Ion implanter control system and hydrogen-helium coaxial ion implanter

Abstract

The invention relates to an ion implanter control system and an ion implanter. The ion implanter control system comprises an upper computer, a lower computer, a gas flow controller and a sensor system, wherein the sensor system is arranged on the ion implanter, the gas flow controllers are connected to a gas inlet pipeline of an ion source of the ion implanter, and the lower computer is connected with the gas flow controllers and the sensor system; the sensor system is used for detecting real-time state parameters of the ion implanter and sending the real-time state parameters to the lower computer; the gas flow controller is used for detecting flow parameters of gas and uploading the flow parameters to the lower computer; the lower computer uploads real-time state parameters of the ion implanter detected by the sensor system to the upper computer, receives control parameters of various gas inflow issued by the upper computer, and adjusts the gas inflow of each gas flow controller according to the control parameters; the upper computer is configured to adjust the air inflow of the ion source; according to the technical scheme, the control efficiency is improved, and the control accuracy is improved.

Description

Technical Field

[0001] The present application relates to the technical field of ion implanters, and particularly to an ion implanter control system and a hydrogen-helium coaxial ion implanter. Background Art

[0002] Ion implanters have a very wide range of application fields and play an important role in many fields such as the nuclear energy field, new energy field, medical field, and environmental protection field; for example, a hydrogen-helium coaxial ion implanter can simultaneously generate two ion beams, namely hydrogen molecular ions H2+ and helium ions He+, in the research and testing of nuclear energy engineering materials, the development of surface materials for space vehicles, ion beam material modification, the research of nuclear medicine and radiation medicine, the research and development of catalysts in the energy field, etc., covering key technologies in the new energy field, medical treatment, environmental protection and other fields.

[0003] Due to the complex ionization process of the ion source, the ionization effect of ions is affected by multiple factors such as the gas concentration, temperature, and electric field in the cavity, and the vacuum effect of the vacuum system is also affected by factors such as the temperature and humidity of the external environment. Therefore, at different time periods, the same gas inlet ratio will produce different ratios of ion beam currents; thus, it is necessary to adjust the gas inlet ratio of the ion source to ensure that the ion implanter can output a stable ratio of ion beams. For example, the ion source used in a hydrogen-helium coaxial ion implanter has relatively obvious characteristics compared with a common ion source. For example, the ionization process of the ion source is complex, and the ionization effect of ions is affected by multiple factors such as the gas concentration, temperature, and electric field in the cavity, making it difficult to accurately establish a corresponding model to describe the ionization process; the ionization of the ion source is easily affected by the external environment, and the ionization effect is greatly affected by the vacuum degree and impurity gas in the ion source cavity, and the vacuum effect of the vacuum system is also affected by factors such as the temperature and humidity of the external environment. Therefore, at different time periods, the same hydrogen-helium gas inlet ratio will produce different ratios of beam currents; the ionization effect is different during the long-term operation of the ion source. The ion source will consume the impurity gas in the cavity while generating ions, and the impurity gas attached to the inner wall of the cavity will also be released during the operation of the ion source, affecting the ionization effect of the ion source; some ion implanters cannot detect the intensity of the beam current and the ratio of each ion in the beam current while outputting the beam to other devices. Therefore, the ion implanter cannot continuously feedback the beam current situation, which also brings certain difficulties to beam tuning. When the ionization effect of the ion source changes, it will cause the intensity of the ion beam current generated by the implanter and the ratio of each ion in the beam current to change.

[0004] At present, for the function of adjusting the beam current in an ion implanter, a manual control method is mainly adopted. According to the beam current intensity and ion ratio feedback by the ion implanter control system, the intake of the ion source is adjusted by combining the operator's own experience. This control method has low efficiency and difficult-to-control effects. It is necessary to arrange operators to adjust the intake of the ion source for a long time. Moreover, the effect of manual adjustment is related to the operator's experience, and certain experience accumulation is required to achieve beam current adjustment quickly. In addition, during the process of measuring the beam current intensity and the ratio of each ion in the beam current of some implanters, the measurement and control equipment will block the transmission of the plasma beam to the subsequent device, resulting in the compression of the time for the implanter to transmit the beam to the subsequent device, thus reducing the efficiency of the device to output the plasma beam. Summary of the Invention

[0005] An object of the present application is to solve one of the above technical defects, and provide an ion implanter control system and a hydrogen-helium coaxial ion implanter to improve the control efficiency and accuracy of the ion implanter.

[0006] An ion implanter control system includes: a host computer, a slave computer, a gas flow controller, and a sensor system; wherein, the sensor system is arranged on the ion implanter, the gas flow controller is connected to the intake pipeline of the ion source of the ion implanter, and the slave computer is connected to each gas flow controller and the sensor system;

[0007] The sensor system is used to detect the real-time state parameters of the ion implanter and send them to the slave computer;

[0008] The slave computer uploads the real-time state parameters of the ion implanter detected by the sensor system to the host computer, and receives the control parameters of the intake of various gases issued by the host computer, and adjusts the intake of each gas flow controller according to the control parameters;

[0009] The host computer is configured to obtain the expected beam current parameters of the ion implanter and the real-time state parameters uploaded by the slave computer, calculate the intake of various gases of the ion source at the next moment by calling a pre-trained neural network according to the real-time state parameters and the expected beam current parameters of the ion implanter at the current moment, and adjust the intake of the ion source at the next moment, so that the output beam current parameters of the ion implanter meet the requirements of the expected beam current parameters.

[0010] In one embodiment, the sensor system includes: a vacuum gauge arranged in the cavity of the ion implanter and a beam current sensor arranged on the ion beam output component; wherein, the vacuum gauge detects the vacuum degree in the cavity of the ion implanter, and the beam current sensor detects the beam current intensity parameter of the ion beam.

[0011] The lower computer is also connected to the pneumatic valve of the ion implanter, the power switch, the molecular pump switch, and the chuck current limiter motor of the ion implanter.

[0012] In one embodiment, the upper computer includes: a server and a client terminal; wherein, the server communicates with the lower computer and the client terminal respectively through a network;

[0013] The client terminal accesses the automatic beam tuning program on the server, receives the desired ion beam current parameters input by the user through the user operation interface of the automatic beam tuning program, and displays various state parameters of the ion implanter.

[0014] In one embodiment, the lower computer includes: a programmable logic controller and a gateway module; wherein, the programmable logic controller is used to convert analog signals and digital signals; the gateway module is connected to the network by wire or wirelessly and conducts data communication with the server.

[0015] In one embodiment, it is characterized in that the upper computer is configured to obtain the desired beam current parameters output by the ion implanter input through the user interface, control the ion implanter to output an ion beam current; detect the initial state data of the current operating state of the ion implanter, input the desired beam current parameters and the initial state data into a prediction neural network to obtain the initial intake ratio of various gases of the ion source, and adjust the intake amount of various gases of the ion source according to the initial intake ratio; detect the real-time state data of the ion implanter, when the output beam current parameters of the ion implanter do not meet the requirements of the desired beam current parameters, calculate the state change data of the ion implanter at the current moment according to the real-time state data, input the state change data and the desired beam current parameters into a correction neural network to calculate the intake amount correction value of various gases of the ion source at the next moment, calculate the intake amount of various gases of the ion source according to the intake amount correction value, and adjust the intake amount of the ion source at the next moment so that the output beam current parameters of the ion implanter meet the requirements of the desired beam current parameters.

[0016] In one embodiment, the desired beam current parameters include the desired beam current intensity and / or the desired beam current ion ratio;

[0017] The initial state data includes the air pressure in the cavity of the ion implanter, the intensity of the current beam, and / or the current beam ion ratio;

[0018] The state change data includes: the change amount of the air pressure in the cavity of the ion implanter, the change amount of the beam current intensity, and / or the change amount of the beam current ion ratio.

[0019] In one embodiment, the host computer is further configured to control the ion implanter to output an ion beam current with corresponding parameters and regularly detect whether the output beam current parameters of the ion beam generated by the ion implanter meet the requirements of the desired beam current parameters when the output beam current parameters of the ion implanter meet the requirements of the desired beam current parameters.

[0020] In one embodiment, when calculating the intake air volume correction values of various gases of the ion source at the next moment, the host computer is configured to obtain the state change data of the ion implanter in multiple time periods as the state change data at the current moment according to the recorded real-time state data, and input the multiple state change data into the input layer of the correction neural network to calculate the intake air volume correction values of various gases of the ion source at the next moment.

[0021] In one embodiment, the host computer is further configured to determine the total intake air volume of the ion source according to the desired beam current intensity; input the desired beam current ion ratio and the initial state data into the prediction neural network to obtain the initial intake air ratio of various gases of the ion source and send it to the gas regulation program; calculate the intake air volume of various gases of the ion source according to the initial intake air ratio and the total intake air volume through the gas regulation program, and control the intake air volume of various gases.

[0022] In one embodiment, the host computer uses the following calculation formula when calculating the intake air volume of various gases of the ion source:

[0023] Q 1t =Q 1t-1 +(k S +k W )Δr e Q

[0024] In the formula, Q 1t is the intake air volume of the gas at the current moment, Q 1t-1 is the intake air volume of the gas at the previous moment, Q is the total intake air volume of all gases of the ion source, Δr e represents the change amount of the ratio of the corresponding gas ions at the next moment output by the correction neural network; k S is the corresponding relationship between the desired gas intake ratio and the desired beam current ion ratio, and k W is the corresponding relationship between the change amount of the gas intake ratio and the change amount of the beam current ion ratio.

[0025] In one embodiment, the prediction neural network includes an input layer, a hidden layer, and an output layer; among them, the parameters in the hidden layer are trained by the historical data of the operation of the ion implanter; the input layer accesses the initial state data and the desired beam current parameters, and the output layer outputs the intake air ratio of each gas in the ion source calculated by the hidden layer;

[0026] The modified neural network includes an input layer, a hidden layer, and an output layer. Among them, the parameters in the hidden layer are obtained by training with the historical data of the ion implanter. The input layer accesses the state change data and the desired beam parameters generated at the current moment after the initial intake ratio adjustment of the ion implanter, and the output layer outputs the correction values of the intake amounts of various gases of the ion source at the next moment calculated by the hidden layer.

[0027] In one embodiment, the host computer is further configured to establish a prediction neural network; obtain the intake amounts of various gases of the ion implanter recorded in history at different time periods and the ion ratios of various ions in the ion beam; use the state parameters and ion ratios of the ion implanter as inputs and the intake ratios of various gases as outputs to train the prediction neural network.

[0028] In one embodiment, the host computer is further configured to establish a modified neural network; use the ion ratios and the real-time state data of the ion implanter as inputs and the intake ratios of various gases as outputs; use the intake ratios of various gases at the current moment as the prediction results of the intake ratios of the previous moment for the next moment to train the neural network, and update the parameters of the modified neural network by backpropagation according to the error between the prediction results and the actual data.

[0029] In one embodiment, the host computer is further configured to:

[0030] Establish a prediction neural network;

[0031] Obtain the intake amounts of various gases of the ion implanter recorded in history at different time periods and the ion ratios of various ions in the ion beam;

[0032] Use the state parameters and ion ratios of the ion implanter as inputs and the intake ratios of various gases as outputs to train the prediction neural network.

[0033] In one embodiment, the host computer is further configured to:

[0034] Establish a modified neural network;

[0035] Use the ion ratios and the real-time state data of the ion implanter as inputs and the intake ratios of various gases as outputs;

[0036] Use the intake ratios of various gases at the current moment as the prediction results of the intake ratios of the previous moment for the next moment to train the neural network, and update the parameters of the modified neural network by backpropagation according to the error between the prediction results and the actual data

[0037] A hydrogen-helium coaxial ion implanter, comprising: an ion source, an ion transport component, and an ion beam output component connected in sequence; wherein, the ion source accesses a variety of gases and outputs ions, the ion transport component transports the ions to the ion beam output component, and the ion beam output component outputs an ion beam current;

[0038] The ion implanter is controlled by the ion implanter control system described above.

[0039] In one embodiment, the gases input by the ion source include hydrogen and helium; the ion beam output component outputs hydrogen ion and helium ion beam currents;

[0040] The gas flow controller includes a hydrogen flowmeter and a helium flowmeter.

[0041] In one embodiment, the ion transport component includes a magnetic analyzer, and the ion beam output component includes an accelerator.

[0042] The technical solution of the above embodiment collects the state parameters of the ion implanter through the lower computer, obtains the desired beam current parameters by the upper computer, combines the real-time state parameters uploaded by the lower computer, calls the pre-trained neural network to calculate the intake amounts of various gases of the ion source at the next moment, and adjusts the intake amounts of the ion source at the next moment, so that the output beam current parameters of the ion implanter meet the requirements of the desired beam current parameters; this technical solution improves the control efficiency, enhances the control accuracy, and also increases the time for the ion implanter to transmit the beam current to the subsequent equipment.

[0043] Furthermore, an architecture for realizing intelligent control functions and network-based remote control is provided. By running a neural network model on a server, a powerful intelligent ability is provided, enabling multiple ion implanters to be remotely controlled by a client terminal through the server, improving the convenience of control operations, enhancing the control accuracy of the ion implanter, and also reducing the control cost.

[0044] Furthermore, the initial intake ratios of various gases of the ion source are calculated using a prediction neural network with the desired beam current parameters and initial state data, and the intake amount correction values of various gases of the ion source at the next moment are calculated using a correction neural network with the state change data of the ion implanter at the current moment, and the intake amounts of various gases of the ion source are corrected; greatly improving the control efficiency and accuracy.

[0045] Additional aspects and advantages of the present application will be given in part in the following description, which will become apparent from the following description, or will be understood through the practice of the present application. Description of the Drawings

[0046] The above and / or additional aspects and advantages of the present application will become apparent and be readily understood from the following description of embodiments in conjunction with the accompanying drawings, in which:

[0047] Figure 1 is a block diagram of the control system structure of an ion implanter in an embodiment;

[0048] Figure 2 is a block diagram of the control system structure of an ion implanter in another embodiment;

[0049] Figure 3 is a block diagram of the control system structure of an ion implanter in yet another embodiment;

[0050] Figure 4 is a structure diagram of the server system in an embodiment;

[0051] Figure 5 is a beam current regulation block diagram of an exemplary ion implanter;

[0052] Figure 6 is a monitoring diagram of beam current stability in an example;

[0053] Figure 7 is a beam current control flow chart in an example;

[0054] Figure 8 is a comparison diagram of training effects at different time steps in an example;

[0055] Figure 9 is a comparison diagram of the training processes of LSTM and ANN;

[0056] Figure 10 is a comparison diagram of the prediction results of LSTM and ANN for recent data;

[0057] Figure 11 is a comparison diagram of the prediction results of LSTM and ANN for long-term data;

[0058] Figure 12 is an architecture diagram of an ion implanter and its control system in an application example. Detailed implementation manners

[0059] Embodiments of the present application will be described in detail below. Examples of the embodiments are shown in the accompanying drawings, where the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below by referring to the accompanying drawings are exemplary only for explaining the present application and should not be construed as limiting the present application.

[0060] Those skilled in the art can understand that, unless specifically stated otherwise, the singular forms "a", "an", "the", and "said" used herein may also include the plural forms. It should be further understood that the term "comprising" used in the specification of this application means the presence of the stated features, integers, steps, operations, but does not exclude the presence or addition of one or more other features, integers, steps, operations.

[0061] Those skilled in the art can understand that, unless otherwise defined, all terms used herein (including technical terms and scientific terms) have the same meaning as the general understanding of those of ordinary skill in the art to which this application belongs. It should also be understood that terms such as those defined in a general dictionary should be understood to have a meaning consistent with the meaning in the context of the prior art, and will not be interpreted with an idealized or overly formal meaning unless specifically defined as here.

[0062] In an ion implanter, since the plasma generation of the ion source is a highly non-linear and complex process affected by many parameters, there are controllable parameters such as the intake amounts of hydrogen and helium gases, the power of the microwave machine, etc., and there are also uncontrollable parameters such as the environmental temperature, the change in the supply pressure of the gas cylinder; at different time periods, the change of these parameters will lead to the change of the state of the plasma, and different beam currents and ion ratios will also appear when the set value of the intake amount remains unchanged, which makes the beam tuning of the ion implanter a difficult and time-consuming task.

[0063] In order to be able to obtain the desired ion beam quickly and stably, the ion implanter control system provided by the technology of this application can calculate and automatically adjust the intake amounts of various gases in the ion source of the ion implanter according to the current beam state, the desired beam current intensity, and the desired ion ratio through a pre-trained neural network, control the ion implanter to generate an ion beam with the expected beam current intensity size and beam ion ratio, and realize the automatic adjustment function of the ion implanter output ion beam.

[0064] As Figure 1 shown, Figure 1 is a structural block diagram of an ion implanter control system of an embodiment, including: a host computer, a slave computer, a sensor system, and a plurality of gas flow controllers arranged at the gas inlet of the ion source; the host computer can be a computer installed with a control software system, the slave computer can adopt a PLC (Programmable Logic Controller), and the programmable logic controller can implement the control function and realize the conversion between analog signals and digital signals. The ion implanter controls the intake amounts of various gases in the ion source through the host computer and the slave computer.

[0065] Exemplarily, the host computer has a control program built in, which can be defined as an automatic beam adjustment program; the sensor system is provided on the ion implanter, and it can be various components such as sensors for detection; the gas flow controller is connected to the intake pipeline of the ion source of the ion implanter, and it can implement the functions of a gas flow meter and flow control; such as Figure 1 In [reference], gas flow controllers 1 to n, where n≥2, are respectively connected to n intake pipelines of the ion source of the ion implanter. The slave computer is connected to each gas flow controller and the sensor system; the sensor system is used to detect the real-time state parameters of the ion implanter and send them to the slave computer; each gas flow controller corresponds to a kind of gas. The slave computer uploads the real-time state parameters of the ion implanter detected by the sensor system to the host computer, and receives the control parameters of the intake amounts of various gases sent by the host computer, and adjusts the intake amounts of each gas flow controller according to the control parameters.

[0066] The host computer is configured to obtain the expected beam current parameters of the ion implanter and the real-time state parameters uploaded by the slave computer, calculate the intake amounts of various gases of the ion source at the next moment by calling a pre-trained neural network according to the real-time state parameters and the expected beam current parameters of the ion implanter at the current moment, and adjust the intake amounts of the ion source at the next moment, so that the output beam current parameters of the ion implanter meet the requirements of the expected beam current parameters.

[0067] Exemplarily, during use, the operator inputs the expected beam current parameters on the user interface of the host computer. The host computer obtains the expected beam current parameters of the ion implanter and controls the ion implanter to output an ion beam. The slave computer can detect the real-time state parameters such as the intake amount, vacuum degree, and beam current intensity parameters of the ion implanter through the gas flow controller and the sensor system, convert them from analog signals to digital signals and then send them to the host computer. After the host computer receives the real-time state parameters of the ion implanter uploaded by the slave computer, it calculates the intake amounts of various gases of the ion source at the next moment as control parameters by calling a pre-trained neural network according to the real-time state parameters and the expected beam current parameters of the ion implanter at the current moment, and issues them to the slave computer at the next moment. The slave computer receives the control parameters issued by the host computer and adjusts the intake amounts of the ion source according to the control parameters, controls the hydrogen flow meter and the helium flow meter to adjust the intake amounts of hydrogen and helium respectively, so that the output beam current parameters of the ion implanter meet the requirements of the expected beam current parameters, thereby realizing the output of the expected ion beam.

[0068] In one embodiment, referring to Figure 2 as shown Figure 2It is a block diagram of the control system structure of an ion implanter for another embodiment. The ion implanter may include an ion source, a transmission component, an ion beam output component, etc. The ion source accesses various reaction gases and outputs ions, which are transmitted to the ion beam output component through the transmission component and finally outputs an accelerated ion beam. The sensor system may include a vacuum gauge disposed in the cavity of the ion implanter and a beam current sensor disposed on the ion beam output component of the ion implanter. The vacuum gauge and the beam current sensor are respectively connected to the lower computer. The vacuum gauge is used to detect the vacuum degree in the cavity of the ion implanter, and the beam current sensor is used to detect the beam current intensity parameter of the ion beam. The vacuum gauge uploads the detected vacuum degree to the lower computer, and the lower computer uploads it to the upper computer. The beam current sensor uploads the detected beam current intensity parameter of the ion beam to the lower computer, and the lower computer uploads it to the upper computer. A control program is built in the upper computer. Through the control program, the upper computer outputs control parameters to the lower computer, and the lower computer controls the gas flow controller to respectively control the intake of various gases.

[0069] In one embodiment, referring to Figure 3 as shown, Figure 3 It is a block diagram of the control system structure of an ion implanter for yet another embodiment. The lower computer is also connected to the pneumatic valve, power switch, molecular pump switch, beam blocker motor, etc. of the ion implanter, reads the status parameters of these ion implanter components and controls them. Among them, the lower computer can control the power switch of the ion source implanter power supply, control the output voltage or current of the ion implanter, and read the voltage and current data in real time. The beam blocker is used to block a part of the ions in the ion beam to play a screening role. The lower computer controls the position of the beam blocker through the beam blocker motor and reads the position data of the beam blocker motor in real time.

[0070] Exemplarily, in a hydrogen-helium coaxial ion implanter, it can be divided into an ion source component, a transmission component, and an ion beam current detection component according to functions. The ion source component can include various ion sources that generate plasma. The transmission component can include a mass analyzer. The ion beam current detection component can include a Faraday cup for detecting the beam current intensity. In the ion source component, a mixed ion beam is generated by feeding different flow rates of hydrogen and helium. After separation and purification by the transmission component, an H2+ and He+ mixed ion beam with an impurity ion content of less than 2% is transmitted to the Faraday cup, the total beam current is recorded, and by moving the beam blocker plate, the ratio of the H2+ and He+ ion beams is calculated. By obtaining data such as the total beam current of the ion implanter, the vacuum degree in the ion source extraction area, the intake of hydrogen and helium, and the microwave power, according to the desired beam current intensity and ion ratio of the user, the corresponding intake of hydrogen and helium in the current ion implanter state is calculated, and the intake of the two gases of hydrogen and helium input to the ion implanter is adjusted, thereby controlling the intensity of the H2+ and He+ ion beams output.

[0071] In the solution of the above embodiment, based on the provided architecture of the ion implanter control device, the ion implanter can achieve automatic control. The host computer provides a parameter calculation and output function, outputs control parameters to the slave computer, and the slave computer executes the control of the gas intake volume of the ion implanter, thereby adjusting the output beam current of the ion beam, improving the control efficiency, enhancing the control accuracy of the ion implanter, and increasing the time for the ion implanter to transmit the beam current to the subsequent equipment.

[0072] In one embodiment, referring to Figure 4 as shown, Figure 4 is a structural diagram of the server system in one embodiment. To improve the accuracy of the host computer in processing control parameters, a server can be built to provide corresponding computing capabilities. The server consists of at least one computer; an automatic beam adjustment program is configured on the server to provide background services. The programmable logic controller communicates with the server through the gateway module, such as gateway modules 1 to m and ion implanters 1 to m in the figure. Thus, the ion implanter can access the network through the programmable logic controller and use the gateway module to communicate with the server. The gateway module is connected to the network by wired or wireless means. The server can also communicate with the client terminal through the network, enabling various client terminals to access and control the ion implanter, which is convenient for operation. A powerful neural network model can be deployed and run on the server to provide intelligent capabilities, improving the control processing efficiency of the ion implanter and also enhancing the control accuracy. Based on the above automatic control architecture, the server can provide control computing capabilities for multiple ion implanters, such as ion implanters 1 to m in the figure. The operator uses the client terminal to input the desired ion beam current parameters through the user operation interface and can then call the software program and neural network model of the server to calculate the control parameters of the intake volume of various gases. At the same time, the client terminal can also display the real-time status parameters of the ion implanter to the operator through the user operation interface.

[0073] As in the technical solution of the above embodiment, an architecture based on network remote control, automatic control, and intelligent control is provided, enabling multiple ion implanters to be remotely called by the client terminal to use the automatic software program and powerful intelligent neural network model of the server. The operator can input the desired beam current parameters and view the real-time status parameters of the ion implanter through multiple client terminals, such as client terminals 1 to k in the figure. These client terminals can be terminal devices such as smartphones, laptops, and personal computers to achieve the control interaction function, improving the convenience of control operations. At the same time, the server can provide control for multiple ion implanters, reducing the control cost.

[0074] In one embodiment, in the ion implanter control system of the present application, when the host computer uses a neural network model to calculate the intake ratio, it can use a prediction neural network to predict the initial intake ratio for initial adjustment and use a correction neural network to perform real-time error correction adjustment.

[0075] For the initial adjustment process, it may include the following:

[0076] According to the desired beam parameters output by the ion implanter input; control the ion implanter to output an ion beam current, and detect the initial state data of the current operating state of the ion implanter; input the desired beam parameters and the initial state data into the prediction neural network to obtain the initial intake ratio of various gases of the ion source, and adjust the intake amount of various gases of the ion source according to the initial intake ratio.

[0077] Specifically, after the ion implanter is normally started, an ion beam current will be generated after the ion implanter is normally started; an operator can set the desired beam parameters, such as the desired beam intensity and the desired beam ion ratio; at the same time, the lower computer receives the initial state data detected by each sensor and sends it to the host computer. The initial state data may include the air pressure in the cavity of the ion implanter, the intensity of the current beam, the current beam ion ratio, etc.; the host computer can input the desired beam intensity, the desired beam ion ratio, the air pressure in the cavity of the ion implanter, the intensity of the current beam, the current beam ion ratio, etc. into the pre-trained prediction neural network to obtain the initial intake ratio of various gases and send it to the lower computer, and then the lower computer adjusts the intake amount of various gases of the ion source according to the initial intake ratio, so that the state parameters of the ion beam current output by the ion implanter reach the specified range.

[0078] Exemplarily, when adjusting the intake amount of various gases of the ion source, the host computer can determine the total intake amount of the ion source according to the desired beam intensity, input the desired beam ion ratio and the initial state data into the prediction neural network to obtain the initial intake ratio of various gases of the ion source and send it to the gas regulation program; wherein, the initial state data includes data reflecting the current state information of the ion implanter, such as the air pressure in the cavity of the ion implanter, the intensity of the current beam, the current beam ion ratio, etc.; the gas regulation program calculates the intake amount of various gases of the ion source according to the initial intake ratio and the total intake amount, and issues it to the lower computer to control the intake amount of various gases, completing the preliminary adjustment process.

[0079] According to the technical solution of the above embodiment, after the operator inputs the desired beam parameters such as the desired beam current intensity and the desired ion ratio of the ion implanter, the host computer inputs the desired beam parameters into the input layer of the prediction neural network, and at the same time transmits the detected initial state data into the input layer of the prediction neural network. The prediction neural network automatically calculates the intake ratios of various gases, and thus can calculate the intake amounts of various gases. The intake amounts of the ion source are adjusted by the lower computer, improving the control efficiency and accuracy.

[0080] For the real-time error correction adjustment process, it may include the following:

[0081] Detect the real-time state data of the ion implanter. The state change data of the ion implanter may include the change amount of the air pressure in the cavity of the ion implanter, the change amount of the beam current intensity, the change amount of the beam ion ratio, etc.; and determine whether the output beam parameters of the ion implanter meet the requirements of the desired beam parameters according to the real-time state data. If the output beam parameters of the ion implanter do not meet the requirements of the desired beam parameters, calculate the state change data of the ion implanter at the current moment according to the real-time state data, input the state change data and the desired beam parameters into the correction neural network to calculate the correction values of the intake amounts of various gases of the ion source at the next moment, calculate the intake amounts of various gases of the ion source according to the intake amount correction values, and adjust the intake amounts of the ion source at the next moment so that the output beam parameters of the ion implanter meet the requirements of the desired beam parameters. If the output beam parameters of the ion implanter meet the requirements of the desired beam parameters, control the ion implanter to output the ion beam with the corresponding parameters, and regularly detect whether the output beam parameters of the ion beam generated by the ion implanter meet the requirements of the desired beam parameters.

[0082] Specifically, after initially adjusting the ion implanter, the lower computer can receive the real-time status data of the ion implanter detected by various sensors and send it to the upper computer. The upper computer can determine whether the output beam parameters of the ion implanter meet the requirements of the expected beam parameters based on the real-time beam intensity and beam ion ratio of the output ion beam. When making a comparison and judgment, it can compare whether the difference between the actual beam intensity and the expected beam intensity reaches within the set threshold range, and whether the difference between the actual beam ion ratio and the expected ion ratio reaches within the set threshold range. By detecting the actual beam parameters, if the requirements of the expected beam parameters are met, the current output ion beam of the ion implanter is maintained, and at the same time, it enters the monitoring state to regularly detect whether the ion beam generated by the ion implanter meets the requirements of the expected beam parameters; when it does not meet the requirements of the expected beam parameters, it enters the error correction process. Using the pre-trained correction neural network, the state change data and the expected beam parameters are input into the correction neural network to calculate the correction values of the intake air volume of various gases of the ion source at the next moment, and then the intake air volume is corrected and adjusted so that the output beam parameters of the ion implanter meet the requirements of the expected beam parameters and approach the beam state expected by the user.

[0083] As the technical solution of the above embodiment, using the correction neural network to calculate the intake air volume ratio and perform error correction adjustment realizes the automatic control function, improves the efficiency of the ion implanter outputting the plasma beam, improves the control accuracy, and can increase the time for the ion implanter to transmit the beam to the subsequent equipment.

[0084] Regarding the functions completed by the upper computer and the lower computer during the control process, more embodiments are described below.

[0085] In one embodiment, the prediction neural network may include: an input layer, a hidden layer, and an output layer; among them, the parameters in the hidden layer are obtained by training with the historical data of the operation of the ion implanter; the input layer accesses the initial state data and the expected beam parameters, and the output layer outputs the intake air ratio of various gases in the ion source calculated by the hidden layer. The process of training the prediction neural network may include the following:

[0086] Establish a prediction neural network; obtain the intake air volume of various gases of the ion implanter in different time periods and the ion ratio of various ions in the ion beam recorded in history; use the state parameters and ion ratio of the ion implanter as the input and the intake air ratio of various gases as the output to train the prediction neural network.

[0087] Among them, the prediction neural network and the correction neural network can use LSTM (Long Short-Term Memory). The LSTM consists of multiple neurons, and each neuron is composed of a forget gate, an input gate, an output gate, and a cell state update. The data information selected by the input gate will be recorded, the data selected by the forget gate will be discarded, the output gate controls the output of the neuron, and the cell state update is used to transfer the data information recorded at the current moment to the neurons at the next moment.

[0088] In one embodiment, the correction neural network may include an input layer, a hidden layer, and an output layer. Among them, the parameters in the hidden layer are obtained by training with the historical data of the operation of the ion implanter. The input layer accesses the state change data and the desired beam current parameters generated at the current moment after the ion implanter completes the initial intake ratio adjustment. The output layer outputs the correction values of the intake amounts of various gases of the ion source at the next moment calculated by the hidden layer. The process of training the correction neural network may include the following:

[0089] Establish a correction neural network; use the ion ratio and the real-time state data of the ion implanter as the input and the intake ratios of various gases as the output; use the intake ratios of various gases at the current moment as the prediction results of the intake ratios for the next moment from the previous moment to train the neural network, and update the parameters of the correction neural network by backpropagation according to the error between the prediction results and the actual data.

[0090] In one embodiment, when calculating the correction values of the intake amounts of various gases of the ion source at the next moment through the correction neural network, the state change data of the ion implanter in multiple time periods can be obtained according to the recorded real-time state data as the state change data at the current moment, and the multiple state change data are input into the input layer of the correction neural network to calculate the correction values of the intake amounts of various gases of the ion source at the next moment.

[0091] In one embodiment, calculating the intake amounts of various gases of the ion source according to the intake amount correction values includes the following formula:

[0092] Q 1t =Q 1t-1 +(k S +k W )Δr e Q

[0093] In the formula, is the intake amount of the gas at the current moment, is the intake amount of the gas at the previous moment, Q is the total intake amount of all gases of the ion source, Δr e represents the change amount of the ratio of the corresponding gas ions at the next moment output by the correction neural network; k SFor the correspondence relationship between the desired gas inlet ratio and the desired beam ion ratio, k W It is the correspondence relationship between the change in the gas inlet ratio and the change in the beam ion ratio.

[0094] In the solution of the above embodiment, when the neural network continuously predicts the gas inlet ratio of the ion source at the next moment at different moments, the ion implanter generates state change data at the corresponding moment. The neural network records the state data at different moments and predicts the gas inlet ratio at the next moment according to the state data of the previous multiple moments, thereby improving the prediction accuracy.

[0095] In one embodiment, when detecting the state data of the ion implanter, the real-time state data of the ion implanter can be obtained through a gas flow controller, a vacuum gauge, a beam sensor (Faraday cup), etc. installed on the ion implanter. The analog signal of the sensor is converted into a digital signal by the lower computer and transmitted to the neural network through the channel of EPICS for application.

[0096] In one embodiment, the beam regulation process of the ion source of the ion implanter can be divided into multiple time segments. The parameters related to beam regulation in each time segment can include: hydrogen gas inlet volume x1, helium gas inlet volume x2, pressure at the ion source x3, ion beam intensity x4, actual current of the extraction high voltage x5, ion ratio of the ion beam x6, change in the desired beam ion ratio in the next time period x7; the above data records the state information of the ion implanter at different time periods. Transmitting the state information at these different time periods into the neural network can predict the relationship between the change in the hydrogen gas inlet ratio and the change in the beam ion ratio at the next moment.

[0097] As Figure 5 shown, Figure 5 is a block diagram of the beam regulation of an example ion implanter; among them, the control system will record the parameters related to beam regulation in the time periods T1, T2, …… Tn, Tn+1. These data record the state information of the ion implanter at different time periods. Transmitting the information at these different time periods into the neural network to calculate the correspondence relationship k between the change in the gas inlet ratio and the change in the beam ion ratio at the next moment. The calculation formula of k can be expressed as: k = f lstm (X); where k maps the change in the gas inlet ratio to the change in the gas ion ratio, and f lstm is a pre-trained neural network model, and X is a two-dimensional array containing the parameters related to beam regulation in different time segments.

[0098] In one embodiment, when adjusting the beam current of an ion implanter, the proportion of ions generated at the ion source is related to the proportion of gases. Taking hydrogen as an example, the prediction result of hydrogen by a prediction neural network can be expressed as:

[0099] Q 1 = Q 1S + Q 1W

[0100] Where Q 1 is the predicted hydrogen intake, Q 1S is the accurate hydrogen intake corresponding to the desired beam current generated by the implanter, and Q 1W is the prediction error. Directly predicting the hydrogen intake will result in an error. It is corrected through a correction neural network to calculate the hydrogen intake at the next moment, including the following:

[0101]

[0102] Where r q is the proportion of hydrogen intake in the total intake, r e is the intake proportion of hydrogen ions in the total intake, Q is the total intake of hydrogen and helium, is the hydrogen intake at the current moment, is the hydrogen intake at the previous moment. Among them, Since the proportion of hydrogen intake increases, the proportion of hydrogen ions will also increase. k is used to map the change in the proportion of hydrogen intake and the change in the proportion of hydrogen ions; comparing the desired ion proportion at the next moment with the ion proportion at the current moment, the change in the ion proportion at the next moment Δr e is obtained. Through k mapping, the change in the proportion of hydrogen intake at the next moment is obtained. Finally, based on the change in the current hydrogen intake proportion, the hydrogen intake at the current moment is calculated Since the k value is predicted by a correction neural network, an error k W will be generated, resulting in an error in the prediction of the hydrogen intake. The smaller the value of Δr e , the smaller the error value of the predicted hydrogen intake; during the beam current adjustment process, by predicting and modifying the intake of hydrogen and helium multiple times, the desired beam ion proportion is continuously approached, thereby reducing the value of Δr e , making the prediction result more accurate.

[0103] In one embodiment, when training a neural network, actual data generated by an ion implanter can be used for training. Taking a hydrogen-helium coaxial ion implanter as an example, the ion implanter modifies the intake amounts of hydrogen and helium at different time periods to obtain the ion ratio of the implanter beam current. The intake ratio of hydrogen and helium at the current moment is the prediction result of the intake ratio for the next moment from the previous moment. During the training of the neural network, the model propagates forward once based on the input at the current moment to obtain the predicted intake amounts of hydrogen and helium for the next moment, compares the predicted result with the actual data and calculates the error, and then updates the parameters of the network through backpropagation.

[0104] In one embodiment, taking a hydrogen-helium coaxial ion implanter as an example, the beam current state generated by the ion implanter is related to the intake amounts of hydrogen and helium input at the ion source. Hydrogen and helium can be expressed as:

[0105] Q 1 = r q Q(1)

[0106] Q 2 =(1 - r q )Q(2)

[0107] Where Q 1 is the intake amount of hydrogen, Q 2 is the intake amount of helium, r q is the intake ratio of hydrogen in the total intake, Q is the total intake of the ion source. By the value of r q and the total intake Q, Q 1 and Q 2 are calculated to realize the calculation of the intake amounts of hydrogen and helium at the ion source. The size of the total intake is determined by the desired beam current intensity. Accordingly, the following calculations are carried out:

[0108]

[0109] Where e represents the error between the desired beam current intensity and the actual beam current intensity, ΔQ is the change in the total intake. Since the total beam current intensity of the ion source is affected by the gas ratio, the higher rq is, the more sensitive the total current intensity is to the change in Q. K P , K i are defined gain coefficients. Since the beam current intensity of the ion source is affected by the gas ratio, the higher the hydrogen proportion, the more sensitive the beam current intensity is to the change in the total intake. Therefore, the calculation of K P , K i is determined by the size of the hydrogen proportion r q . C p , C i , b p1 , b p2 , b i1 , b i2 , Kd is a constant obtained empirically; when r q increases, K P and K i decrease accordingly. By adjusting the total intake air volume, the beam current intensity of the implanter can be controlled.

[0110] For example, in order to improve the operation stability of the ion source of an ion implanter, the automatic beam tuning off and on were monitored for 24 hours. As Figure 6 shown, Figure 6 is an example beam current stability monitoring graph. The ordinate in the graph is the monitored beam current intensity at the end of the ion implanter. Figure a is the monitoring graph when the automatic beam tuning is off, and Figure b is the monitoring graph when the automatic beam tuning program is on. It can be seen that when the automatic beam tuning is off, the beam current intensity of the ion implanter fluctuates to a certain extent over time. After the automatic beam tuning program is turned on, the beam current of the ion implanter is stably maintained at 20.5 μA for a long time.

[0111] The relationship between the ion ratio change and the hydrogen intake ratio change can be expressed as:

[0112] Δr q = k t Δr e (6)

[0113]

[0114] Q 1,t = Q 1,t-1 + k t Δr e Q (9)

[0115] Q 2,t = Q 1,t-1 +(1 - k t Δr e )Q (10)

[0116] where Δr q is the change amount of the hydrogen intake ratio, Δr e is the change amount of the H2+ ion ratio. At a certain moment, k t can map Δr e to Δr q , E 1 is the H2+ ion beam current intensity, E is the total beam current intensity, and the H2+ ion ratio r e is calculated. After obtaining the value of Δr q at time t, the hydrogen intake volume Q 1,t at the current moment is calculated through the Q 1,t at the previous moment. Similarly, the helium intake volume Q 2,t at the current moment can be calculated. Since kt is not a fixed value. Due to the influence of the plasma state at different time periods, k t will change. When using deep learning to predict k t there will be errors in the value. However, according to formulas (9) and (10), when Δre becomes smaller, the errors in predicting Q1,t and Q2,t also decrease. Therefore, during the automatic beam tuning process, by adjusting the intake amounts of hydrogen and helium multiple times, the obtained ion ratio gradually approaches the expected value.

[0117] Since Δr q and Δr e have a more complex mapping relationship, simply using k t to represent the accuracy of their mapping relationship may be insufficient. Therefore, relying solely on deep learning to predict the k t value may result in deviations and lead to unsatisfactory prediction effects. Accordingly, in practical applications, the prediction of the k t value can be skipped, and deep learning can be directly used to predict Δr e through Δr q

[0118] As Figure 7 shown, Figure 7 is a flowchart of an example beam current control; when the value of Δr e is too large, deep learning is used to directly predict the Δr q at the next moment, and then the value of Δr e is recalculated. If the value of Δr e is within the error tolerance range, the value of Δr t is directly predicted through the k q value. The k t value is set based on debugging experience and is a fixed value between 0 and 1. Deep learning is mainly used to quickly reduce the range of the Δr e value. When the value of Δr e is small enough, the intake ratio is adjusted multiple times through the k t value, so that the ion ratio generated by the ion source gradually approaches the ratio expected by the user.

[0119] After completing the prediction calculation of the intake amounts of hydrogen and helium, the gas regulation program sends the calculated results to the lower computer, and the lower computer controls the gas flow meters of hydrogen and helium to adjust the gas flow rate, thereby realizing the control of the gas intake of the ion source.

[0120] ​Exemplarily, in view of the cumbersome, difficult and low degree of automation in manually adjusting the beam current for the current ion implanter, the present application proposes an automatic beam current adjustment method for an ion implanter based on an LSTM network; the LSTM network in deep learning is used to dynamically predict the change amount of the hydrogen intake ratio in the ion source of the ion implanter, and the total intake air volume is adjusted using PID control according to the prediction result, so as to realize the automatic regulation of the beam current of the ion implanter.

[0121] Exemplarily, since the state of the plasma is affected by various uncontrollable and difficult-to-measure factors, the intensity and ion ratio of the ion beam generated by the ion source change; however, the change of the ion source state is slow, and the current state of the ion source can be mapped through some parameters at the current moment, and the state of the ion source at the next moment can be predicted according to the change of the parameter values. Accordingly, in order to obtain the best intelligent beam adjustment scheme, the beam adjustment methods of ANN (Artificial Neural Network) and LSTM (Long Short-Term Memory) can be introduced in the solution of the present application. Among them, ANN predicts the output value according to the input parameters and the activation function added according to the hidden layers of different layers. The neurons of the LSTM neural network are composed of an input gate, a forget gate, an output gate, a memory unit and a hidden layer unit. In the process of adjusting the beam current of the ion implanter, corresponding data will be generated in different time segments. After being screened by the input gate, these data are stored in the memory of the LSTM network, and the LSTM network will predict the next moment Δr q value.

[0122] When using the LSTM network to predict the Δr q value, it is necessary to dynamically predict the next moment according to the change of the ion source state online, so as to predict the output results in multiple subsequent time periods. The data structure of the network in LSTM is represented as follows:

[0123]

[0124] Among them, matrix A p is the data structure input into the LSTM network at time t p , which is a two-dimensional matrix of n×m; n is the number of features representing the ion source state, and m is the step length of recording the data of the ion source state in different time periods. Among them, x(1,t p ) is the error between the actual ion ratio and the desired ion ratio of the ion source at time t p , and x(n,t p ) represents at time t pOther characteristic parameters of the ion source at a certain moment, such as data on the vacuum degree of the extraction area, the total intake air volume, the total beam current intensity, etc. During the automatic beam tuning process, according to the data matrix A in the previous m time periods p Predict t p+1 The value of Δr at a certain moment q value. After adjusting the intake air volumes of hydrogen and helium according to the prediction results, the ion implanter measures the new ion proportion, and stores the ion source state data at time t p+1 into the matrix A p+1 for the prediction of time t p+2 Since the step size m determines the prediction of the data at the next moment by the LSTM network based on the data at the previous m moments, the larger the step size, the smaller the error of the trained model. Therefore, a larger step size should be selected.

[0125] Exemplarily, as Figure 8 shown Figure 8 is a comparison chart of the training effects of different time step sizes in an example. The vertical coordinate is the model loss Loss, and the horizontal coordinate is the number of iterations Epochs. When training using data structures with step sizes of 1, 3, 5, and 7, it can be clearly seen that for each increase in the step size by one step, the data matrix input to the network needs to increase by one row, and the computational amount of the network also increases accordingly. When the step size increases from 5 to 7, although the error decreases, it gradually tends to be consistent. Therefore, considering the trade-off between the consumption of computer memory and computing power and the computing accuracy, the step size of 5 is finally selected.

[0126] The parameters input to the LSTM network include the intake air volumes of hydrogen and helium, the vacuum degree of the extraction area, the total beam current intensity output by the ion implanter, the load of the extraction high-voltage power supply, the current hydrogen intake ratio, the current beam ion ratio, and the error between the current ion ratio and the desired ratio. The number of layers of the LSTM network is 2, the number of nodes is 128, the probability of dropout is 0.2, and the number of nodes output by the last layer of the network is 32. Then, the predicted value of the hydrogen intake ratio is output through a fully connected layer. After the ion source randomly adjusts the intake air volumes of hydrogen and helium, the corresponding beam ion ratio value is generated, and the randomly adjusted data is used for the training of the model. The beam ion ratio at the next moment is used as the expected value of the beam ion ratio at the previous moment, and the LSTM network uses the data at the previous moment and the corresponding expected value to predict the hydrogen intake ratio at the next moment. The hydrogen intake ratio at the next moment in the training data is used as the true value to calculate the prediction error of the LSTM network, and finally, it is trained through the Adam optimization algorithm with a dynamically updated learning rate. To ensure the generality of the model, the data three weeks later is used to verify the prediction results of the LSTM network and calculate the prediction error.

[0127] Using an ANN and an LSTM network for comparison, the ANN has 3 hidden layers with 34, 64, and 16 nodes respectively. The probability of dropout is 0.2 for all layers, and the ReLU activation function is used between each layer of the network. Refer to Figure 9 as shown Figure 9 Figure 4 is a comparison chart of the training processes of the LSTM and the ANN; it can be seen from the figure that as the number of iterations increases, the error of the model prediction of the ANN begins to gradually increase after training for a certain number of times, while the LSTM network can still maintain the error within a small range. The LSTM network and the ANN are respectively used to predict Δrq, and the data obtained from the test a few days later and the data obtained from the test three weeks later are predicted respectively. The prediction results are as Figure 10 and Figure 11 shown. Among them, Figure 10 Figure 5 is a comparison chart of the prediction results of the LSTM and the ANN for recent data, Figure 11 Figure 6 is a comparison chart of the prediction results of the LSTM and the ANN for long-term data; the red dots in the figure are the true values of the Δr q value. Connecting the predicted values with broken lines, it can be seen that for recent data, due to the small change in the ion source state, the errors of the LSTM network and the ANN predictions are both small. For long-term data, due to the change in the ion source state, the prediction accuracy of the ANN drops significantly, while the prediction of the LSTM network still maintains a high accuracy. Therefore, the LSTM network is more suitable for use in the automatic beam adjustment of ion implanters.

[0128] For the technical solution of comparing the prediction accuracy using an ANN and an LSTM network as in the above embodiment, by combining the method of intelligent beam adjustment, the influence of parameter changes such as environmental temperature and inlet gas supply pressure on the ion beam stability is reduced, and unattended operation is achieved.

[0129] Refer to Figure 12 as shown Figure 12 Figure 7 is an architecture diagram of an ion implanter and its control system for an application example. In the software part, the host computer can run the automatic beam adjustment program and the gas regulation program of the ion implanter. The automatic beam adjustment program is configured to calculate the intake ratio and intake volume through a prediction neural network and a correction neural network. The gas regulation program is configured to adjust the intake volume of various gases in the ion source of the ion implanter to achieve the automatic adjustment function of the ion beam current output by the ion implanter. The lower computer can include a single-chip microcomputer or a PLC. The lower computer controls each component in the ion implanter and can obtain various data collected by sensors in the ion implanter; data transmission between the host computer and the lower computer is carried out through a TCP / IP channel.

[0130] Specifically, a hydrogen-helium coaxial ion implanter can be divided into an ion source component, a transmission component, and an ion beam current detection component (such as a Faraday cup) according to its functions. In the ion source component, a mixed ion beam current is generated by feeding in different flow rates of hydrogen and helium. After separation and purification by the transmission component, an H2+ and He+ mixed ion beam with an impurity ion content of less than 2% is transmitted to the Faraday cup, where the total current intensity is recorded. By moving the beam stopper plate, the ratio of the H2+ and He+ ion beams is calculated. In the control and sampling process, the PLC controls the ion implanter through analog signals and digital signals (such as RS-485), and the PLC is connected to the control program through TCP / IP; under the EPICS control framework, the server realizes the data transmission and reception process with the PLC through TCP / IP. The intelligent beam tuning software and the user interface act as clients to access the data of the server through channels under the EPICS framework to achieve control of the ion implanter from the client. The user inputs parameters such as the hydrogen and helium intake amounts and the microwave power through the control interface and transmits them to the server through Channel Access. The server modifies the corresponding parameters in the PLC through the TCP / IP network connection. The PLC controls the gas flow controllers of hydrogen and helium in the ion source to regulate the gas flow rate and simultaneously controls the microwave machine according to the parameters of the microwave machine obtained. The beam current detection component transmits the detected beam current intensity value back to the PLC. The server obtains this data through TCP / IP, differentiates the total current intensity and the calculated ion beam ratio, and displays them in the user interface. By accessing the server, data such as the total beam current intensity of the ion implanter, the vacuum degree in the ion source extraction area, the hydrogen and helium intake amounts, and the microwave power are obtained. According to the desired beam current intensity and ion ratio of the user, the corresponding hydrogen and helium intake amounts in the current state of the ion implanter are calculated and adjusted.

[0131] The automatic beam tuning program runs on a computer. After the operator inputs the desired beam parameters in the user interface, the automatic beam tuning program inputs the desired beam parameters into the input layer of the prediction neural network. The lower computer collects the status data of the ion implanter through various sensors. The collected status data is simply processed by the lower computer and then input into the automatic beam tuning program of the upper computer. The automatic beam tuning program analyzes and processes the collected status data and inputs it into the input layer of the prediction neural network. The intake ratios of various gases in the ion source are calculated through the prediction neural network and the intake amounts are calculated. The lower computer controls the intake amounts of the gas flow controllers according to the calculated intake amounts. When the ion beam output by the ion implanter changes, the automatic beam tuning program can judge and process the need to correct the ion beam. Based on the status information of the ion beam collected by the sensors uploaded by the lower computer, the correction neural network calculates the change amounts of the intake ratios of various gases in the ion source. The automatic beam tuning program calculates the intake amounts according to the change amounts of the intake ratios. The ion beam intensity and ion ratio of the ion implanter are automatically adjusted by the gas regulation program through the lower computer, and the intake amounts of the ion source are adjusted by the lower computer to complete the correction process.

[0132] In the technical solution of the above example, the operator runs the automatic beam tuning program of the ion implanter on the computer terminal. After waiting for the ion implanter to be normally turned on, the operator inputs the desired beam parameters in the user operation interface of the automatic beam tuning program. The automatic beam tuning program automatically completes the adjustment of the ion beam of the ion implanter and feeds back the situation after the adjustment of the ion beam of the implanter on the user interface, thus greatly improving the efficiency of the control operation and having high control precision, and being suitable for popularization and use in ion implanters.

[0133] In the technical solution of the above example, the operator connects to the automatic beam tuning program of the server on the client terminal. After waiting for the ion implanter to be normally turned on, the operator inputs the desired beam parameters in the user operation interface of the automatic beam tuning program. The automatic beam tuning program automatically completes the adjustment of the ion beam of the ion implanter and feeds back the situation after the adjustment of the ion beam of the implanter on the user interface, thus greatly improving the efficiency of the control operation and having high control precision, and being suitable for popularization and use in ion implanters.

[0134] The embodiments of the hydrogen-helium coaxial ion implanter are described below.

[0135] Refer to Figures 1 to 12As shown, the hydrogen-helium coaxial ion implanter provided by the present application includes an ion source, a transmission component, and an ion beam output component that are connected in sequence. The ion source is connected to access various reaction gases and output ions. The transmission component transmits the ions to the ion beam output component, and the ion beam output component outputs an accelerated ion beam. The ion implanter is controlled by the ion implanter control system of any of the foregoing embodiments. Exemplarily, the ion implanter of this embodiment may be a hydrogen-helium coaxial ion implanter. The transmission component includes a magnetic analyzer, and the ion beam output component includes an accelerator. When the ion implanter is a hydrogen-helium coaxial ion implanter, two gas flow controllers are used to provide hydrogen and helium, and the expected beam parameters of hydrogen ion and helium ion beams are output by controlling the intake air volume of the two gas flow controllers.

[0136] The hydrogen-helium coaxial ion implanter of the present application realizes an automatic control function, improves the control efficiency, provides powerful intelligent capabilities through a neural network model, realizes an intelligent control function, and improves the control accuracy of the ion implanter. Based on the network-based remote control architecture, multiple ion implanters can be remotely controlled by a client terminal calling a server, which improves the convenience of control operations and also reduces the control cost.

[0137] The above are only partial embodiments of the present application. It should be noted that for those of ordinary skill in the art, without departing from the principle of the present application, several improvements and refinements can be made, and these improvements and refinements should also be regarded as the protection scope of the present application.

Claims

1. An ion implanter control system, characterized in that: include: A host computer, a slave computer, a gas flow controller and a sensor system; wherein the sensor system is arranged on an ion implanter, the gas flow controller is connected to an air inlet pipe of an ion source of the ion implanter, and the slave computer is connected to each gas flow controller and the sensor system; The sensor system is used to detect the real-time status parameters of the ion implanter and send them to the lower computer; The lower computer uploads the real-time status parameters of the ion implanter detected by the sensor system to the upper computer, receives the control parameters of the various gas intakes issued by the upper computer, and adjusts the intake of each gas flow controller according to the control parameters; The upper computer is configured to obtain the expected beam parameters of the ion implanter and the real-time status parameters uploaded by the lower computer, and call a pre-trained neural network to calculate the intake volume of various gases of the ion source at the next moment according to the real-time status parameters and the expected beam parameters of the ion implanter at the current moment, and adjust the intake volume of the ion source at the next moment so that the output beam parameters of the ion implanter meet the requirements of the expected beam parameters.

2. The ion implanter control system according to claim 1, characterized in that: The sensor system comprises: a vacuum gauge arranged in the cavity of the ion implanter and a beam current sensor arranged in the ion beam output component; wherein the vacuum gauge detects the vacuum degree in the cavity of the ion implanter, and the beam current sensor detects the beam current intensity parameter of the ion beam current; The lower computer is also connected to the pneumatic valve, power switch, molecular pump switch and beam controller motor of the ion implanter.

3. The ion implanter control system according to claim 1, characterized in that: The host computer includes: a server and a client terminal; wherein the server communicates with the lower computer and the client terminal respectively through a network; The client terminal accesses the automatic beam adjustment program on the server, receives the desired ion beam flow parameters input by the user through the user operation interface of the automatic beam adjustment program, and displays various status parameters of the ion implanter.

4. The ion implanter control system according to any one of claims 1 to 3, characterized in that: The host computer is configured to obtain the desired beam parameters of the ion implanter output input by the user interface, and control the ion implanter to output the ion beam; detect the initial state data of the current operating state of the ion implanter, input the desired beam parameters and the initial state data into the prediction neural network to obtain the initial intake ratio of various gases of the ion source, and adjust the intake amount of various gases of the ion source according to the initial intake ratio; detect the real-time state data of the ion implanter, when the output beam parameters of the ion implanter do not meet the requirements of the desired beam parameters, calculate the state change data of the ion implanter at the current moment according to the real-time state data, input the state change data and the desired beam parameters into the correction neural network to calculate the intake amount correction value of various gases of the ion source at the next moment, calculate the intake amount of various gases of the ion source according to the intake amount correction value, and adjust the intake amount of the ion source at the next moment, so that the output beam parameters of the ion implanter meet the requirements of the desired beam parameters.

5. The ion implanter control system according to claim 4, characterized in that: The desired beam parameters include desired beam intensity and / or desired beam ion ratio; The initial state data includes the gas pressure in the cavity of the ion implanter, the intensity of the current beam and / or the current beam ion ratio; The state change data includes: a change in the gas pressure in the chamber of the ion implanter, a change in the beam intensity and / or a change in the beam ion ratio.

6. The ion implanter control system according to claim 4, characterized in that: When calculating the intake amount correction value of various gases in the ion source at the next moment, the upper computer is configured to obtain the state change data of the ion implanter in multiple time periods according to the recorded real-time state data as the state change data at the current moment, and input the multiple state change data into the input layer of the correction neural network to calculate the intake amount correction value of various gases in the ion source at the next moment.

7. The ion implanter control system according to claim 4, characterized in that: The host computer uses the following calculation formula to calculate the intake volume of various gases in the ion source: In the formula, is the gas intake volume at the current moment, is the gas intake at the last moment, Q is the total intake of all gases in the ion source, Δr e Indicates the proportional change of gas ions corresponding to the next moment of the corrected neural network output; k S is the correspondence between the desired gas feed ratio and the desired beam ion ratio, k W It is the corresponding relationship between the change of gas intake ratio and the change of beam ion ratio.

8. The ion implanter control system according to claim 4, characterized in that: The prediction neural network comprises: an input layer, a hidden layer and an output layer; wherein each parameter in the hidden layer is obtained by training the historical data of the ion implanter operation; the input layer accesses the initial state data and the expected beam parameters, and the output layer outputs the intake ratio of each gas in the ion source calculated by the hidden layer; The correction neural network includes: an input layer, a hidden layer and an output layer; wherein the parameters in the hidden layer are obtained by training the historical data of the ion implanter operation; the input layer is connected to the state change data and the expected beam parameters generated at the current moment after the ion implanter completes the initial intake ratio adjustment, and the output layer outputs the intake amount correction value of various gases of the ion source at the next moment calculated by the hidden layer.

9. The ion implanter control system according to claim 8, characterized in that: The host computer is also configured to: Building a predictive neural network; Obtain the historical records of the intake volume of various gases in the ion implanter at different time periods and the ion ratio of various ions in the ion beam; The prediction neural network is trained with the state parameters and ion ratio of the ion implanter as input and the intake ratio of various gases as output; or Establishing a revised neural network; Taking the ion ratio and the real-time status data of the ion implanter as input and the intake ratio of various gases as output; The neural network is trained with the intake ratio of various gases at the current moment as the prediction result of the intake ratio of the next moment at the previous moment, and the parameters of the neural network are updated and corrected by back propagation based on the error between the prediction result and the actual data.

10. A hydrogen-helium coaxial ion implanter, characterized in that: include: An ion source, an ion transmission component and an ion beam output component connected in sequence; wherein the ion source is connected to a plurality of gases and outputs ions, the ion transmission component transmits the ions to the ion beam output component, and the ion beam output component outputs an ion beam flow; The ion implanter is controlled by the ion implanter control system according to any one of claims 1 to 9.

Images