Remote plasma source load impedance perturbation protection control method and device

Abstract

The present application belongs to the technical field of remote plasma source, and provides a remote plasma source load impedance disturbance protection control method and device. The method comprises: collecting the gas flow, gas pressure and reaction chamber temperature of the reaction chamber; obtaining a load impedance prediction value according to a load impedance prediction model; and controlling the system to work in a constant power phase or a constant current phase according to whether the load impedance prediction value F is in a fluctuation stable interval. The present application can predict the load impedance value F under different gas flow, pressure and chamber temperature by using the load impedance prediction model, and determine whether the working state is stable according to the load impedance value, so as to control the system to work in a constant power phase or a constant current phase. The present application realizes real-time collaborative regulation of gas load fluctuation and plasma dynamics, effectively suppresses the mixed disturbance during gas switching, solves the problem of plasma instability, avoids damage to power devices caused by excessive current due to gas load fluctuation, and improves the service life of the remote plasma source.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of remote plasma source, and particularly relates to a remote plasma source load impedance disturbance protection control method and device. BACKGROUND

[0002] In the high-end industrial fields such as semiconductor manufacturing, thin film deposition and material surface modification, remote plasma source (RPS) is widely used due to its unique advantages. Unlike traditional direct plasma technology, remote plasma source effectively avoids the direct bombardment of high-energy ions on sensitive substrates by physically separating the plasma generation zone from the reaction zone, and simultaneously realizes the uniform distribution of active particles through long-range transmission. However, this decoupling design also introduces new challenges: the fluctuation of gas load (such as instantaneous changes in flow, pressure or composition) will significantly affect the stability and reaction characteristics of the plasma, thereby leading to a decrease in process repeatability, degradation of film uniformity and even a reduction in equipment life.

[0003] Studies have shown that the gas transmission path of the remote plasma source is relatively long, and the gas is easily disturbed by factors such as pipeline impedance, vacuum pump pumping speed and precursor injection timing during the flow process, forming dynamic load fluctuations. This fluctuation will affect the process quality through two mechanisms: first, the non-steady-state change of the gas state directly leads to fluctuations in the plasma ionization efficiency, causing the drift of active particle concentrations such as free radicals and excited state substances; second, the transient fluctuation of the gas load will change the plasma impedance characteristics, causing a decrease in the coupling efficiency of the radio frequency (RF) or microwave power, further amplifying the instability of plasma parameters such as electron density and ion energy distribution. Although existing research has achieved macro flow control through gas mass flow controllers (MFCs) and pressure feedback systems, there is still a lack of effective means to suppress transient fluctuations on the microsecond to millisecond scale.

[0004] Although the gas load fluctuation problem of the remote plasma source (RPS) has attracted widespread attention, the existing protection control technology still has the following key defects:

[0005] 1. Insufficient dynamic response speed: the response time of existing gas control systems (such as mass flow controllers (MFCs) and pressure sensors) is usually on the order of milliseconds to hundreds of milliseconds, making it difficult to effectively capture transient fluctuations (such as pressure oscillations during pulsed gas supply or mixing disturbances during gas switching) on the microsecond to millisecond scale. This lag results in the inability of the control system to match the dynamic changes of the plasma impedance in real time, thereby causing radio frequency power reflection or plasma instability.

[0006] 2. Limitations of control model linearization assumptions: Traditional control strategies (such as PID algorithms) are based on linear or weakly nonlinear assumptions, while the gas-plasma coupling process of a remote plasma source has strong nonlinear and time-varying characteristics (such as the exponential relationship between gas dissociation rate and flow rate, dynamic fluctuations in plasma sheath impedance). Such models are difficult to accurately describe the interaction effects of multiple physical fields, resulting in control parameters deviating from the optimal interval.

[0007] 3. Lack of multi-parameter coupling regulation: The impact of gas load fluctuations usually involves the coordinated changes of multiple parameters such as gas flow, pressure, temperature, etc., while existing protection control often uses isolated parameter feedback mechanisms, lacking global optimization of gas-plasma-electromagnetic field coupling relationships. For example, when adjusting the flow rate only through pressure compensation, the reverse interference of plasma impedance changes on power coupling efficiency may be ignored.

[0008] 4. Insufficient adaptability to complex conditions: In advanced processes such as pulsed modulation plasma and multi-gas path alternate injection, gas load fluctuations exhibit high-frequency and non-periodic characteristics, while existing control algorithms rely on fixed thresholds or empirical rules, making it difficult to dynamically adjust control strategies, resulting in a decline in process repeatability (such as a film thickness deviation of more than ±5%). SUMMARY

[0009] To overcome the shortcomings of the prior art, the present application provides a remote plasma source load impedance disturbance protection control method and device to solve the problem of the need for high-precision regulation mechanism of multiple physical field cooperation in the current remote plasma source.

[0010] In a first aspect, the present application provides a remote plasma source load impedance disturbance protection control method, after the reaction chamber is normally ignited, the method comprises:

[0011] Collecting the gas flow Q, gas pressure P and reaction chamber temperature T of the reaction chamber;

[0012] Obtaining a load impedance prediction value F according to a load impedance prediction model;

[0013] According to whether the load impedance prediction value F is in the fluctuation stable interval, the control system works in the constant power stage or the constant current stage; if the load impedance prediction value is in the fluctuation stable interval, the control system works in the constant power stage, otherwise the control system works in the constant current stage.

[0014] From the above technical solutions can be known, the remote plasma source load impedance disturbance protection control method provided by the application can predict the load impedance value F by using the load impedance prediction model under different gas flow, gas type and cavity temperature changes, the system works in the constant current stage when the ignition working stage or the predicted load impedance is unstable, constant current output is carried out near the resonance frequency point, current damage to the power device is avoided; when the predicted load impedance is stable, the system automatically enters the constant power mode to keep the power output constant.

[0015] Optionally, the load impedance prediction model is ; Z is the load impedance, k1, k2, k3 are variable coefficients, and k0 is a constant term;

[0016] The objective function is , Q i , P i , t i , Z i is n groups of data, ;

[0017] The load impedance prediction value F is determined by fitting the variable coefficient and the constant term.

[0018] Optionally, the fitting method of the variable coefficient and the constant term comprises:

[0019] The data matrix X and the observation vector Z are constructed; ;

[0020] The variable vector to be solved is , and the derivative is zero, to obtain the standardization equation ;

[0021] The optimal solution of the variable to be solved is .

[0022] Optionally, the fluctuation stable interval is , is the average value of a plurality of the impedance prediction values F obtained within the interval sampling period T s in the unit time T0, and d is the load impedance fluctuation error.

[0023] Optionally, the constant current stage comprises:

[0024] The current error e1 is obtained according to the resonant output current reference value and the current i pri obtained by sampling, and the duty cycle d is output through the current PI modulator;

[0025] The PWM drive signals D1-D4 of the inverter switching tube are output through the PWM modulation.

[0026] Optionally, the constant power stage comprises:

[0027] The output power reference value P ref is multiplied by the resonant output voltage V loop and the current i pri to obtain the actual power Po, and the output power error e0 is compared, and the resonant output current reference value i ref is obtained after power PI modulation.

[0028] The current reference value i ref is compared with the current value i pri sampled by the power circuit to obtain the current error e1, and the duty cycle d is output by the current PI modulator.

[0029] The current error e1 is PI modulated and inverted to obtain the switching frequency f s , and the PWM drive signal D1-D4 of the inverter switching tube is output by the PWM modulation.

[0030] In a second aspect, the application provides a remote plasma source load impedance disturbance protection control device, comprising:

[0031] A collection module is configured to collect the gas flow Q, gas pressure P and reaction chamber temperature T of the reaction chamber.

[0032] An impedance prediction module is configured to obtain a load impedance prediction value F according to a load impedance prediction model.

[0033] A disturbance control module is configured to control the system to work in a constant power stage or a constant current stage according to whether the load impedance prediction value F is in a fluctuation stable interval; if the load impedance prediction value is in the fluctuation stable interval, the system works in the constant power stage, otherwise the system works in the constant current stage.

[0034] The above technical solutions have the following beneficial effects:

[0035] The application can predict the load impedance value F by using the load impedance prediction model under different gas flow, gas type and cavity temperature changes, determine whether the working state is stable according to the load impedance value, work in the constant current stage when the system is in the ignition working stage or the predicted load impedance is unstable, perform constant current output near the resonant frequency point, and avoid damaging the power device due to excessive current; when the predicted load impedance is stable, the system automatically enters the constant power mode to keep the power output constant. BRIEF DESCRIPTION OF DRAWINGS

[0036] In order to more clearly illustrate the technical solutions in the specific embodiments of the present application or the prior art, the drawings required to be used in the description of the specific embodiments or the prior art will be briefly introduced below. In all the drawings, similar elements or parts are generally identified by similar reference signs. In the drawings, the elements or parts are not necessarily drawn according to the actual proportions.

[0037] Figure 1 A control block diagram of a remote plasma source provided by an embodiment of the present application is shown;

[0038] Figure 2 A flow chart of a remote plasma source load impedance disturbance protection control method provided by an embodiment of the present application is shown;

[0039] Figure 3 An application flow chart of a remote plasma source load impedance disturbance protection control method provided by an embodiment of the present application is shown;

[0040] Figure 4 A structural block diagram of a remote plasma source load impedance disturbance protection control device provided by an embodiment of the present application is shown. DETAILED DESCRIPTION

[0041] The embodiments of the technical solutions of the present application will be described in detail below with reference to the drawings. The following embodiments are only used to more clearly illustrate the technical solutions of the present application, and therefore are only examples, and cannot limit the protection scope of the present application.

[0042] It should be noted that, unless otherwise specified, the technical terms or scientific terms used in the present application should be the usual meanings understood by the skilled in the art to which the present application belongs.

[0043] As Figure 1As shown, a control block diagram of a remote plasma source is provided, including a remote plasma source power circuit, a constant current stage unit, and a constant power stage unit. The remote plasma source power circuit includes a control unit, a PWM drive unit, a DC / AC resonant converter, an ignition circuit, a sustaining circuit, a sampling unit, and a reaction chamber. The control unit is used to control the output pulse signal of the PWM drive unit. The PWM drive unit controls the output current of the resonant converter. The output current of the resonant converter affects the input current of the ignition circuit. The resonant converter converts the pulse signal output by the PWM drive unit into an AC source and uses it to provide an input AC source for the ignition circuit. The ignition circuit generates a high-voltage ignition signal, causing the plasma load to oscillate and ionize at high frequency. The resonant converter is used to provide the input AC source. The resonant converter includes a transformer T1, and the ignition circuit and the secondary coil N3 of the transformer T1 are connected. The power circuit also includes a sustaining circuit, which is connected in series with the secondary coil N2 of transformer T. The vacuum reaction chamber is used as the secondary side of the high-frequency transformer, which can only provide io to the chamber to meet the low-voltage energy input during the ionization sustaining stage. The sustaining circuit is used to maintain the bus voltage stability after the constant current stage (ignition stage) is switched to the constant power stage (ionization stage) when the relay is disconnected. The sampling unit is used to collect the ignition current of the ignition circuit, as well as the output voltage and output current of the resonant converter. The relay is controlled by the control unit to be engaged or disengaged.

[0044] like Figure 2 As shown, a remote plasma source load impedance disturbance protection control method is provided. After normal ignition of the reaction chamber, the method includes:

[0045] S1. Collect the gas flow rate Q, gas pressure P, and reaction chamber temperature T in the reaction chamber.

[0046] The above parameters are acquired through a multi-sensor module, which includes a gas flow meter, a pressure sensor, and a temperature sensor to monitor the gas flow rate Q, gas pressure P, and reaction chamber temperature T of the reaction chamber, respectively.

[0047] S2. Based on the load impedance prediction model, obtain the predicted load impedance value F.

[0048] In a remote plasma system, the effects of gas flow rate Q, gas pressure P, and reaction chamber temperature T on the load impedance can be described by the following ideal gas model:

[0049] When the pumping speed S of the vacuum pump is fixed, the pressure P is directly proportional to the flow rate P=Q / S, the load impedance Z is directly proportional to the flow rate and inversely proportional to the temperature T, and the effect of the gas flow rate Q on the load impedance Z can be expressed by equation (1):

[0050] (1)

[0051] in is a proportionality constant, which is related to system parameters such as electrode area, collision cross section, etc.

[0052] At constant temperature, the load impedance is proportional to the pressure:

[0053] (2)

[0054] where is a proportionality constant, which is related to plasma density and collision frequency.

[0055] At constant pressure, the load impedance is inversely proportional to the reaction chamber temperature due to the decrease in gas density:

[0056] (3)

[0057] where is a proportionality constant.

[0058] where the load impedance prediction model in step S2 is

[0059] (4)

[0060] Z is the load impedance, k1, k2, k3 are variable coefficients, and k0 is a constant term; Q i , P i , t i , Z i is the n group of observation data, ;

[0061] The load impedance prediction value F is determined by fitting the variable coefficients and the constant term.

[0062] Optionally, the fitting method of the variable coefficients and the constant term includes:

[0063] Constructing a data matrix X and an observation vector Z;

[0064] (5)

[0065] The objective function of the least squares method is to minimize the sum of squares of residuals:

[0066] (6)

[0067] The variable vector to be solved is Take the partial derivative and set the derivative to zero to get the standardization equation:

[0068] (7)

[0069] The optimal solution of the variable to be solved is (8). The parameter optimal solution is substituted into equation (4) to obtain the current load impedance prediction value F.

[0070] S3. According to whether the load impedance prediction value F is in the fluctuation stable interval, the control system works in the constant power phase or the constant current phase; if the load impedance prediction value is in the fluctuation stable interval, the control system works in the constant power phase, otherwise the control system works in the constant current phase.

[0071] In step S3, the fluctuation stable interval is , is the interval sampling period T s The average of the obtained plurality of impedance prediction values F, and d is the load impedance fluctuation error.

[0072] Using the above method, the load impedance value F can be predicted by the load impedance prediction model under different gas flow, gas type and cavity temperature changes. The system works in the constant current phase when the ignition working phase or the predicted load impedance is unstable, and works in the constant current output near the resonance frequency point to avoid damage to the power device caused by excessive current. When the predicted load impedance is stable, the system automatically enters the constant power mode to maintain constant power output.

[0073] Referring to Figure 1 , the constant current phase includes:

[0074] According to the resonant output current reference value and the sampled current i pri , the current error e1 is obtained, and the duty cycle d is output through the current PI modulator;

[0075] The PWM drive signal D1~D4 of the inverter switching tube is output through PWM modulation.

[0076] Referring to Figure 1 , the constant power phase includes:

[0077] The output power reference value P ref is multiplied by the resonant output voltage V loop and the current i pri to obtain the actual power Po, and the output power error e0 is compared and output, and the resonant output current reference value i ref is obtained after power PI modulation.

[0078] The current reference value i ref is compared with the current value i pri sampled by the power circuit to output the current error e1, and the duty cycle d is output through the current PI modulator.

[0079] The current error e1 is modulated by PI and inverted to obtain the switching frequency f sThe PWM drive signals D1-D4 of the inverter switching tubes are outputted by PWM modulation.

[0080] As shown in the figure, the remote plasma source load impedance disturbance protection control method provided by the embodiment has the following steps in specific application: Figure 3

[0081] Step A1: After the remote plasma source is started, the system first sets the gas current excitation threshold imin, and inputs the excitation gas. Through the stage adjustment unit, the resonant output current reference value Iref is reduced to the current reference value Iref0 at 50 ms before the ignition operation, which can effectively avoid the inrush current and protect the relay S.

[0082] Step A2: It is judged whether the ignition signal is received. If the ignition instruction is received, the suction relay S is driven to output, otherwise the system inputs the excitation gas. The resonant output current reference value Iref is compared with the current value ipri sampled by the power circuit, and the current error e is outputted. After the current PI modulation, the PWM drive unit is started. It is judged whether imin≤ipri≤imax is met and maintained for more than 500 ms. If it is met, it is judged whether the system is successfully ignited, otherwise the system is reported as an error and the relay S is disconnected.

[0083] Step A3: If it is judged that the ignition is not normal, the duty cycle d in the constant current stage is increased, so that the resonant output voltage Vloop is increased, the output power is increased, the ignition voltage Vcd in the ignition circuit of the coupling transformer T1 is also increased, and it is judged again whether the ignition is normal. When the ignition is successful, the step of disconnecting the relay S is jumped to. If the ignition fails, the cycle ignition processing is performed, and the upper limit value of the cycle number is N times, so as to improve the ignition success rate. If the ignition is not successful after exceeding the cycle upper limit value, the system is reported as an error and the relay S is disconnected.

[0084] Step A4: The predicted value of the load impedance is outputted by using the load impedance prediction model. In actual working conditions, increasing the output power of the plasma source will cause the fluctuation of the load impedance. Therefore, the system works in the constant current stage when the ignition working stage or the predicted load impedance is unstable, that is, or The system enters the constant current stage. Due to the circuit characteristics of the LCL resonant converter, after the fixed resistance inductance, capacitance and other hardware parameters, the system works near the resonant frequency point to output the constant current, avoids the damage of the power device caused by the excessive current. When the predicted load impedance is stable, that is, ​​​​​When the load impedance fluctuation error d is negative, the system automatically enters the constant power mode to keep the power output constant;

[0085] The constant power stage control flow includes: output power reference value P ref is multiplied by the resonant output voltage V loop and the current i pri to obtain the actual power P o , and the output power error e0 is compared, and the resonant output current reference value i ref is obtained after power PI modulation, and the current reference value i ref is compared with the current value i pri sampled by the power circuit to output the current error e1, and the duty cycle d is output through the current PI modulator; The switching frequency f s is obtained by taking the opposite of the current error e1 through the PI modulator, and finally the PWM drive signal D1~D4 of the inverter switch tube is output through the PWM drive unit.

[0086] In one embodiment, as shown in Figure 4 , a remote plasma source load impedance disturbance protection control device 40 is provided, comprising:

[0087] The acquisition module 401 is configured to acquire the gas flow Q, the gas pressure P and the reaction chamber temperature T of the reaction chamber.

[0088] The impedance prediction module 402 is configured to obtain a load impedance prediction value F according to a load impedance prediction model.

[0089] The anti-disturbance control module 403 is configured to control the system to work in a constant power stage or a constant current stage according to whether the load impedance prediction value F is in a fluctuation stable interval; if the load impedance prediction value is in the fluctuation stable interval, the system works in the constant power stage, otherwise the system works in the constant current stage.

[0090] The remote plasma source load impedance disturbance protection control device 40 provided by the embodiments of the present application adopts the same inventive concept as the remote plasma source load impedance disturbance protection control method described above, and can achieve the same beneficial effects, which will not be described here.

[0091] The technical features of the above embodiments can be combined in any way. To make the description concise, not all possible combinations of the technical features in the above embodiments are described, but as long as the combinations of the technical features do not exist contradictory, they should be considered as the scope of the present application.

[0092] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the invention patent. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this patent application should be determined by the appended claims.

Claims

1. A remote plasma source load impedance disturbance protection control method, characterized in that, include: Collect the gas flow rate Q, gas pressure P, and reaction chamber temperature T in the reaction chamber; Based on the load impedance prediction model, the predicted load impedance value F is obtained; The load impedance prediction model is as follows: Z represents the load impedance, k1, k2, and k3 are variable coefficients, and k0 is a constant term. The objective function is Q i P i , t i Z i Given n sets of data, ; The predicted load impedance F is determined by fitting variable coefficients and constant terms; Depending on whether the predicted load impedance value F is within the fluctuating and stable range, the control system operates in either the constant power stage or the constant current stage; if the predicted load impedance value is within the fluctuating and stable range, the control system operates in the constant power stage; otherwise, the control system operates in the constant current stage.

2. The method according to claim 1, characterized in that, The fitting method for the variable coefficients and constant terms includes: Construct the data matrix X and the observation vector Z; ; For the variable vector Taking the partial derivatives and setting them to zero, we obtain the standardized equation. ; Optimal solution of the variable to be solved .

3. The method according to claim 1, characterized in that, The fluctuation stability range is: , The sampling period T within a unit time T0 s The average of the multiple obtained impedance prediction values ​​F, where d is the load impedance fluctuation error.

4. The method according to claim 3, characterized in that, The constant current stage includes: Based on the resonant output current reference value and the current i obtained by sampling pri The current error e1 is obtained, and the duty cycle d is output through the current PI modulator. The PWM drive signals D1~D4 of the inverter switching transistors are output through PWM modulation.

5. The method according to claim 4, characterized in that, The constant power stage includes: Based on the output power reference value P ref The resonant output voltage V obtained by sampling loop and current i pri The actual power Po is obtained, and the output power error e0 is compared with it. After power PI modulation, the resonant output current reference value i is obtained. ref ; Use the current reference value i ref The current value i sampled by the power circuit pri After comparison, the output current error e1 is used to output the duty cycle d through the current PI modulator; The switching frequency f is obtained by modulating the current error e1 with a PI sensor and then inverting it. s The PWM drive signals D1~D4 of the inverter switching transistors are output through PWM modulation.

6. A remote plasma source load impedance disturbance protection and control device, characterized in that, include: The data acquisition module is used to acquire the gas flow rate Q, gas pressure P, and reaction chamber temperature T of the reaction chamber. The impedance prediction module is used to obtain the predicted load impedance value F based on the load impedance prediction model; the load impedance prediction model is... Z represents the load impedance, k1, k2, and k3 are variable coefficients, and k0 is a constant term. The objective function is Q i P i , t i Z i Given n sets of data, ; The predicted load impedance F is determined by fitting variable coefficients and constant terms; The disturbance rejection control module is used to control the system to operate in a constant power stage or a constant current stage based on whether the predicted load impedance value F is within the fluctuation and stability range; if the predicted load impedance value is within the fluctuation and stability range, the control system operates in the constant power stage, otherwise the control system operates in the constant current stage.

Images