shape is not as desired). The temperature is controlled using the atmosphere.

[Effect]

Changes in the etching atmosphere are immediately reflected in the wafer surface temperature, making it more sensitive than conventional temperature control based on changes in cooling capacity from the back side of the wafer. In other words, precise temperature control is possible. In addition, when the temperature tends to rise due to insufficient cooling capacity, it is possible to temporarily reduce the amount of heat generated and suppress the rise. The temperature control range can be very low, close to the temperature of the refrigerant itself. That is, when using the same cooling mechanism, instead of using it in an equilibrium state between the cooling capacity and the amount of heat generated as in the conventional case, it is possible to precisely control the upper and lower limits by determining the upper and lower limits within the range of the non-equilibrium state during temperature rise. [Example 1 Example 1] One embodiment of the present invention will be described below with reference to FIG. FIG. 1 is a cross-sectional view of the main parts of the etching apparatus according to the ninth embodiment. This device consists of a magnetron 1. Waveguide 2. Discharge tube 3. Electromagnetic coil 4. High frequency power application power source 5. Etching gas introduction system6. A conventional microwave plasma etching apparatus having an exhaust system 7 as a main component, a cooling electrode connected to a coolant supply device 8, a wafer temperature measuring device 10, and a control system 11 that changes etching condition settings based on the wafer temperature. It has been established. Here, liquid nitrogen was used as the coolant, so that the inside of the cooling electrode 9 could be cooled to about -196°C, and the surface temperature of the aluminum wafer mounting surface, which has good thermal conductivity, could be cooled to about -170°C. The wafer to be etched is placed on this cooling electrode and fixed with a presser foot 12 made of ceramic such as silicon nitride, which has poor thermal conductivity. The back surface of the wafer is filled with a gas such as He as a cooling gas at a pressure of about 1 kPa. Wafer temperature measurement is performed by temperature measurement 11 using thermocouple 13 as a measuring element.
10. Although not shown in the diagram. From the preliminary evacuation chamber, the wafer 14 is placed on the cooling electrode by an automatic transfer system without breaking the vacuum in the etching chamber. After installing the wafer, the wafer temperature will rise to -130℃ for about 10 to 20 seconds.
Cooled to ~-150°C. In this state, for example, sulfur hexafluoride (hereinafter referred to as SF) as an etching gas is introduced at a pressure of about IPa, and microwave power and high frequency power are also applied as necessary to remove silicon (hereinafter referred to as Si). ), the wafer temperature changes as shown in FIG. Namely. After the start of cooling, the wafer temperature is asymptotically cooled to become equal to the cooling electrode surface temperature. Etching is then started when the temperature has cooled to a desired temperature. If the amount of heat generated in the wafer during etching is small compared to the cooling capacity, the wafer temperature may drop even after etching has started, so reduce the cooling capacity (for example, by lowering the cooling gas pressure) to bring the wafer to the desired temperature. Keep it. On the other hand, if the amount of heat generated is large, the wafer temperature will rise after etching starts, resulting in
It is saturated at a temperature T that satisfies the following equation. (T To) · h−s =Q Here, To is the surface temperature of the cooling electrode, h is the thermal resistance between the wafer and the electrode, S is the contact area, that is, the wafer area, and Q is the amount of heat generated. To and h are parameters expressing cooling capacity, and the lower To is and the higher h is, the lower the wafer temperature T can be kept against a certain amount of heat generation Q, but there is a limit determined by tTo is the refrigerant and h is the cooling gas pressure etc. . Generally, in low-temperature etching, processing accuracy improves as the wafer temperature is controlled as low as possible within a range that does not cause agglomeration of etching gas, etc. Therefore, it is important to devise ways to control the wafer temperature to a desired low temperature with high precision under a practical cooling mechanism and etching conditions. In the etching apparatus of the present invention shown in FIG. 1, the wafer temperature during etching can be measured and various etching conditions can be changed based on the measured temperature. That is, by specifying the temperature range to be managed and the etching conditions to be changed in the control system 11, effective etching can proceed only within the desired temperature range.
Advanced temperature control that exceeds cooling capacity is now possible. Although a microwave etching apparatus is taken as an example here, other etching apparatuses such as reactive ion etching (abbreviated as RIE) may be equipped with a similar temperature control mechanism. Further, the same control can be performed using liquid helium, which is lower in temperature than liquid nitrogen, or antifreeze, which is higher in temperature but simpler, as the refrigerant. Wafer temperature measurement. In addition to thermocouples, use fluorescent thermometers, resistance thermometers, etc. It can be estimated from direct contact with the wafer or from indirect measurements of the return gas temperature of the circulated cooling gas. Example 2 An example of etching processing will be described with reference to FIG. This figure shows changes in wafer temperature and accompanying etching switching when etching a Si substrate to a desired depth using a sample in which approximately half of the surface area of a 4-inch diameter Si wafer is covered with a resist pattern. Figure 3 shows this. The setting values of the etching control factors are as follows. First, the magnetron power, which is the microwave discharge power, is 5
00W, SF was introduced as an etching gas at 50 cc/min, and the gas pressure in the etching chamber was adjusted to IPa by adjusting the exhaust speed with a conductance valve. A high frequency power of 5 W was applied by a 2 MHz power source to provide an accelerating electric field for etching ions. In the case of Si etching using SFG, it is known that the wafer temperature is preferably in the range of about 25 DEG C. to -135 DEG C. (for example, JP-A-64-32628). According to the above etching conditions, the etching speed of Si is 3 to 4
It is μm/min. The heat of reaction of S i +4 F-+s i F is about 45
Since it was as large as 0 kcal/mol, it was estimated that the reaction heat generated in the wafer during such high-speed etching was equivalent to about 40 W. In addition, there are several tens of W of incident heat from the plasma, and due to this heat generation, the temperature reaches -135℃.
When etching was started, the temperature rose to -125°C after 6 seconds. Thereafter, the discharge for etching was stopped for 2 seconds, and the temperature was again cooled to -135°C. As shown in FIG. 2, in this way, the upper limit temperature was set to -125°C and the lower limit temperature was set to -135°C, and the discharge power was automatically controlled to be 0N10FF to the sea where these temperatures were detected. In this case, temperature control was achieved by repeating etching for about 6 seconds and stopping for 2 seconds. When the etching was completed by setting the total etching time to 1 minute, a highly accurate Si groove with vertical side walls was obtained, as shown in FIG. 6, and the etching depth was 4 μm in the wide groove portion 23. The etching amount of the resist is 0.2 .mu.m, and the selection ratio of Si to reregistration is 1 to 20, which is a large value that is good for normal selective etching. Since a value about 50 times larger than that of SiO2 can be obtained, it is preferable to use Sin as the etching mask material when performing deeper etching. In addition, as shown in FIG. It was observed that the etching depth of the submicron-sized fine grooves 24 tended to be shallower than that of the wide grooves. This point will be described in detail in Example 4 later. If continuous etching is performed under the same discharge conditions as the main etching without temperature control, the wafer temperature will be -90°C.
It took 1 minute to reach a certain level, and the processed shape became one with slight side etching as shown in Figure 5. In general, when the wafer temperature cannot be controlled and reaches -100° C. or higher, side etching tends to be noticeable and shape reproducibility tends to deteriorate. Furthermore, if etching is simply carried out intermittently, cooling time is added to the effect of slowing down the temperature rise, but the longer the etching is carried out, the more unstable the processed shape becomes due to temperature fluctuations. This seems to be because the temperature of the cooling electrode surface and the etching chamber wall surface rises to some extent, so the wafer temperature tends to rise as the etching time elapses. Embodiment 3 As a method of not directly measuring the wafer temperature during etching, it is also possible to calculate the etching and cooling times based on the amount of heat generated and the cooling characteristics, and control the etching based on this. The generated heat ff1Q is the sum of the incident heat Q1 of ions and light from the plasma and the chemical reaction heat Q2 on the etching surface,
The Q process can be determined in advance from the discharge conditions, and the Q2 can be determined in advance from the area to be etched that is treated in a predetermined etching atmosphere. As for the cooling properties. The wafer temperature drop characteristics with respect to time after the wafer is placed on the cooling electrode, and the wafer temperature rise under constant heat generation Q (i.e., under constant discharge conditions) - temperature rise) is measured in advance and input into the control system. Based on such control data, the etching switching time is determined by specifying the etching conditions, etching material, area, and control temperature range. When processed according to this control, temperature control was achieved in the same manner as in Example 2, and high-precision processing was achieved. Note that when using this method, it is important to prevent the temperature of the entire etching chamber from rising and the cooling capacity from changing. For this reason, the walls of the etching chamber are kept at a constant temperature (Lo
It is desirable to provide a heater system that heats the electrode to a constant temperature (approximately 100 ft), a cooling system that constantly fills the cooling mold with refrigerant to keep the surface temperature of the electric heron constant, and minimizes the thermal resistance from the refrigerant to the electrode surface. . Embodiment 4 During the time to recover from the temperature increase during etching, it is possible to not only stop etching on the body but also use various discharge conditions that generate less heat. Figure 4 is an example of this, and 1
Five conditions (a) to (e) are switched in a cycle. (a) After the wafer is cooled to a predetermined temperature of −135° C., CF, gas IPa, and microwave power ioow are discharged for 2 seconds. Under these conditions, etching of Si does not proceed and a small amount of deposit containing C occurs. This was a condition with little heat generation, and the wafer temperature remained almost constant at -135°C. (b) Then, the discharge is paused for 2 seconds to switch the gas from CF4 to He. (c) He gas IPa, R for microwave 200W discharge
F was applied at 10 W and etching was carried out for about 2 seconds until the etching temperature reached the lower limit of -140°C. This does not remove Si etching, but only removes the deposited film deposited on the etched surface. ((1) Next, while switching the gas to SF6, IPa,
Si etching with microwave 500W. The upper limit temperature becomes -130°C in 5 seconds. (e) Next, the discharge is stopped for 2 seconds and the gas exchange period begins. This results in two-condition switching temperature-controlled etching with a cycle time t of about 13 seconds. When etching is performed for a long time, the temperature rises quickly and the etching time tends to become gradually shorter (d'), but the etching depth can be controlled by the total etching time. Furthermore, in order to perform etching atmosphere switching in such a short time, the apparatus is configured to have a high gas exhaust speed and a high gas pressure adjustment speed. The purpose of setting the deposition conditions in the etching cycle is to increase the etching anisotropy due to the sidewall protection effect of the deposited film remaining on the etched sidewalls, and to increase the etching anisotropy by increasing the amount of deposition, as shown in FIG. This is because it has an effect that can be controlled. In addition to this, the effect of correcting the etching delay of the fine grooves shown in FIG. 6 as shown in FIG. 7 was also obtained. This seems to be because when the deposit material adheres, the amount of deposit inside the microgrooves is small, so the bottoms of the microgrooves are kept relatively clean. If etching is performed using two different conditions while controlling the temperature as described above, the etching characteristics can be controlled more favorably. Although we talked about Si etching here,
It goes without saying that this etching method can be used in combinations of polycrystalline silicon, tungsten, tungsten silicide, the fffi film of the multilayer resist 1-, and other various materials and various etching gases therefor. Effects of the Invention 1 According to Akira, the temperature of the wafer during etching can be highly controlled, so that the processing characteristics of low-temperature etching, which is sensitive to the wafer temperature, can be stably improved.

[Brief explanation of the drawing]

FIG. 1 is a sectional view of the etching apparatus of the present invention, and FIG. 2 is a diagram showing changes in etching wafer temperature. 3 and 4 are diagrams showing changes in wafer temperature during etching according to an embodiment of the present invention, FIG. 5 is a sectional view of a processed portion of a wafer showing an example of a conventional etching shape, and FIGS. 6 and 7 1 is a sectional view of a processed portion of a wafer according to an embodiment of the present invention. Explanation of symbols 1... Magnetron, 2... Waveguide, 3... Discharge tube. 5... High frequency power supply, 8... Refrigerant supply machine, 9... Cooling type tM. 10... Wafer temperature measuring device, 11... Control system. 14...Wafer, 21...Si, 22...Resist No. 1 Fig./-77' Netron 2- Doji Suga 3- Discharge tube 4-11 Electric fiber coil 5-11 High-use rinsing power supply g- 'x Twins η Su, - E people thread 2---Mu1 Bohide 411 Sheet Turn 9ヱrtth> lθ・-Raeha warmth; I'l regular job If ---s'I profit l - m - To Tsuney Fig. Eckenbu section order ↓ Etsuchinfu Cape 3 Le 4 Easy (minutes) 14 hours time - Fig. 5 Fig. 6 Fig. 7 Fig. 21-4; !

Claims (6)

[Claims] 1. In a low-temperature etching method in which the wafer is exposed to an etching atmosphere while being cooled, the wafer temperature rise due to incident heat and reaction heat from the etching atmosphere is measured, and the atmosphere is changed so as not to deviate from the desired wafer temperature control range. Etching method. 2. The above atmosphere to be changed is an etching gas atmosphere,
The temperature of the wafer placed on the cooling electrode can be changed in accordance with the cycle by inserting a rest period in which the wafer is left in either a depot gas atmosphere and a non-reactive atmosphere as necessary. An etching method characterized in that the temperature is controlled so as to move back and forth between a desired upper and lower temperature limit. 3. In the etching method described in claim 1, etching is started when the wafer has been cooled to the lower limit of the temperature control range, and when the upper limit is reached, the etching atmosphere is shut off, and the etching is stopped after the wafer has been cooled to the lower limit again. An etching method characterized by repeating a cycle of starting and continuing until a desired total etching time is reached. 4. An etching apparatus comprising means for measuring wafer temperature and means for changing etching conditions based on the measured value. 5. The etching apparatus is characterized in that the etching apparatus is a plasma etching apparatus, and the etching conditions to be changed are any of the type of reaction gas, flow rate, etching chamber pressure, and discharge power. 6. An etching method characterized by estimating the amount of heat generated during processing in a predetermined etching atmosphere from the area to be etched, predicting temperature rise, and setting etching conditions and time to be changed in advance.