An upper electrode plasma device, a method of using the same, and a method of manufacturing the same

Abstract

This invention provides an upper electrode plasma device, a method for using the upper electrode plasma device, and a method for manufacturing the device, relating to the field of plasma reactor technology. The device includes: a composite plasma generating unit, an insulating plate, a spray hole frame, and an upper electrode plate. An insulating ring and a conductive induction ring are nested axially to form a composite nested unit. The insulating cylinder has a first end and a second end that are axially opposite each other. The composite nested unit is disposed inside the first end of the insulating cylinder, and the inner circumferential wall of the first end of the insulating cylinder covers the outer circumference of the composite nested unit. A coil is arranged around the outer circumference of the second end of the insulating cylinder. The conductive induction ring generates a ring-shaped electric field, and the coil generates an induced coupling electric field. A capacitive coupling electric field is formed between the insulating ring and the conductive induction ring, generating capacitively coupled plasma and inductively coupled plasma within the gas channel. This achieves reliable ignition under low pressure and flexible switching between multiple modes.

Description

Technical Field

[0001] This invention relates to the field of plasma reactor technology, specifically to an upper electrode plasma device, a method of using the upper electrode plasma device, and a method of manufacturing the device. Background Technology

[0002] In the manufacturing process of semiconductor devices, processes such as plasma etching, physical vapor deposition, and chemical vapor deposition are commonly used to micro-machin the wafers. Plasma-assisted processes are generally carried out in a vacuum reaction chamber, among which plasma etching is a key step in processing the wafers into the designed patterns.

[0003] Under specific conditions, a gas can undergo glow discharge under the influence of a strong electric field. During this process, the electrons of the gas molecules are excited and jump to higher energy levels. Since these electrons are not stable in the high energy state, they will fall back to their original low energy orbits on their own, releasing photons corresponding to the energy difference according to the law of conservation of energy.

[0004] Besides the common solid, liquid, and gaseous states, plasma is another state of matter existing in the universe. When a gas is subjected to high temperatures, collisions between molecules cause outer electrons to escape their orbits, forming a plasma composed of charged particles. This mixture of charged ions is conductive.

[0005] Currently, most plasma generation structures used in equipment are based on a single principle: capacitively coupled plasma (CCP) sources or inductively coupled plasma (ICP) sources. CCP sources have a simple structure and controllable ion energy, but their plasma density is relatively low; ICP sources can generate high-density plasma, but their ion energy is not easily controlled independently, and ignition is difficult under low pressure. There is currently no device that combines these two principles in the same ring structure, nor is there a corresponding method for coordinated control. Summary of the Invention

[0006] In view of this, embodiments of this specification provide an upper electrode plasma device, a method for using the upper electrode plasma device, and a method for manufacturing the device. This achieves the goal of simultaneously generating high-density plasma and independently controlling ion energy within the same device, while also enabling reliable ignition under low pressure and flexible switching between multiple modes.

[0007] The embodiments in this specification provide the following technical solutions:

[0008] An upper electrode plasma device, comprising: Composite plasma generating unit, insulating plate, spray hole frame and upper electrode plate; The composite plasma generating unit includes an insulating ring, a conductive induction ring, an insulating cylinder, and a coil; An insulating ring and a conductive induction ring are nested along the axial direction to form a composite nested unit. The insulating cylinder has a first end and a second end that are axially opposite. The composite nested unit is disposed inside the first end of the insulating cylinder, and the inner circumferential wall of the first end of the insulating cylinder covers the outer circumference of the composite nested unit. The coil is arranged around the outer circumference of the second end of the insulating cylinder. A gas channel is formed inside the insulating cylinder, extending from the first end to the second end; The conductive induction ring is connected to the first radio frequency power supply and is used to generate a ring electric field. The coil is connected to the second radio frequency power supply and is used to generate an induced coupling electric field. A capacitive coupling electric field is formed between the insulating ring and the conductive induction ring, and capacitively coupled plasma and inductively coupled plasma are generated in the gas channel. The composite plasma generating unit is mounted on the upper electrode plate, and the spray hole frame is fixed below the upper electrode plate and supports the lower part of the composite plasma generating unit. An insulating plate is placed between the upper electrode plate and the spray hole frame to isolate the potential between the upper electrode plate and the spray hole frame.

[0009] A method of using an upper electrode plasma device, which generates composite plasma based on the upper electrode plasma device and is used for substrate processing, includes the following steps: Process gas is introduced into the gas channel inside the insulating cylinder; Select the working mode according to the process requirements. The working modes include low density mode, sensing mode, high density mode and cleaning mode. When the operating mode is low-density mode, the first radio frequency power is applied to the conductive induction loop through the first radio frequency power supply, and the second radio frequency power supply is turned off to generate pure capacitively coupled plasma. When the operating mode is induction mode, a second radio frequency power is applied to the coil through the second radio frequency power supply, and the first radio frequency power supply is turned off to generate pure inductively coupled plasma. When the operating mode is high-density mode, a first radio frequency power is applied to the conductive induction ring through the first radio frequency power supply, and a second radio frequency power is applied to the coil through the second radio frequency power supply at the same time, and the power amplitude of the second radio frequency power is greater than the power amplitude of the first radio frequency power, so as to generate high-density composite plasma. When the working mode is cleaning mode, the first radio frequency power supply is turned off, and the second radio frequency power is applied to the coil only through the second radio frequency power supply. Oxygen-containing cleaning gas is introduced into the gas channel to generate pure inductively coupled oxygen plasma to remove the deposits on the inner wall of the gas channel. The plasma flow is distributed to the surface of the substrate to be treated.

[0010] Furthermore, when the operating mode is low-density mode, a first radio frequency power is applied to the conductive induction loop through the first radio frequency power supply, and the second radio frequency power supply is turned off to generate purely capacitively coupled plasma, including: When applying the first radio frequency power to the conductive induction loop, the first radio frequency power is applied in a pulse manner; The current gas pressure value in the gas channel is detected in real time. The duty cycle of each pulse cycle is calculated independently based on the current gas pressure value. When the gas pressure value is higher than the predetermined high gas pressure threshold, the duty cycle is reduced until the predetermined low duty cycle is reached. When the gas pressure value is lower than the predetermined low gas pressure threshold, the duty cycle is increased until the predetermined high duty cycle is reached, so that the duty cycle can be continuously or stepwise adjusted within the predetermined duty cycle range.

[0011] Furthermore, when the operating mode is induction mode, before applying the second radio frequency power to the coil via the second radio frequency power supply, the following steps are also included: Detect the current gas pressure in the gas channel. If the current gas pressure is lower than the predetermined ignition gas pressure threshold, execute the auxiliary ignition procedure, including: Temporarily turn on the first radio frequency power supply, and apply a set of auxiliary radio frequency pulses to the conductive induction loop in the form of a pulse train. The width of each auxiliary radio frequency pulse is within a predetermined pulse width range, the pulse interval is within a predetermined pulse interval range, the number of pulses in the pulse train is within a predetermined pulse number range, and the power amplitude of each pulse increases sequentially from the first to the last, with the increase slope being within a predetermined increase slope range. During the application of the auxiliary radio frequency pulse train, the initial electron flux density in the gas channel is detected in real time by a Langmuir probe set on the inner wall of the insulating cylinder. After each auxiliary radio frequency pulse ends, the detected initial electron flow density is compared with a predetermined ignition density threshold. If the current density value reaches or exceeds the predetermined ignition density threshold, the remaining auxiliary radio frequency pulses are immediately terminated, the first radio frequency power supply is turned off, and the second radio frequency power supply is turned on within a predetermined switching delay time range, so that the coil starts to output the second radio frequency power. If the current density value is lower than the predetermined ignition density threshold, the next auxiliary radio frequency pulse is applied.

[0012] Furthermore, when the operating mode is high-density mode, a first radio frequency power is applied to the conductive induction loop through a first radio frequency power supply, and a second radio frequency power is simultaneously applied to the coil through a second radio frequency power supply, with the power amplitude of the second radio frequency power being greater than the power amplitude of the first radio frequency power, thereby generating high-density recombination plasma, including: Before simultaneously applying the first radio frequency power and the second radio frequency power, the initial electron flow density value of the inner wall of the insulating cylinder is obtained, and the initial electron flow density value is compared with the predetermined target density value to calculate the density deviation. The initial amplitude ratio of the first RF power and the second RF power is dynamically set according to the density deviation, so that the amplitude of the second RF power is greater than the amplitude of the first RF power, and the initial amplitude ratio is positively correlated with the density deviation. During the simultaneous application of the first radio frequency power and the second radio frequency power, the output amplitude of the first radio frequency power supply and the output amplitude of the second radio frequency power supply are adjusted respectively to independently change the actual amplitude of the first radio frequency power and the second radio frequency power. The amplitude of the first radio frequency power is adjusted to control the ion energy bombarding the substrate surface, and the amplitude of the second radio frequency power is adjusted to control the plasma density in the gas channel. The emission spectrum intensity or ion current density of the composite plasma in the gas channel is detected in real time. The detected value is compared with the process target value. Based on the comparison result, the amplitude ratio of the first radio frequency power and the second radio frequency power is repeatedly adjusted to stabilize the characteristic parameters of the composite plasma within the predetermined error range.

[0013] Furthermore, it also includes: During the simultaneous application of the first and second radio frequency powers, the self-bias voltage on the substrate surface and the plasma density in the gas channel are measured. The self-bias voltage is compared with a predetermined target ion energy value: if the self-bias voltage is higher than a first positive deviation threshold of the target value, the current amplitude of the first radio frequency power is decreased by a first adjustment step; if the self-bias voltage is lower than a first negative deviation threshold of the target value, the current amplitude of the first radio frequency power is increased by a first adjustment step; otherwise, the first radio frequency power remains unchanged. The plasma density is compared with a predetermined density target value. If the plasma density is higher than the second positive deviation threshold of the target value, the current amplitude of the second radio frequency power is reduced by the second adjustment step. If the plasma density is lower than the second negative deviation threshold of the target value, the current amplitude of the second radio frequency power is increased by the second adjustment step. Otherwise, the second radio frequency power is kept unchanged. After each adjustment, wait for the predetermined stabilization time and repeat the above steps until the self-bias voltage and plasma density simultaneously meet their respective allowable deviation ranges.

[0014] Furthermore, it also includes: When switching from low-density mode to high-density mode, or from sensing mode to high-density mode, or from high-density mode to low-density mode or sensing mode, the following transition control steps are performed: While maintaining the output of the currently activated RF power supply, the output amplitude of the RF power supply to be activated is gradually increased in a ramp manner. The rise time of the ramp is set to a predetermined ramp rise time range, and the curve of the ramp is set to linear or exponential. When the output amplitude of the RF power supply to be turned on does not reach the first threshold ratio of the target amplitude, the output amplitude of the currently turned-on RF power supply remains unchanged. When the output amplitude of the RF power supply to be turned on enters the second threshold ratio range of the target amplitude, the output amplitudes of the two RF power supplies are adjusted synchronously so that the amplitude ratio of the two reaches the target ratio within the predetermined synchronous adjustment time range. If it is necessary to turn off the currently outputting RF power supply, after the output amplitude of the RF power supply to be turned on reaches the third threshold ratio of the target amplitude, the output amplitude of the RF power supply to be turned off will be reduced to zero in a ramp manner within a predetermined fall time range. During the switching process, the gas pressure value is monitored in real time and the process gas flow rate is adjusted to ensure that the fluctuation of the gas pressure value does not exceed the predetermined allowable range of gas pressure fluctuation of the set target value.

[0015] A method for manufacturing an upper electrode plasma device, the method comprising the following steps: An insulating ring and a conductive sensing ring are nested axially to form a composite nested unit, and the insulating ring and the conductive sensing ring are fixedly connected by high-temperature brazing or interference fit. The composite nested unit is inserted into the interior of the first end of the insulating cylinder, so that the inner peripheral wall of the first end covers the outer periphery of the composite nested unit and is fixed by ceramic adhesive or mechanical retaining ring. The coil is wrapped around the outer periphery of the second end of the insulating cylinder and fixed to the insulating cylinder by an insulating fastener; Install the insulating cylinder onto the upper electrode plate, so that the gas passage inside the insulating cylinder is connected to the air hole of the upper electrode plate. An insulating plate is placed between the upper electrode plate and the spray hole frame to isolate the potential between the upper electrode plate and the spray hole frame. Fix the spray hole frame below the upper electrode plate and support the insulating plate; The assembled upper electrode plasma device was subjected to overall sintering and performance testing.

[0016] Furthermore, the high-temperature brazing uses Ag-Cu-Ti active brazing filler metal, the brazing temperature is 850℃~950℃, and the vacuum degree is not less than 1×10⁻⁶. -3 Pa, after brazing, cool to room temperature at a rate of 2℃ / min to 5℃ / min; The radial interference of the interference fit is 0.02 mm to 0.10 mm, and the fitting force is 500 N to 2000 N. Before nesting the insulating ring and the conductive sensing ring, the conductive sensing ring is surface-plated with silver, with a silver plating layer thickness of 5 μm to 20 μm, and then annealed at 300℃ to 500℃ for 1 to 2 hours in an inert gas atmosphere. The ceramic adhesive is a phosphate-based or silicate-based ceramic adhesive with a temperature resistance of not less than 800℃, and the curing conditions are to keep it at 200℃~300℃ for 1~2 hours; The insulating fasteners are ceramic clips or mica gaskets. The coil is wound with hollow copper tubes with a diameter of 3 mm to 8 mm, 2 to 5 turns, and a helix angle of 5° to 15°. After winding, a polyimide insulating layer with a thickness of 0.1 mm to 0.5 mm is sprayed onto the surface of the coil, with a withstand voltage of not less than 5 kV.

[0017] Furthermore, the assembled upper electrode plasma device undergoes overall sintering and performance testing, including: The assembled upper electrode plasma device is placed in a sintering furnace and sintered at 1200℃~1400℃ for 1~3 hours in a hydrogen or argon atmosphere to form a metallurgical bond or densification at each fixed connection interface. After sintering, the material is cooled to room temperature at a rate of 2℃ / min to 5℃ / min, and then stress-reduced annealed at 200℃ to 300℃ for 2 hours. The following tests were performed on the upper electrode plasma device after overall sintering, and the qualification of the upper electrode plasma device was determined: Helium gas at a pressure of 0.2 MPa to 0.5 MPa is introduced into the gas channel, and the leak rate is detected using a helium mass spectrometer leak detector. The leak rate should not exceed 1 × 10⁻⁶ MPa. -7 Pa·m 3 When the plasma output is / s, the upper electrode plasma device is qualified; Apply 500 V DC between the insulating ring and the conductive induction ring, and measure the insulation resistance to be no less than 100 MΩ. Apply 500 V DC between the insulating plate and the upper electrode plate. When the measured insulation resistance is no less than 100 MΩ, the upper electrode plasma device is qualified. Connect the radio frequency source and test the coupling efficiency between the capacitively coupled plasma and the inductively coupled plasma generated by the upper electrode plasma device. When the coupling efficiency is not less than 85%, the upper electrode plasma device is qualified.

[0018] Compared with the prior art, the beneficial effects that at least one technical solution adopted in the embodiments of this specification can achieve include at least: By integrating the capacitively coupled insulating ring with the conductive induction ring and the inductively coupled coil onto the same insulating cylinder and sharing a gas channel, a physical composite of two plasma generation principles is achieved, providing a compact and synergistic structural basis for subsequent high-density, highly controllable plasma processes. Attached Figure Description

[0019] To more clearly illustrate the technical solutions of the embodiments of this application, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0020] Figure 1 This is an overall structural diagram of the upper electrode plasma device according to an embodiment of the present invention; Figure 2 This is a top view of the upper electrode plasma device according to an embodiment of the present invention; Figure 3 yes Figure 2 A sectional view of AA; Figure 4 This is an overall structural diagram of the composite plasma generating unit according to an embodiment of the present invention; Figure 5 This is a side view of the composite plasma generating unit according to an embodiment of the present invention; Figure 6 yes Figure 5 A sectional view of AA.

[0021] The attached diagram is labeled as follows: 1. Insulating ring; 2. Conductive induction ring; 3. Insulating cylinder; 4. Coil; 5. Insulating plate; 6. Spray hole frame; 7. Upper electrode plate. Detailed Implementation

[0022] The embodiments of this application will now be described in detail with reference to the accompanying drawings.

[0023] It should be noted that, unless otherwise specified, the embodiments and features described in this application can be combined with each other. The present invention will now be described in detail with reference to the accompanying drawings and embodiments.

[0024] I. Terminology Definitions.

[0025] In the embodiments of the present invention, unless otherwise stated, the following terms and preset parameters have the following meanings: Capacitively coupled plasma (CCP): Plasma generated by applying a radio frequency voltage between two electrodes, utilizing the capacitive coupling effect. Its characteristic is that the ion energy can be independently controlled by the radio frequency power, but the plasma density is relatively low (typically 10). 9 ~10 10 cm -3 ).

[0026] Inductively coupled plasma (ICP): Plasma excited by an induced electric field generated through electromagnetic induction by passing a high-frequency alternating current through a coil. Its characteristics include high plasma density (typically 10⁻⁶). 11 ~1012 cm -3 However, ion energy is not easily controlled independently, and ignition is difficult at low pressure (<1 Pa).

[0027] Controller: Includes a processor (such as MCU, DSP, FPGA), memory, and corresponding input / output interfaces. It is used to receive sensor signals, execute control algorithms, and output control commands.

[0028] Predetermined duty cycle range: preferably 5% to 95%, to ensure stable pulse CCP discharge without overheating.

[0029] Preset high pressure threshold: preferably set to 120% to 130% of the target pressure value. For example, when the target pressure is 2.0 Pa, the threshold is 2.5 Pa, which is used to determine that the pressure is too high and the duty cycle needs to be reduced.

[0030] Preset low pressure threshold: preferably set to 70% to 80% of the target pressure value. For example, when the target pressure is 2.0 Pa, the threshold is 1.5 Pa, which is used to determine that the pressure is too low and the duty cycle needs to be increased.

[0031] Predetermined low duty cycle: preferably 40% to 50%. When the air pressure is higher than the predetermined high air pressure threshold, the duty cycle is reduced to this range to reduce capacitive coupling energy injection.

[0032] Predetermined high duty cycle: preferably 60% to 70%. When the gas pressure is lower than the predetermined low gas pressure threshold, the duty cycle is expanded to this range to maintain plasma ignition stability.

[0033] Predetermined ignition gas pressure threshold: preferably 1.0 Pa. When the gas pressure in the gas channel is lower than this threshold, pure inductively coupled plasma ignition is difficult and an auxiliary ignition step needs to be performed.

[0034] Predetermined pulse width range: preferably 10 microseconds to 100 microseconds, as the duration of the auxiliary ignition pulse, to ensure sufficient energy injection without causing electrode overheating.

[0035] Predetermined pulse interval range: preferably 50 microseconds to 500 microseconds, the time interval between adjacent auxiliary ignition pulses, used for relaxation of plasma intermediate states.

[0036] The predetermined number of pulses is preferably 3 to 10, which is the number of pulses in a set of auxiliary pulse trains to achieve reliable ignition with the fewest possible pulses.

[0037] The predetermined incremental slope range is preferably 5% to 20%, which refers to the percentage increase in the power amplitude of each auxiliary pulse relative to the previous pulse, ensuring that the ignition energy gradually increases without overshoot.

[0038] Predetermined ignition density threshold: preferably 1×108 cm -3 When the initial electron flux density reaches this value, the inductively coupled plasma can independently sustain discharge.

[0039] The predetermined switching delay time range is preferably 0.1 milliseconds to 1 millisecond, which is the waiting time between turning off the first radio frequency power supply and turning on the second radio frequency power supply to avoid plasma extinction.

[0040] Preferred target density value: preferably 5 × 10 8 cm -3 Up to 1×10 9 cm -3 The initial electron flux density target value set according to process requirements is used to calculate density deviation and adjust power ratio.

[0041] Predetermined error range: preferably ±5%, which refers to the maximum relative deviation of composite plasma characteristic parameters (such as emission spectrum intensity) from the process target value.

[0042] First positive deviation threshold: preferably +5%, when the self-bias voltage measurement value is more than 5% higher than the predetermined ion energy target value, the first radio frequency power is reduced.

[0043] First negative deviation threshold: preferably -5%, when the self-bias voltage measurement value is more than 5% lower than the predetermined ion energy target value, the first radio frequency power is increased.

[0044] First adjustment step size: preferably 5% of the current first RF power amplitude, the proportion of power change during each adjustment, taking into account both response speed and stability.

[0045] The second positive deviation threshold is preferably +10%, which triggers a reduction in the second radio frequency power when the plasma density measurement value is more than 10% higher than the predetermined density target value.

[0046] The second negative deviation threshold is preferably -10%, which triggers an increase in the second radio frequency power when the plasma density measurement value is more than 10% lower than the predetermined density target value.

[0047] The second adjustment step size is preferably 8% of the current second RF power amplitude. This is the percentage change in power during each adjustment to ensure the smoothness of density adjustment.

[0048] Preset stabilization time: preferably 0.2 seconds. After each power adjustment, wait for the plasma response to reach a stable state before proceeding to the next round of measurement and adjustment.

[0049] The predetermined ramp rise time range is preferably 0.2 seconds to 2 seconds, which is the time for the RF power supply to rise from zero output to the target amplitude during mode switching. The ramp curve can be linear or exponential.

[0050] First threshold ratio: preferably 80%. During mode switching, the amplitude of the power supply to be turned on remains unchanged until the output amplitude reaches 80% of the target amplitude.

[0051] The second threshold ratio range is preferably 80% to 100%. When the output amplitude of the power supply to be turned on enters this range, the amplitude ratio of the two power supplies is adjusted synchronously.

[0052] The predetermined synchronous adjustment time range is preferably 0.5 seconds to 1 second, which is the time required to adjust the amplitude ratio of the two power supplies to the target ratio within the second threshold ratio range.

[0053] The third threshold ratio is preferably 95%. The RF power supply that needs to be turned off can only start to decrease after the output of another power supply reaches more than 95% of the target amplitude.

[0054] Predetermined fall time range: preferably 0.1 seconds to 0.5 seconds, the time it takes for the RF power supply to drop from the current output amplitude to zero.

[0055] Preset allowable range of gas pressure fluctuation: preferably ±10%. During the entire mode switching process, the gas pressure fluctuation in the gas channel is kept within the set target value range by adjusting the process gas flow rate to ensure plasma stability.

[0056] Allowable deviation range: refers to the range in which the self-bias voltage is allowed to deviate from the predetermined target value of ion energy, which is limited by the first positive deviation threshold and the first negative deviation threshold; and the range in which the plasma density is allowed to deviate from the predetermined target value of density, which is limited by the second positive deviation threshold and the second negative deviation threshold.

[0057] II. Upper Electrode Plasma Device.

[0058] like Figures 1 to 6 As shown, an upper electrode plasma device includes: a composite plasma generating unit, an insulating plate 5, a spray hole frame 6, and an upper electrode plate 7.

[0059] The composite plasma generating unit includes an insulating ring 1, a conductive induction ring 2, an insulating cylinder 3, and a coil 4; The insulating ring 1 and the conductive induction ring 2 are nested along the axial direction to form a composite nested unit. The insulating cylinder 3 has a first end and a second end that are axially opposite. The composite nested unit is disposed inside the first end of the insulating cylinder 3, and the inner peripheral wall of the first end of the insulating cylinder 3 covers the outer periphery of the composite nested unit. The coil 4 is disposed around the outer periphery of the second end of the insulating cylinder 3.

[0060] A gas channel is formed inside the insulating cylinder 3, which extends from the first end to the second end; The conductive induction ring 2 is connected to the first radio frequency power supply and is used to generate a ring electric field. The coil 4 is connected to the second radio frequency power supply and is used to generate an induced coupling electric field. A capacitive coupling electric field is formed between the insulating ring 1 and the conductive induction ring 2, and capacitively coupled plasma and inductively coupled plasma are generated in the gas channel.

[0061] The composite plasma generating unit is mounted on the upper electrode plate 7. The spray hole frame 6 is fixed below the upper electrode plate 7 and supports the lower part of the composite plasma generating unit. The insulating plate 5 is disposed between the upper electrode plate 7 and the spray hole frame 6 to isolate the potential between the upper electrode plate 7 and the spray hole frame 6, preventing discharge between the metal parts due to the potential difference.

[0062] Example 1: Insulating ring 1 is made of 95% alumina ceramic, with an outer diameter of 80 mm, an inner diameter of 60 mm, and a thickness of 5 mm. Conductive sensing ring 2 is made of oxygen-free copper, with an outer diameter of 78 mm, an inner diameter of 62 mm, a thickness of 3 mm, and a 10 μm silver plating on the surface. The two are fixed together by an interference fit with a radial interference of 0.05 mm.

[0063] The insulating cylinder 3 is made of 99% alumina ceramic, with a total length of 120mm, an outer diameter of 90mm, and an inner diameter of 70mm. The first end (upper section) is 40mm long, and its inner circumferential wall is fixed to the outer wall of the composite nested unit with phosphate ceramic adhesive. The curing conditions are: 250℃ for 1.5 hours. The second end (lower section) is 80mm long and has an outer diameter of 90mm.

[0064] Coil 4 is made of a hollow copper tube with an outer diameter of 6mm and a wall thickness of 1mm, wound 3 turns with a helix angle of 10°. After winding, the surface is coated with a 0.3mm polyimide insulation layer. The coil is fixed around the outer circumference of the second end of the insulating tube and pressed tightly with ceramic clips.

[0065] The insulating plate 5 is made of alumina ceramic, is a solid plate without holes, has an outer diameter of 100mm and a thickness of 4mm, and is set between the upper electrode plate 7 and the spray hole frame 6 for electrical isolation. The spray hole frame 6 is made of stainless steel, is fixed below the upper electrode plate 7, and supports the lower part of the composite plasma generating unit (i.e., below the second end of the insulating cylinder 3).

[0066] Example 2: Insulating ring 1 is made of aluminum nitride ceramic, with an outer diameter of 120mm, an inner diameter of 90mm, and a thickness of 8mm. Conductive sensing ring 2 is made of oxygen-free copper, with an outer diameter of 115mm, an inner diameter of 95mm, a thickness of 5mm, and a surface gold plating of 15μm. The interference fit is 0.08mm.

[0067] The insulating cylinder 3 is made of aluminum nitride ceramic, with a total length of 150mm, an outer diameter of 130mm, and an inner diameter of 100mm. The coil 4 is made of a hollow copper tube with an outer diameter of 8mm wound 5 times, with a helix angle of 15°. Other structures are the same as in Example 1.

[0068] III. The principle and phenomena of plasma generation.

[0069] Within a sealed reaction chamber, process gases are ionized under a strong electric field through the upper electrode plasma circular structure of this invention, forming plasma. These conductive gas plasmas, under the influence of an external ring-shaped electric field, form stable ring structures along the gas channels. In the plasma, collisions of charged particles generate glow discharges, emitting a pale purple light. This phenomenon can serve as a direct criterion for the ignition and stable existence of plasma.

[0070] The superimposed plasma formed by the simultaneous operation of CCP and ICP in the same gas channel has the characteristics of both high density and adjustable ion energy.

[0071] IV. How to use the upper electrode plasma device.

[0072] The upper electrode plasma device provided by this invention can operate in four modes: low-density mode (pure CCP), induction mode (pure ICP), high-density mode (CCP+ICP composite), and cleaning mode (ICP+oxygen). The operation process, control principle, and process effect of each mode are described in detail below through specific embodiments.

[0073] (I) Working principle and implementation examples of low-density mode (pure capacitively coupled plasma) In low-density mode, only the first radio frequency power supply connected to the conductive induction ring 2 is turned on, while the second radio frequency power supply connected to the coil 4 is turned off. At this time, a capacitively coupled electric field is formed between the conductive induction ring 2 and the insulating ring 1, and its electric field intensity distribution is described by Maxwell's equations. ,in Here, is the radio frequency voltage amplitude, and d is the radial gap between the insulating ring and the conductive ring. This electric field accelerates free electrons within the gas channel; these electrons collide with neutral gas molecules, ionizing them and generating and sustaining plasma. Due to the characteristics of capacitively coupled discharge, the plasma density is typically low (approximately 10⁻⁶). 9 ~10 10 cm -3 However, the ion energy can be independently adjusted by the radio frequency power, making it suitable for processes with precise requirements for ion energy, such as low-damage etching and oxide layer growth.

[0074] Example 3 (Low-damage etching of SiO2 in low-density mode): This embodiment uses a 200mm silicon wafer as the substrate, with a 500nm thermally oxidized SiO2 layer covering the surface. The process gas is a CHF3 / Ar / O2 mixed gas with flow rates of 10 sccm, 200 sccm, and 5 sccm, respectively, for a total flow rate of 215 sccm. The gas pressure in the reaction chamber is set to 2.0 Pa and is monitored in real time by a capacitive thin-film gauge.

[0075] The operation steps are as follows: First, switch to low-density mode and set the output parameters of the first RF power supply: frequency 13.56 MHz, initial power 400 W, pulse output, initial duty cycle 50%. Turn off the second RF power supply.

[0076] Next, initiate pressure feedback control. The pressure sensor samples the actual pressure in the gas channel every 10 milliseconds. The controller compares the measured pressure with the set value (2.0 Pa): When the air pressure rises to 2.5 Pa (i.e. 125% of the set value), it is determined that the air pressure is too high. At this time, excessive energy injection may cause arc discharge or cavity damage. The controller gradually reduces the duty cycle in 2% increments until it drops to 40% (the predetermined low duty cycle).

[0077] When the gas pressure drops to 1.5 Pa (i.e. 75% of the set value), it is determined that the gas pressure is too low. At this time, the plasma may be extinguished. The controller gradually increases the duty cycle in 2% increments until it reaches 60% (the predetermined high duty cycle).

[0078] The duty cycle adjustment range is 5% to 95% (predetermined duty cycle range). The duty cycle for each pulse cycle is calculated independently, and the adjustment method is continuous or step (step size 2%).

[0079] The physical basis of this feedback control is that changes in gas pressure reflect fluctuations in process conditions. By dynamically adjusting the duty cycle, plasma impedance matching can be maintained, avoiding increased reflected power and unstable discharge. Experiments have shown that after adopting this feedback control, the gas pressure fluctuation amplitude is reduced to within ±0.1 Pa.

[0080] Simultaneously, the Langmuir probe monitors the electron density and electron temperature within the gas channel in real time. In this embodiment, the measured plasma density is approximately 6.5 × 10⁻⁶. 9 cm -3 The electron temperature is approximately 2.5 eV. The substrate self-bias voltage, measured by a voltage probe, is approximately -220 V to -280 V (varying with duty cycle), corresponding to ion energies of approximately 200–250 eV.

[0081] Etching results: The etching rate was 24 nm / min, the on-wafer uniformity (measured at 49 points) was 2.1%, and the selectivity to the underlying silicon reached 12.5:1. In contrast, with a fixed duty cycle of 50%, a pressure fluctuation of ±0.3 Pa resulted in an etching rate fluctuation of ±8% and a decrease in uniformity to 5.2%. Pressure feedback duty cycle adjustment significantly improved process stability.

[0082] (II) Working principle and auxiliary ignition technology of induction mode (pure inductively coupled plasma).

[0083] In induction mode, only the second radio frequency power supply connected to coil 4 is turned on, while the first radio frequency power supply is turned off. The high-frequency alternating current flowing through the coil generates an alternating magnetic field. According to Faraday's law of electromagnetic induction, the changing magnetic field induces a vortex electric field within the gas channel. The vortex electric field accelerates electrons, causing them to ionize upon collision with gas molecules. Pure ICP can generate high-density plasma (10⁻⁶). 11 ~10cm -3 However, it has an inherent drawback: at low pressures (typically below 1 Pa), the induced electric field cannot effectively heat the electrons due to insufficient initial free electrons, leading to ignition difficulties. The composite structure of this invention precisely solves this problem: it utilizes the CCP electrode to temporarily generate an initial electron flow to assist ICP ignition.

[0084] Example 4 (In-situ cavity cleaning in induction mode, low-pressure assisted ignition).

[0085] This embodiment is used to remove fluorocarbon polymers deposited on the inner wall of the gas channel in the previous etching process. The process gas is oxygen (O₂). 2 The flow rate was 500 sccm, and the reaction chamber pressure was set to 0.9 Pa, which is lower than the self-sustaining ignition threshold of pure ICP by 1 Pa (the predetermined ignition pressure threshold).

[0086] Step 1: Pressure detection and determination of auxiliary ignition conditions.

[0087] The controller reads the pressure sensor value (0.9 Pa) and determines that it is lower than the predetermined ignition pressure threshold (1 Pa), so it automatically switches to the auxiliary ignition program.

[0088] Step 2: Pulse train assisted ignition.

[0089] Temporarily turn on the first radio frequency power supply (CCP source), outputting 5 auxiliary radio frequency pulses (predetermined pulse count range of 3-10) to the conductive induction loop 2 in the form of a pulse train. Parameters for each pulse: Pulse width: 50 μs (predetermined pulse width range 10-100 μs).

[0090] Pulse interval: 200 μs (predetermined pulse interval range 50-500 μs).

[0091] Pulse power amplitude: An incremental strategy is adopted, starting from 100 W of the first pulse, the power of each subsequent pulse increases by 20% compared to the previous one (the predetermined incremental slope range is 5% to 20%), and the values ​​are 100 W, 120 W, 144 W, 173 W, and 208 W respectively.

[0092] Gradually increasing the electric field strength can prevent local breakdown or sputtering of the electrodes caused by excessive initial pulse power, while gradually accumulating seed electrons to enable soft start of the plasma.

[0093] Step 3: Real-time monitoring of initial electron flux density.

[0094] During the application of the auxiliary pulse train, a Langmuir probe positioned on the inner wall of the insulating cylinder 3 detects the initial electron flux density n in the gas channel at a sampling interval of 10 μs. e The test results are as follows: After the first pulse ends, the initial electron flux density n e = 2.1×10 7 cm -3 .

[0095] After the second pulse ends, the initial electron flux density n e = 4.5×10 7 cm -3 .

[0096] After the third pulse ends, the initial electron flux density n e = 9.8 × 10 7 cm -3 .

[0097] After the fourth pulse ends, the initial electron flux density n e = 2.3×10 8 cm -3 .

[0098] The detected n e Compared with the predetermined ignition density threshold (1×10 8 cm -3 (Compare.) After the 4th pulse ends, n e The threshold has been exceeded, indicating that there are enough initial electrons in the gas channel to enable the ICP to self-sustain discharge.

[0099] Step 4: Switch to ICP power supply.

[0100] Immediately terminate the remaining auxiliary pulses (the 5th pulse is canceled), turn off the first RF power supply, and turn on the second RF power supply within 0.5 ms (the predetermined switching delay time range is 0.1 to 1 ms). The purpose of this short delay time is to prevent plasma extinction (if the delay is too long, the initial electrons will disappear due to recombination, resulting in ignition failure).

[0101] The second RF power supply provides continuous power to coil 4 at 13.56 MHz and 1200 W. With sufficient initial electrons already present, the induced electric field rapidly accelerates them, inducing avalanche ionization and establishing a stable pure ICP discharge within approximately 50 μs.

[0102] Step 5: Cleaning process.

[0103] Maintaining ICP discharge for 300 seconds resulted in oxygen dissociation, generating a high concentration of oxygen free radicals (O). Oxygen emission at 777 nm was monitored using optical emission spectroscopy (OES), showing an intensity approximately 25 times higher than the baseline. These oxygen free radicals then reacted with the fluoropolymer: C xFy +O → COF2+ CO2+ HF, and the generated gaseous products are removed by the vacuum system.

[0104] The etching rate was approximately 150 nm / min, and XPS analysis after 300 seconds showed that the residual carbon and fluorine on the inner wall of the gas channel decreased below the detection limit. Comparative experiments showed that if the ICP (0.9 Pa) was directly activated without auxiliary ignition, the RF reflection power reached as high as 80%, triggering the power supply protection and preventing plasma ignition. The auxiliary ignition technology of this invention completely solves the problem of low-pressure ICP ignition.

[0105] Example 5 (SiN deposition in induction mode, without auxiliary ignition): When the process gas pressure is higher than 1 Pa, pure ICP can ignite automatically without assistance. This embodiment uses SiN thin film deposition as an example: the process gas is SiH4 / N2 / He (flow rate 30 / 100 / 500 sccm), and the pressure is 5 Pa. The second RF power supply (13.56 MHz, 800 W) is directly turned on. The measured plasma density is 1.1 × 10⁻⁶. 11 cm -3 The deposition rate was 55 nm / min, the film uniformity was 2.3%, and the refractive index was 2.02. This indicates that the induction mode can be directly started under high pressure, and the operation is simple.

[0106] (III) Synergistic effects and decoupling control of high-density mode (CCP and ICP combined).

[0107] In the high-density mode, both the first and second RF power supplies are simultaneously activated, causing the capacitively coupled electric field and the inductively coupled electric field to superimpose within the gas channel. The composite mode exhibits the following synergistic effects: 1. Ignition coordination.

[0108] The initial electron flow generated by the CCP provides seed electrons for the ICP, enabling the ICP to reliably ignite even below 0.5 Pa (without auxiliary pulses).

[0109] 2. Density-energy decoupling.

[0110] CCP power primarily controls ion energy (through self-bias), while ICP power primarily controls plasma density. By independently adjusting the amplitude ratio of these two sources, independent control of ion energy and ion flux can be achieved, which is impossible with a single-principle source.

[0111] 3. Super-cumulative effect.

[0112] The process rate of the composite mode is often greater than the sum of the rates of pure CCP and pure ICP, proving that there is a nonlinear synergistic enhancement between the two.

[0113] Example 6 (Bosch process for deep silicon etching in high-density mode): The Bosch process alternates between etching and passivation steps to achieve high aspect ratio through-silicon via (TSV) etching. This embodiment uses a 200mm silicon wafer and a SiO2 hard mask.

[0114] Etching step parameters: Gas: SF6 / O2 = 300 / 20 sccm, gas pressure 1.5 Pa.

[0115] First RF power supply (CCP): 13.56 MHz, 400 W.

[0116] Second RF power supply (ICP): 13.56 MHz, 1500 W.

[0117] The initial power ratio was set to 1:3.75.

[0118] Density bias feedforward control is as follows: before applying dual power, the Langmuir probe measures the initial electron flux density n. e0 = 3×10 8 cm -3 Preset target density value n target = 5×10 8 cm -3 Calculate the density deviation Δ = (n e0 - n target ) / n target= -40%. The controller adjusts the initial power ratio based on the deviation: the larger the negative deviation, the less initial electrons are needed, requiring a stronger ICP to compensate. Therefore, the initial power ratio is set to 1:4 (i.e., CCP 400 W, ICP 1600 W), which increases the ICP share compared to the default ratio. This feedforward control allows the composite plasma to approach the target state immediately upon ignition, reducing subsequent settling time.

[0119] Closed-loop regulation during the process: Monitoring SF using emission spectroscopy (OES) x The emission intensity of free radicals (520 nm) is positively correlated with the concentration of fluorine free radicals, reflecting the etching rate. The process target value is set as OES intensity I. target Measured value I meas The controller compares the two every 0.5 seconds. If the deviation exceeds ±5% (a predetermined error range), it adjusts the ratio of the first and second RF power in 2% increments (keeping the total power approximately constant), so that I... meas Reply to I target nearby.

[0120] Decoupled control of ion energy and density: Meanwhile, the substrate self-bias (V bias The plasma density (n) is measured by a voltage probe. e ) Measured by the Langmuir probe. Target value V bias = -150 V (corresponding to an ion energy of approximately 150 eV), n e =5×10 10 cm -3 .

[0121] Self-bias adjustment rule: If V bias If the voltage is higher than -142.5 V (i.e., 5% higher than the target value), the controller will reduce the first RF power by 5% (first adjustment step); if V bias If the voltage drops below -157.5 V (5% below the target value), increase the first RF power by 5%. After adjustment, wait 0.2 seconds for it to stabilize before proceeding to the next test.

[0122] Density adjustment rule: If n e Higher than 5.5×10 10 cm -3 (10% higher than the target value), reduce the second RF power by 8% (second adjustment step); if n e Below 4.5×10 10 cm -3 If the value is 10% lower than the target value, then increase the second radio frequency power by 8%.

[0123] Decoupling verification: During the process, the first RF power was artificially changed from 400 W to 450 W (+12.5%), and the self-bias voltage was observed to increase from -150 V to -185 V (+23%), while the plasma density only increased from 5.0 × 10⁻⁶. 10 Increased to 5.3×10 10 cm -3 (+6%), demonstrating that ion energy is primarily controlled by the CCP. Conversely, increasing the second radio frequency power from 1500 W to 1800 W (+20%) and the density from 5.0 × 10⁻⁶ W... 10 Increased to 7.5×10 10 cm -3 (+50%), the self-bias voltage only dropped from -150 V to -135 V (-10%), proving that the density is mainly controlled by ICP. The decoupling effect between the two is good.

[0124] Etching results: The etching rate was 2.8 μm / min (compared to approximately 2.0 μm / min for the traditional Bosch process), with an aspect ratio of 25:1 (2 μm aperture, 50 μm depth), a sidewall roughness RMS of 15 nm, no bottom microgrooves, and a uniformity of 2.5%. Comparative experiments showed that the etching rate of pure ICP (without CCP) was only 1.9 μm / min, and sidewall bending occurred due to insufficient ion energy; the etching rate of pure CCP (without ICP) was only 0.5 μm / min. The combined etching rate was 1.47 times that of pure ICP and 5.6 times that of pure CCP, achieving both high density and anisotropy.

[0125] (iv) Smooth transition control for mode switching.

[0126] When a process needs to switch from one mode to another (e.g., from low-density etching to high-density etching, or from inductive deposition to composite etching), it is essential to ensure that the plasma does not extinguish and the process parameters do not change drastically during the switching process. This invention designs a five-stage smooth transition control strategy.

[0127] Example 7 (Switching from low-density mode to high-density mode).

[0128] Initial state: Low-density mode, process conditions are the same as in Example 3 (CCP 400 W, duty cycle 50%, gas pressure 2.0 Pa). It needs to be switched to high-density mode (CCP+ICP, target power CCP 500 W, ICP 2000 W, ratio 1:4).

[0129] Phase 1 (t = 0 ~ 200 ms, t is time): Maintain the current power output and ramp up another power supply.

[0130] After the controller issues a switching command, the first RF power supply remains constant at 400 W (to maintain plasma stability). Simultaneously, the second RF power supply is started with a linear ramp, its output amplitude rising from 0 at a rate of 8 W / ms. The ramp rise time is set to 0.5 seconds (predetermined ramp rise time range 0.2–2 seconds). The ramp curve can be selected as linear (in this embodiment) or exponential (to reduce initial impact).

[0131] Phase Two (t=200 ms): The percentage of those reaching the first threshold (80%).

[0132] When the output of the second radio frequency power supply reaches 80% of the target amplitude of 2000 W (i.e., 1600 W), the second stage begins. Before this (t=0~200 ms), the first radio frequency power supply remains constant at 400 W to maintain sufficient plasma density and ion energy, and to prevent shutdown due to excessively low dual power.

[0133] Phase 3 (t=200~600 ms): Adjust synchronously to the target ratio.

[0134] During the range where the second RF power supply output increases from 1600 W (80%) to 2000 W (100%), the two power supplies enter a synchronous adjustment phase. The controller adjusts simultaneously: The first RF power supply slowly increases from 400 W to 500 W at a rate of 0.5 W / ms.

[0135] The second RF power supply increased from 1600 W to 2000 W at a rate of 1.33 W / ms.

[0136] The total adjustment time was 400 ms (0.4 seconds), falling within the predetermined synchronous adjustment time range of 0.5 to 1 second (slightly lower than 0.4 seconds in this embodiment, but still within a reasonable range). The amplitude ratio of the two power supplies gradually transitioned from the initial 1:4 to the target 1:4 (the actual ratio was kept constant).

[0137] The key to this stage is maintaining plasma impedance matching: by monitoring the reflected power, the controller fine-tunes the adjustment rate to keep the matching network within an acceptable range.

[0138] Phase 4 (t=600 ms): Switching complete.

[0139] The second RF power supply reached 2000 W, and the first RF power supply reached 500 W, completing the switching. The plasma OES intensity fluctuated by less than 5% throughout the process, and the self-bias voltage smoothly transitioned from approximately -220 V to -310 V (naturally changing according to power variation) without any jumps or oscillations.

[0140] Phase 5 (Air pressure stabilizes).

[0141] Throughout the switching process, the pressure sensor monitors the pressure within the gas channel in real time. The controller adjusts the opening of the intake valve using a PID algorithm to ensure that the pressure fluctuation does not exceed ±10% of the set target value (2.0 Pa) (i.e., 1.8–2.2 Pa). The actual measured maximum deviation was 2.12 Pa (+6%), which meets the requirements.

[0142] Example 8 (Switching from sensing mode to high-density mode).

[0143] Initial state: Induction mode, ICP power 800 W, SiN deposition (same as Example 5), requiring switching to composite mode etching. The transition steps are similar: maintain ICP output at 800 W, start the CCP power supply with an exponential ramp (time constant 0.3 seconds), increasing from 0 to 500 W. When CCP reaches 400 W (80%), simultaneously adjust ICP from 800 W to 1500 W, and CCP from 400 W to 500 W, transitioning the power ratio from 1:2 to 1:3. After switching, plasma density increases, and self-bias voltage rises from approximately -50 V to -180 V; the plasma remains extinguished throughout the process.

[0144] V. Manufacturing method of upper electrode plasma device.

[0145] Example 9: Insulating ring 1 and conductive induction ring 2 are brazed at high temperature. Ag-Cu-Ti active solder is used, the brazing temperature is 900℃, and the vacuum degree is 5×10⁻⁶. -4 Pa, hold at that temperature for 30 minutes, then cool to room temperature at 3℃ / min. Before brazing, the surface of the conductive induction ring 2 is plated with 10 μm of silver and annealed at 400℃ in an argon atmosphere for 1.5 hours.

[0146] The composite nested unit is inserted into the first end of the insulating cylinder 3 and cured at 250°C for 1.5 hours using a phosphate ceramic adhesive. After the coil 4 is wound, a 0.2mm polyimide insulating layer is sprayed on it and fixed to the outer periphery of the second end of the insulating cylinder with ceramic clips.

[0147] The entire structure was sintered. Under a hydrogen atmosphere, it was sintered at 1300℃ for 2 hours, then cooled to room temperature at 3℃ / min, and then stress-relieved annealed at 250℃ for 2 hours.

[0148] Performance testing was performed. Helium gas chromatography-mass spectrometry was used for leak detection, with a leak rate of 8 × 10⁻⁶. -8 Pa·m 3 / s; Insulation resistance test, 800MΩ@500V DC between the insulation ring and the conductive induction ring; Coupling efficiency test, 91% efficiency at a 13.56MHz RF source.

[0149] Beneficial effects of the embodiments of the present invention: This invention integrates the CCP electrode and ICP coil at both ends of the same insulating cylinder, sharing a gas channel. This eliminates the need for two separate cavities or power interfaces, significantly saving equipment space and manufacturing costs. The nested axial segmented design spatially separates the two electric fields, preventing mutual interference and ensuring the stability of the composite mode. An insulating plate is positioned between the upper electrode plate and the spray hole frame, achieving reliable electrical isolation between them and preventing discharge due to potential differences between metal components, while not affecting plasma flow from the gas channel. Through high-temperature brazing, interference fit, and integral hydrogen sintering, the device achieves excellent airtightness and insulation performance, meeting the stringent requirements of semiconductor manufacturing.

[0150] In terms of control methods, in low-density mode, the duty cycle of the pulse CCP is dynamically adjusted using gas pressure feedback, effectively suppressing the impact of gas pressure fluctuations on discharge stability and significantly improving etching uniformity. In induction mode, addressing the inherent defect of ICP sources—difficulty in ignition under low gas pressure—this invention utilizes an auxiliary pulse train with increasing power applied to the CCP electrode and monitors the initial electron flux density in real time. Once the ignition threshold is reached, the power supply is quickly switched to the ICP, completely solving the low-pressure ignition problem. Simultaneously, the increasing power strategy avoids electrode damage. The high-density composite mode is the core breakthrough of this invention: through density deviation feedforward control, the initial power ratio of CCP and ICP is dynamically set before ignition, allowing the plasma to quickly enter the target state; by independently adjusting the power amplitude of the two RF power supplies, decoupled control of ion energy and plasma density is achieved for the first time in the same device—adjusting the CCP power mainly changes the ion energy, and adjusting the ICP power mainly changes the plasma density, filling a long-standing technological gap in this field. Experiments show that the process rate of the composite mode exceeds the sum of the pure CCP and pure ICP rates, producing a super-superposition effect, demonstrating the nonlinear synergistic enhancement of the two principles. Meanwhile, closed-loop modulation of the emission spectrum ensures the consistency of plasma chemical activity. During mode switching, a five-stage smooth transition control is employed to prevent plasma extinction and parameter jumps, ensuring process continuity. Furthermore, the same device can achieve four operating modes: pure CCP, pure ICP, recombination, and cleaning, greatly expanding the process window and reducing reliance on multiple dedicated devices.

[0151] Compared with existing single-principle plasma sources, this invention has comprehensive advantages such as high density, independently adjustable ion energy, reliable ignition at low pressure, and flexible switching between multiple modes. It solves the technical problems that have long plagued the semiconductor etching field and has outstanding substantial progress and broad industrial application prospects.

[0152] The above description is merely a specific embodiment of the present invention and should not be construed as limiting the scope of the invention. Therefore, any substitution of equivalent components or equivalent changes and modifications made within the scope of protection of this patent should still fall within the scope of this patent. Furthermore, the technical features, technical features and technical solutions, and technical solutions in this invention can be freely combined and used.

Claims

1. A top electrode plasma device, characterized in that, include: Composite plasma generating unit, insulating plate (5), spray hole frame (6) and upper electrode plate (7); The composite plasma generating unit includes an insulating ring (1), a conductive induction ring (2), an insulating cylinder (3), and a coil (4). The insulating ring (1) and the conductive induction ring (2) are nested along the axial direction to form a composite nested unit. The insulating cylinder (3) has a first end and a second end that are axially opposite each other. The composite nested unit is disposed inside the first end of the insulating cylinder (3), and the inner peripheral wall of the first end of the insulating cylinder (3) covers the outer periphery of the composite nested unit. The coil (4) is disposed around the outer periphery of the second end of the insulating cylinder (3). A gas channel is formed inside the insulating cylinder (3), and the gas channel extends from the first end to the second end; The conductive induction ring (2) is connected to the first radio frequency power supply and is used to generate a ring electric field. The coil (4) is connected to the second radio frequency power supply and is used to generate an induced coupling electric field. A capacitive coupling electric field is formed between the insulating ring (1) and the conductive induction ring (2), and capacitively coupled plasma and inductively coupled plasma are generated in the gas channel. The composite plasma generating unit is disposed on the upper electrode plate (7), and the spray hole frame (6) is fixed below the upper electrode plate (7) and supports the lower part of the composite plasma generating unit; The insulating plate (5) is disposed between the upper electrode plate (7) and the spray hole frame (6) to isolate the potential between the upper electrode plate (7) and the spray hole frame (6).

2. A method of using an upper electrode plasma device, wherein the upper electrode plasma device of claim 1 generates composite plasma for processing a substrate, characterized in that, This includes the following steps performed by the controller: Process gas is introduced into the gas channel inside the insulating cylinder (3); The working mode is selected according to the process requirements, including low-density mode, sensing mode, high-density mode and cleaning mode. When the operating mode is the low-density mode, the first radio frequency power is applied to the conductive induction ring (2) through the first radio frequency power supply, and the second radio frequency power supply is turned off to generate pure capacitively coupled plasma; When the working mode is the induction mode, the second radio frequency power is applied to the coil (4) through the second radio frequency power supply, and the first radio frequency power supply is turned off to generate pure inductively coupled plasma; When the working mode is the high-density mode, the first radio frequency power is applied to the conductive induction ring (2) through the first radio frequency power supply, and the second radio frequency power is applied to the coil (4) through the second radio frequency power supply at the same time, and the power amplitude of the second radio frequency power is greater than the power amplitude of the first radio frequency power, so as to generate high-density composite plasma. When the working mode is the cleaning mode, the first radio frequency power supply is turned off, and the second radio frequency power is applied to the coil (4) only through the second radio frequency power supply, and oxygen-containing cleaning gas is introduced into the gas channel to generate pure inductively coupled oxygen plasma to remove the deposits on the inner wall of the gas channel; The plasma flow is distributed to the surface of the substrate to be processed.

3. The method of use according to claim 2, characterized in that, When the operating mode is the low-density mode, a first radio frequency power is applied to the conductive induction ring (2) through the first radio frequency power supply, and the second radio frequency power supply is turned off to generate pure capacitively coupled plasma, including: When applying the first radio frequency power to the conductive sensing ring (2), the first radio frequency power is applied in a pulse manner; The current gas pressure value in the gas channel is detected in real time, and the duty cycle of each pulse cycle is independently calculated based on the current gas pressure value. When the gas pressure value is higher than a predetermined high gas pressure threshold, the duty cycle is reduced until a predetermined low duty cycle is reached. When the gas pressure value is lower than the predetermined low gas pressure threshold, the duty cycle is increased until a predetermined high duty cycle is reached, so that the duty cycle is continuously or stepwise adjusted within the predetermined duty cycle range.

4. The method of use according to claim 2, characterized in that, When the operating mode is the induction mode, before applying the second radio frequency power to the coil (4) through the second radio frequency power supply, the following steps are also included: Detect the current gas pressure value in the gas channel. If the current gas pressure value is lower than the predetermined ignition gas pressure threshold, execute the auxiliary ignition step, including: Temporarily turn on the first radio frequency power supply and apply a set of auxiliary radio frequency pulses to the conductive induction ring (2) in the form of a pulse train. The width of each auxiliary radio frequency pulse is within a predetermined pulse width range, the pulse interval is within a predetermined pulse interval range, the number of pulses in the pulse train is within a predetermined number of pulses range, and the power amplitude of each pulse increases sequentially from the first to the last, with the increasing slope being within a predetermined increasing slope range. During the application of the auxiliary radio frequency pulse train, the initial electron flow density in the gas channel is detected in real time by a Langmuir probe set on the inner wall of the insulating cylinder (3); After each auxiliary radio frequency pulse ends, the detected initial electron flow density is compared with a predetermined ignition density threshold. If the current density value reaches or exceeds the predetermined ignition density threshold, the remaining auxiliary radio frequency pulses are immediately terminated, the first radio frequency power supply is turned off, and the second radio frequency power supply is turned on within a predetermined switching delay time range, so that the coil (4) starts to output the second radio frequency power. If the current density value is lower than the predetermined ignition density threshold, the next auxiliary radio frequency pulse is applied.

5. The method of use according to claim 2, characterized in that, When the operating mode is the high-density mode, a first radio frequency power is applied to the conductive induction ring (2) through the first radio frequency power supply, and a second radio frequency power is applied to the coil (4) through the second radio frequency power supply, and the power amplitude of the second radio frequency power is greater than the power amplitude of the first radio frequency power, thereby generating high-density recombination plasma, including: Before simultaneously applying the first radio frequency power and the second radio frequency power, the initial electron flow density value of the inner wall of the insulating cylinder (3) is obtained, and the initial electron flow density value is compared with the predetermined target density value to calculate the density deviation; The initial amplitude ratio of the first RF power to the second RF power is dynamically set according to the density deviation, so that the amplitude of the second RF power is greater than the amplitude of the first RF power, and the initial amplitude ratio is positively correlated with the density deviation. During the simultaneous application of the first radio frequency power and the second radio frequency power, the output amplitude of the first radio frequency power supply and the output amplitude of the second radio frequency power supply are adjusted respectively to independently change the actual amplitude of the first radio frequency power and the second radio frequency power. The amplitude of the first radio frequency power is adjusted to control the ion energy bombarding the substrate surface, and the amplitude of the second radio frequency power is adjusted to control the plasma density in the gas channel. The emission spectrum intensity or ion current density of the composite plasma in the gas channel is detected in real time. The detected value is compared with the process target value. Based on the comparison result, the amplitude ratio of the first radio frequency power and the second radio frequency power is repeatedly adjusted to stabilize the characteristic parameters of the composite plasma within a predetermined error range.

6. The method of use according to claim 5, characterized in that, Also includes: While simultaneously applying the first RF power and the second RF power, the self-bias voltage on the substrate surface and the plasma density in the gas channel are measured. The self-bias voltage is compared with a predetermined target ion energy value: if the self-bias voltage is higher than a first positive deviation threshold of the target value, the current amplitude of the first radio frequency power is reduced by a first adjustment step; if the self-bias voltage is lower than a first negative deviation threshold of the target value, the current amplitude of the first radio frequency power is increased by the first adjustment step; otherwise, the first radio frequency power remains unchanged. The plasma density is compared with a predetermined density target value. If the plasma density is higher than the second positive deviation threshold of the target value, the current amplitude of the second radio frequency power is reduced by the second adjustment step. If the plasma density is lower than the second negative deviation threshold of the target value, the current amplitude of the second radio frequency power is increased by the second adjustment step. Otherwise, the second radio frequency power is kept unchanged. After each adjustment, wait for a predetermined stabilization time and repeat the above steps until the self-bias voltage and the plasma density simultaneously meet their respective allowable deviation ranges.

7. The method of use according to any one of claims 2 to 6, characterized in that, Also includes: When switching from the low-density mode to the high-density mode, or from the sensing mode to the high-density mode, or from the high-density mode to the low-density mode or the sensing mode, the following transition control steps are performed: While maintaining the output of the currently activated RF power supply, the output amplitude of the RF power supply to be activated is gradually increased in a ramp manner. The rise time of the ramp is set to a predetermined ramp rise time range, and the curve of the ramp is set to linear or exponential. When the output amplitude of the radio frequency power supply to be turned on does not reach the first threshold ratio of the target amplitude, the output amplitude of the currently turned-on radio frequency power supply remains unchanged. When the output amplitude of the radio frequency power supply to be turned on enters the second threshold ratio range of the target amplitude, the output amplitudes of the two radio frequency power supplies are adjusted synchronously so that the amplitude ratio of the two reaches the target ratio within a predetermined synchronous adjustment time range. If it is necessary to turn off the currently outputting RF power supply, after the output amplitude of the RF power supply to be turned on reaches the third threshold ratio of the target amplitude, the output amplitude of the RF power supply to be turned off is reduced to zero in a ramp manner within a predetermined fall time range. During the switching process, the gas pressure value is monitored in real time and the process gas flow rate is adjusted to ensure that the fluctuation of the gas pressure value does not exceed the predetermined allowable range of gas pressure fluctuation of the set target value.

8. A method for manufacturing an upper electrode plasma device, the method being used to manufacture the upper electrode plasma device according to claim 1, characterized in that, Includes the following steps: The insulating ring (1) and the conductive sensing ring (2) are nested along the axial direction to form the composite nested unit, and the insulating ring (1) and the conductive sensing ring (2) are fixedly connected by high-temperature brazing or interference fit. The composite nested unit is inserted into the interior of the first end of the insulating cylinder (3), so that the inner peripheral wall of the first end covers the outer periphery of the composite nested unit and is fixed by ceramic adhesive or mechanical retaining ring; The coil (4) is arranged around the outer periphery of the second end of the insulating cylinder (3), and the coil (4) is fixed on the insulating cylinder (3) by an insulating fastener; The insulating cylinder (3) is installed on the upper electrode plate (7) so that the gas channel inside the insulating cylinder (3) is connected to the air hole of the upper electrode plate (7); The insulating plate (5) is placed between the upper electrode plate (7) and the spray hole frame (6) to isolate the potential between the upper electrode plate (7) and the spray hole frame (6); The spray hole frame (6) is fixed below the upper electrode plate (7) and supports the insulating plate (5). The assembled upper electrode plasma device was subjected to overall sintering and performance testing.

9. The manufacturing method according to claim 8, characterized in that, The high-temperature brazing uses Ag-Cu-Ti active brazing filler metal, with a brazing temperature of 850℃~950℃ and a vacuum degree of not less than 1×10⁻⁶. -3 Pa, after brazing, cool to room temperature at a rate of 2℃ / min to 5℃ / min; The radial interference of the interference fit is 0.02 mm to 0.10 mm, and the fitting force is 500 N to 2000 N. Before nesting the insulating ring (1) and the conductive sensing ring (2), the conductive sensing ring (2) is subjected to surface silver plating treatment with a silver plating layer thickness of 5 μm to 20 μm, and is annealed at 300℃ to 500℃ for 1 to 2 hours in an inert gas atmosphere. The ceramic adhesive is a phosphate-based or silicate-based ceramic adhesive with a temperature resistance of not less than 800℃, and the curing conditions are to keep it at 200℃~300℃ for 1~2 hours. The insulating fastener is a ceramic clip or a mica gasket. The coil (4) is wound with a hollow copper tube with a diameter of 3 mm to 8 mm, 2 to 5 turns, a helix angle of 5° to 15°, and a polyimide insulating layer with a thickness of 0.1 mm to 0.5 mm and a withstand voltage of not less than 5 kV.

10. The manufacturing method according to claim 8, characterized in that, The assembled upper electrode plasma device undergoes overall sintering and performance testing, including: The assembled upper electrode plasma device is placed in a sintering furnace and sintered at 1200℃~1400℃ for 1~3 hours in a hydrogen or argon atmosphere to form a metallurgical bond or densification at each fixed connection interface. After sintering, the material is cooled to room temperature at a rate of 2℃ / min to 5℃ / min, and then stress-reduced annealed at 200℃ to 300℃ for 2 hours. The upper electrode plasma device after integral sintering is subjected to the following tests to determine whether the upper electrode plasma device is qualified: Helium gas at a pressure of 0.2 MPa to 0.5 MPa is introduced into the gas channel, and the leak rate is detected using a helium mass spectrometer leak detector. The leak rate is considered acceptable when it is not greater than 1 × 10⁻⁶ MPa. -7 Pa·m 3 When the plasma device is at a rate of / s, the upper electrode plasma device is qualified; Apply 500 V DC between the insulating ring (1) and the conductive induction ring (2), and measure the insulation resistance to be no less than 100 MΩ. Apply 500 V DC between the insulating plate (5) and the upper electrode plate (7). When the measured insulation resistance is no less than 100 MΩ, the upper electrode plasma device is qualified. Connect an RF source and detect the coupling efficiency between the capacitively coupled plasma and the inductively coupled plasma generated by the upper electrode plasma device. When the coupling efficiency is not less than 85%, the upper electrode plasma device is qualified.

Images