Method for regulating gain of photodetector, optical detection module and in-situ detection device

Abstract

The present invention provides a method for regulating the gain of a photodetector, an optical detection module, and an in-situ detection device. Among them, the method for regulating the gain of the photodetector includes obtaining a set gain value of the photodetector and the ambient temperature of the environment where the photodetector is located; inputting the ambient temperature and the set gain value into a gain regulation model to output a reverse bias voltage; transmitting the reverse bias voltage to the photodetector to regulate the current gain value of the photodetector so that it returns to the set gain value. Through this method, a same non-linear relationship can be established between the reverse bias voltage and the gain of the APD detector, so that the APD detector is no longer affected by the ambient temperature. At the same time, the in-situ detection device of the present invention adopts an optical detection module based on the above regulation method, thereby achieving a wider temperature detection range.

Description

Technical Field

[0001] The present invention relates to the technical field of semiconductor detection, and particularly to an in-situ detection device. Background Art

[0002] The epitaxy process refers to the process of precisely growing a single-crystal material with the same lattice arrangement on a single-crystal substrate (usually a wafer). Since the newly grown single-crystal layer strictly follows the extension growth of the substrate crystal phase, it is called the epitaxial layer. The epitaxial layer can be a homoepitaxial layer (Si / Si) or a heteroepitaxial layer (SiGe / Si or SiC / Si, etc.); there are also many methods to achieve epitaxial growth, including molecular beam epitaxy (MBE), ultra-high vacuum chemical vapor deposition (UHV / CVD), atmospheric and reduced-pressure epitaxy (ATM&RP EPI), and so on. According to the growth method, the epitaxy process can be divided into two categories: full epitaxy and selective epitaxy.

[0003] During the epitaxial growth process, an in-situ detection device can detect the sample (substrate or epitaxial growth layer) in real time, that is, obtain relevant information while growing, without removing the sample from the reaction chamber. This is achieved through special designs and technical means, such as coaxial beam design, window materials with high temperature and corrosion resistance, high-precision signal transmission and processing, etc., enabling the detection device to work stably in complex growth environments such as high temperature, high pressure, and chemical atmosphere, and timely reflecting the growth dynamics and changes of the epitaxial growth layer. In this way, operators can adjust the process parameters of the machine, such as temperature, gas flow rate, pressure, etc., at any time according to the real-time monitored data to optimize the epitaxial growth process and thereby improve the quality of the epitaxial growth layer.

[0004] Traditional in-situ detection devices rely on silicon detectors to detect the infrared radiation signal and reflection signal on the surface of the sample. The surface temperature of the wafer or epitaxial growth layer can be obtained through the infrared radiation signal, and the growth dynamics and changes of the epitaxial growth layer can be obtained through the reflection signal. However, due to the low sensitivity of the silicon detector itself and its limited ability to detect weak infrared radiation signals, during the process, when the sample temperature is low, traditional silicon detectors often cannot detect the weak infrared radiation on the surface of the sample, and thus cannot obtain the surface temperature of the sample in the low-temperature state. This drawback as a whole leads to a small temperature measurement range and a high lower limit of temperature measurement for traditional in-situ detection devices for samples, and further leads to the inability of an in-situ detection device to meet the requirements of different epitaxial growth processes for the temperature detection of epitaxial growth samples. Often, different specifications of in-situ detection devices need to be used for different epitaxial growth processes. Summary of the Invention

[0005] In view of the above-mentioned disadvantages of the prior art, the purpose of the present invention is to provide a method for regulating the gain of a photodetector, an optical detection module, and an in-situ detection device, which are used to solve the problems in the prior art that the temperature measurement range of the in-situ detection device is small, the lower limit of temperature measurement is high, and one in-situ detection device cannot meet the different temperature measurement ranges required by different types of epitaxial growth processes.

[0006] To solve the above technical problems, the first aspect of the present invention provides a method for regulating the gain of a photodetector, including:

[0007] Obtain the set gain value of the photodetector and the ambient temperature of the environment where the photodetector is located;

[0008] Input the ambient temperature and the set gain value into a gain regulation model to output a reverse bias voltage; the gain regulation model is constructed using a neural network algorithm and trained with sample ambient temperatures, sample reverse bias voltages, and sample gain values;

[0009] Transmit the reverse bias voltage to the photodetector to regulate the current gain value of the photodetector to return to the set gain value.

[0010] The gain regulation model includes: an input layer, an intermediate hidden layer, and an output layer;

[0011] The step of inputting the ambient temperature and the set gain value into the gain regulation model to output a reverse bias voltage includes:

[0012] Input the ambient temperature and the set gain value into the input layer to obtain input layer data;

[0013] The intermediate hidden layer processes the data of the input layer to obtain an intermediate value;

[0014] The output layer processes the intermediate value to obtain the reverse bias voltage.

[0015] Through the above technical solutions, the achieved technical effect is that the method for regulating the gain of a photodetector in this solution is used to solve the problem that the internal gain of some photodetectors (such as ADP detectors) is easily affected by the ambient temperature. In this solution, by introducing a gain regulation model, according to the ambient temperature of the photodetector, the relationship between the reverse bias voltage and the gain of the photodetector at different ambient temperatures is quantified through machine learning, and the gain drift of the APD detector caused by the ambient temperature is compensated by adjusting the reverse bias voltage, so that the signal output is not affected by the ambient temperature.

[0016] Further, in this solution, by constructing a gain regulation model, the ambient temperature and the set gain value are used as the input layer data. The data of the input layer is processed through the middle hidden layer, and the intermediate value is processed through the output layer, and finally the reverse bias voltage for adjusting the gain value of the photodetector is obtained. The reverse bias voltage acts on the photodetector to compensate for the gain drift caused by the ambient temperature, so that the detection signal output is not affected by the ambient temperature.

[0017] Preferably, the training using the sample ambient temperature, the sample reverse bias voltage, and the sample gain value includes:

[0018] Making a data set from multiple groups of known sample ambient temperature, sample reverse bias voltage, and sample gain value;

[0019] Dividing the data set into a training set and a validation set; using the training set to train the gain regulation model, and using the validation set to verify the accuracy of the gain regulation model;

[0020] Adopting the stochastic gradient descent algorithm as the optimization algorithm in the training process to adjust the parameters of the gain regulation model.

[0021] Through the above technical solution, the achieved technical effect is that in this solution, by training the gain regulation model with a data set made from the sample ambient temperature, the sample reverse bias voltage, and the sample gain value, the best parameters of the gain regulation model can be obtained, so that the model is more targeted and accurate, and a more accurate predicted value (reverse bias voltage) can be obtained.

[0022] The second aspect of the present invention provides an optical detection module, including: a photodetector and an ambient temperature sensing module;

[0023] The photodetector includes a variable gain unit, and the variable gain unit has a set gain value; the photodetector is used to receive an optical signal and convert the optical signal into an amplified electrical signal according to the set gain value;

[0024] The ambient temperature sensing module includes a temperature sensor and a microprocessor. The temperature sensor is used to obtain the ambient temperature data of the environment where the photodetector is located; the microprocessor is electrically connected to the temperature sensor, and the microprocessor is used to input the ambient temperature data and the set gain value into the gain regulation model to output a reverse bias voltage for regulating the photodetector, and the reverse bias voltage is used to adjust the current gain value to return it to the set gain value.

[0025] The gain regulation model includes:

[0026] An input layer for inputting the ambient temperature and the set gain value;

[0027] An intermediate hidden layer for processing the data of the input layer to obtain an intermediate value;

[0028] An output layer for processing the intermediate value to obtain the reverse bias voltage.

[0029] Through the above technical solution, the achieved technical effect is that the optical detection module in this technical solution uses a photodetector with variable gain, which can adapt to the detection of light signals with different intensities. For example, when the surface temperature of the sample to be measured is relatively low and the infrared radiation light is relatively weak, the gain value can be increased to obtain a stronger signal; when the surface temperature of the sample to be measured is relatively high and the infrared radiation light is relatively strong, the gain value can be appropriately reduced to obtain a stable signal intensity.

[0030] Furthermore, to avoid the gain value of the photodetector being affected by the ambient temperature, the ambient temperature sensing module is introduced. The ambient temperature sensing module can detect the ambient temperature where the photodetector is located in real time, and according to the ambient temperature and the set gain value of the photodetector, output in real time a reverse bias voltage for regulating the photodetector. By adjusting the reverse bias voltage, the gain drift of the photodetector caused by the ambient temperature can be compensated, so that the signal output is not affected by the ambient temperature.

[0031] Furthermore, by constructing a gain regulation model, the ambient temperature and the set gain value are used as the data of the input layer. The data of the input layer is processed by the intermediate hidden layer, and the intermediate value is processed by the output layer. Finally, a reverse bias voltage for adjusting the gain value of the photodetector is obtained. This reverse bias voltage acts on the photodetector to compensate for the gain drift caused by the ambient temperature, so that the detection signal output is not affected by the ambient temperature.

[0032] The third aspect of the present invention provides an in-situ detection device, including the optical detection module provided by the second aspect of the present invention, and further including:

[0033] A signal acquisition module, arranged above the sample in the process chamber for epitaxial growth, for acquiring the light signal from the surface of the sample;

[0034] A receiving optical module, the receiving optical module is connected to the signal acquisition module through an optical fiber, for receiving the light signal and collimating, splitting, and focusing the light signal; the optical detection module is arranged on the optical path of the light signal for converting the light signal into an amplified electrical signal;

[0035] A process control module for sending process information; in this embodiment, the process information includes the machine table rotation speed and the wafer layout.

[0036] A signal processing module, electrically connected to the optical detection module and the process control module respectively, is configured to receive the electrical signal and the process information, and generate corresponding data information according to the process information and the electrical signal;

[0037] A display module, electrically connected to the signal processing module, is configured to receive and display the data information in real time.

[0038] Through the above technical solution, the achieved technical effect is that: by using the optical detection module provided in the second aspect of the present invention, the in-situ detection device in this solution can effectively improve the detection of the lower temperature signal on the surface of the sample by the in-situ detection device, thereby greatly broadening the temperature detection range of the in-situ detection device, enabling an in-situ detection device to meet the temperature detection requirements of various epitaxial growth processes for the sample. It effectively solves the problems of high lower limit of the surface temperature detection of the sample (wafer or epitaxial growth layer) and small temperature measurement range in the in-situ detection device in the prior art.

[0039] Preferably, it further includes a signal emission module. The signal emission module is connected to the signal acquisition module through an optical fiber. The signal emission module is configured to generate an emitted optical signal, and the emitted optical signal will generate a reflected optical signal when irradiated on the sample;

[0040] The optical signal includes the reflected optical signal and the infrared radiation signal on the surface of the sample.

[0041] Through the above technical solution, the achieved technical effect is that: by setting the signal emission module, a stable, reliable and appropriately intense signal is provided for the detection process to meet the requirement of accurate temperature measurement of the epitaxial growth layer by the in-situ detection device.

[0042] Preferably, the signal emission module includes:

[0043] A light source electronic module, configured to generate the emitted optical signal;

[0044] An emission optical module, disposed on the optical path of the emitted optical signal, is configured to receive the emitted optical signal and perform collimation, beam splitting and focusing processing on the emitted optical signal.

[0045] Through the above technical solution, the achieved technical effect is that: the light source electronic module, as the energy source of the entire signal emission, can provide a stable, reliable and appropriately intense optical signal for the detection process to meet the requirement of accurate measurement of the optical characteristics of the epitaxial growth layer by the in-situ detection device. The emission optical module can perform collimation, beam splitting and focusing processing on the light emitted by the light source electronic module, so that the optical signal can be vertically irradiated on the detection area of the epitaxial growth layer, improving the spatial resolution and accuracy of the detection, and ensuring that the obtained optical signal can accurately reflect the local characteristics of the epitaxial growth layer.

[0046] Preferably, the signal acquisition module includes a probe, the probe includes a signal transmitting end and a signal collecting end, and the optical fiber includes a transmitting optical fiber and a collecting optical fiber;

[0047] The signal transmitting end of the probe is connected to the transmitting optical module through the transmitting optical fiber, and the receiving optical module is connected to the signal collecting end of the probe through the collecting optical fiber.

[0048] Through the above technical solution, the achieved technical effect is that by arranging the signal transmitting end and the signal collecting end on one probe, the structural integration degree of the probe is effectively improved, and the installation space of the probe is saved.

[0049] Preferably, the signal processing module includes:

[0050] A signal amplifier for receiving the electrical signal and amplifying the electrical signal;

[0051] A signal converter for converting the electrical signal processed by the signal amplifier into a digital signal;

[0052] A central processor for receiving the electrical signal and the process information, generating corresponding data information according to the process information and the electrical signal, and transmitting the data information to the display module for display.

[0053] Through the above technical solution, the achieved technical effect is that by arranging the signal amplifier, the received electrical signal can be amplified; by arranging the signal converter, the electrical signal processed by the signal amplifier can be converted into a digital signal, which is convenient for the central processor to process. The central processor generates corresponding data information according to the process information and the electrical signal, and transmits the data information to the display module for display.

[0054] Preferably, the display module includes a graphics processing system, and the graphics processing system receives the data information and displays the data information as a curve image.

[0055] Through the above technical solution, the achieved technical effect is that the display module can display the received data information as a curve image, improving the readability of the data information. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] Figure 1 It is a schematic flowchart of the method for regulating the gain of a photodetector according to the first embodiment of the present invention;

[0057] Figure 2 It is a specific schematic flowchart of step 300 in the method for regulating the gain of a photodetector according to the first embodiment of the present invention;

[0058] Figure 3 It is a schematic flowchart of step 400 in the method for regulating the gain of the photodetector according to the first embodiment of the present invention;

[0059] Figure 4 It is a schematic diagram showing the relationship between the gain and the reverse bias voltage of the APD detector in the prior art of the present invention at different ambient temperatures (from left to right are -20°C, 0°C, 20°C, 40°C, and 60°C) in sequence;

[0060] Figure 5 It is a schematic diagram showing the non-linear relationship between the reverse bias voltage and the gain of the APD detector under the regulation of the gain regulation model in the first embodiment of the present invention;

[0061] Figure 6 It is a schematic diagram of the module structure of the optical detection module in the second embodiment of the present invention;

[0062] Figure 7 It is a schematic diagram of the module structure of the in-situ detection device in the third embodiment of the present invention;

[0063] Figure 8 It is a schematic diagram showing the connection relationship between the in-situ detection device, the process chamber, and the display in the third embodiment of the present invention.

[0064] Reference numerals:

[0065] 1 - Main chassis;

[0066] 2 - Probe; 21 - Transmitting optical fiber; 22 - Collecting optical fiber;

[0067] 3 - Process chamber; 31 - Machine platform; 32 - Wafer;

[0068] 4 - Display. Detailed implementation manners

[0069] To make the objectives, technical solutions, and advantages of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention. Unless otherwise defined, the technical terms or scientific terms used herein shall have the ordinary meaning as understood by those of ordinary skill in the art in the field to which the present invention pertains. The words such as "including" used herein mean that the elements or objects appearing before the word cover the elements or objects listed after the word and their equivalents, without excluding other elements or objects. The "connection" described herein may be a direct connection or an indirect connection, that is, a connection through an intermediate.

[0070] Embodiment 1:

[0071] Please refer to Figure 1 , the first embodiment of the present invention provides a method for regulating the gain of a photodetector, and the method includes the following steps:

[0072] Step 100: Obtain the set gain value of the photodetector and the ambient temperature of the environment where the photodetector is located; Step 200, construct a gain regulation model using a neural network algorithm.

[0073] Step 300, train the gain regulation model with sample ambient temperature, sample reverse bias voltage, and sample gain value;

[0074] Step 400: Input the ambient temperature and the set gain value into the gain regulation model to output a reverse bias voltage;

[0075] Step 500: Transmit the reverse bias voltage to the photodetector to regulate the current gain value of the photodetector so that it returns to the set gain value.

[0076] In this embodiment, the photodetector is an APD (Avalanche Photo – Diode) detector. The APD detector is a highly sensitive photodetector. Its working principle is based on a photodiode. By applying a reverse bias voltage, the photo-generated carriers undergo an avalanche multiplication effect in the strong electric field region. When the electron-hole pairs generated by photons accelerate in the high electric field and collide with lattice atoms, new electron-hole pairs are excited. This process is like an avalanche, amplifying the photocurrent. Due to the avalanche multiplication effect, the APD detector can detect weak optical signals, but its internal gain is easily affected by the external ambient temperature, such as Figure 4 shows the relationship between the gain and the reverse bias voltage of the APD detector at different ambient temperatures (from left to right are -20°C, 0°C, 20°C, 40°C, and 60°C) in turn. It can be seen that at different ambient temperatures, the gain and the reverse bias voltage of the APD detector maintain different non-linear relationships, which easily causes inaccurate measured temperatures at different ambient temperatures.

[0077] In this solution, by referring to the gain regulation model, according to the ambient temperature of the photodetector, through machine learning, the relationship between the reverse bias voltage and the gain of the photodetector at different ambient temperatures is quantified. By adjusting the reverse bias voltage, the gain drift of the APD detector caused by the ambient temperature is compensated, ensuring that at different ambient temperatures, there is the same non-linear relationship between the reverse bias voltage and the gain of the APD detector (as shown in Figure 5 ), thereby making the APD detector no longer affected by the ambient temperature.

[0078] In this embodiment, the gain regulation model includes: an input layer, an intermediate hidden layer, and an output layer; wherein, the input layer is used to input the environmental temperature and the set gain value; the intermediate layer includes three layers of neurons, and the number of neurons in each layer is 8, 16, and 4 in sequence. A nonlinear factor is introduced into the model by adding an activation function after each layer of neurons to enhance the expression ability; specifically, in this embodiment, the Relu function is used as the activation function; in some embodiments, the Sigmoid function or the Tanh function can be used as the activation function, and a suitable activation function can be selected according to actual needs. The output layer is a single fully connected layer, and the output layer is used to process the intermediate value to obtain the reverse bias voltage.

[0079] Please refer to Figure 2 , in step 300, the process of training using the sample environmental temperature, the sample reverse bias voltage, and the sample gain value includes:

[0080] Step 310, making a data set from multiple groups of known sample environmental temperature, sample reverse bias voltage, and sample gain value.

[0081] Step 320, dividing the data set into a training set and a validation set; using the training set to train the gain regulation model, and using the validation set to verify the accuracy of the gain regulation model. Further, in this step, the K-fold cross-validation method is used to divide the data set.

[0082] Step 330, using the stochastic gradient descent algorithm as the optimization algorithm in the training process to adjust the parameters of the gain regulation model.

[0083] Through the above technical solution, the achieved technical effect is: in this solution, the gain regulation model is trained with a data set made from the sample environmental temperature, the sample reverse bias voltage, and the sample gain value, so that the optimal parameters of the gain regulation model can be obtained, and then the model is more targeted and accurate, and a more accurate predicted value (reverse bias voltage) can be obtained.

[0084] Further, this model uses the root mean square error function as the loss function; in terms of hyperparameter settings, the learning rate is set to 0.0001, the batch size is set to 64, and the number of training times is set to 50.

[0085] Please refer to Figure 3 , in step 400, the process of inputting the environmental temperature and the set gain value into the gain regulation model to output the reverse bias voltage includes:

[0086] Step 410: Input the environmental temperature and the set gain value into the input layer to obtain the input layer data.

[0087] Step 420: The intermediate hidden layer processes the data of the input layer to obtain an intermediate value.

[0088] Step 430: The output layer processes the intermediate value to obtain the reverse bias voltage.

[0089] In this embodiment, by constructing the gain regulation model, the environmental temperature and the set gain value are used as the data of the input layer. The intermediate hidden layer processes the data of the input layer, and the output layer processes the intermediate value. Finally, the reverse bias voltage for adjusting the gain value of the photodetector is obtained. The reverse bias voltage acts on the photodetector to compensate for the gain drift caused by the environmental temperature, so that the output of the detection signal is not affected by environmental factors.

[0090] Embodiment 2:

[0091] Please refer to Figure 6 , the second embodiment of the present invention provides an optical detection module, including: a photodetector and an ambient temperature sensing module.

[0092] The photodetector includes a variable gain unit having a set gain value; the photodetector is configured to receive an optical signal and convert the optical signal into an amplified electrical signal according to the set gain value; specifically, in this embodiment, the photodetector is an APD detector. The variable gain unit is a programmable gain amplifier (PGA).

[0093] The ambient temperature sensing module includes a temperature sensor and a microprocessor. The temperature sensor is configured to obtain the ambient temperature data of the environment where the photodetector is located; the microprocessor is electrically connected to the temperature sensor, and the microprocessor is configured to input the ambient temperature data and the set gain value into the gain regulation model to output a reverse bias voltage for regulating the photodetector, and the reverse bias voltage is used to adjust the current gain value to return it to the set gain value.

[0094] The optical detection module in this embodiment uses a PAD detector with variable gain and can be adapted to the detection of optical signals with different intensities. For example, when the surface temperature of the sample to be measured is relatively low and the infrared radiation light is relatively weak, the gain value can be increased to obtain a stronger signal; when the surface temperature of the sample to be measured is relatively high and the infrared radiation light is relatively strong, the gain value can be appropriately reduced to obtain a stable signal intensity.

[0095] In this embodiment, the temperature sensor is a prior art and will not be described herein; the microprocessor can be an MCU (micro control unit).

[0096] In this embodiment, the gain regulation model includes: an input layer, an intermediate hidden layer, and an output layer.

[0097] The input layer is used to input the ambient temperature and the set gain value; the intermediate hidden layer is used to process the data of the input layer to obtain an intermediate value; the output layer is used to process the intermediate value to obtain the reverse bias voltage.

[0098] Embodiment Three:

[0099] Refer to Figure 7 , the third embodiment of the present invention provides an in-situ detection device, which includes an optical detection module provided in the second embodiment of the present invention, and further includes: a signal acquisition module, a receiving optical module, a process control module, a signal processing module, and a display module.

[0100] The signal acquisition module is disposed above the sample in the process chamber for epitaxial growth, and is used to collect the optical signal from the surface of the sample; in this embodiment, the sample is a wafer or an epitaxial growth layer.

[0101] The receiving optical module is connected to the signal acquisition module through an optical fiber, and is used to receive the optical signal and collimate, split, and focus the optical signal; the optical detection module is disposed on the optical path of the optical signal, and is used to convert the optical signal into an amplified electrical signal.

[0102] The process control module is used to send process information. In this embodiment, the process information includes process type, machine table rotation speed, and wafer layout. The process information is formulated according to different types of epitaxial growth processes. This embodiment is an in-situ detection applied to the GaAs (gallium arsenide) process.

[0103] The signal processing module is electrically connected to the optical detection module and the process control module respectively, and is used to receive the electrical signal and the process information, and generate corresponding data information according to the process information and the electrical signal.

[0104] The display module is electrically connected to the signal processing module, and is used to receive and display the data information in real time.

[0105] The in-situ detection device in this embodiment can effectively improve the detection of the lower temperature signal on the surface of the sample by using the optical detection module provided in the second aspect of the present invention, and further greatly broaden the temperature detection range of the in-situ detection device, so that an in-situ detection device can meet the temperature detection requirements of various epitaxial growth processes for the sample. It effectively solves the problems of high lower limit of the surface temperature of the sample (wafer or epitaxial growth layer) and small temperature measurement range in the in-situ detection device in the prior art.

[0106] In this embodiment, it further includes a signal transmitting module, which is connected to the signal acquisition module through an optical fiber. The signal transmitting module is used to generate an emitted optical signal, and when the emitted optical signal irradiates the sample, a reflected optical signal will be generated.

[0107] The optical signal includes the reflected optical signal and the infrared radiation signal on the surface of the sample. By collecting the infrared radiation signal, the temperature of the surface of the sample can be obtained. By collecting the reflected optical signal, the growth condition of the epitaxial growth layer can be obtained.

[0108] Furthermore, the signal transmitting module includes: a light source electronic module and an emission optical module. Among them, the light source electronic module is used to generate the emitted optical signal; the emission optical module is arranged on the optical path of the emitted optical signal and is used to receive the emitted optical signal and perform collimation, beam splitting, and focusing processing on the emitted optical signal.

[0109] Specifically, the light source electronic module, as the energy source of the entire signal emission, can provide a stable, reliable, and appropriately intense optical signal for the detection process to meet the requirement of accurately measuring the optical characteristics of the epitaxial growth layer by the in-situ detection device. In addition, the light source electronic module can also precisely control parameters such as the wavelength, frequency, and pulse width of the light it emits, enabling it to emit specific optical signals adapted to different detection requirements (such as detecting the epitaxial growth of different materials or different growth stages), thereby achieving more targeted and accurate detection. In this embodiment, the frequency range of the signal light that the light source electronic module can provide is 3 - 1000 Hz, and the pulse width range is 1.6 - 5 ms. The light source electronic module can provide three kinds of light, and the wavelength ranges of the three kinds of light are 390 - 410 nm, 610 - 630 nm, and 940 - 960 nm respectively.

[0110] The emission optical module enables the optical signal to accurately irradiate the detection area of the epitaxial growth layer in a specific shape and spot size, improving the spatial resolution and accuracy of the detection, and ensuring that the obtained optical signal can accurately reflect the local characteristics of the epitaxial growth layer.

[0111] In this embodiment, the signal acquisition module includes a probe. The probe includes a signal emission end and a signal acquisition end. The optical fiber includes an emission optical fiber and a collection optical fiber. The signal emission end of the probe is connected to the emission optical module through the emission optical fiber. The emitted optical signal enters the signal emission end through the emission optical fiber and is emitted from the signal emission end to irradiate the sample.

[0112] The receiving optical module is connected to the signal acquisition end of the probe through the acquisition optical fiber. After the reflected optical signal or the infrared radiation signal is received by the signal acquisition end, it enters the receiving optical module through the acquisition optical fiber.

[0113] In this embodiment, the signal transmitting end and the signal acquisition end are arranged on one probe, effectively improving the structural integration degree of the probe and saving the installation space of the probe.

[0114] In this embodiment, the signal processing module includes: a signal amplifier, a signal converter and a central processor. Among them, the signal amplifier is used to receive the electrical signal and amplify the electrical signal; the signal converter is used to convert the electrical signal processed by the signal amplifier into a digital signal, which is convenient for the central processor to process; in this embodiment, the signal converter is an A / D converter.

[0115] The central processor is used to receive the electrical signal and the process information, generate corresponding data information according to the process information and the electrical signal, and transmit the data information to the display module for display.

[0116] In this embodiment, the display module includes a graphics processing system. The graphics processing system receives the data information and displays the data information as a curve image, thereby improving the readability of the data information. In this embodiment, the display module is a display. The curve image is displayed on the screen of the display.

[0117] Please refer to Figure 8 , further, the in-situ detection device further includes a main chassis 1. The signal transmitting module, the signal acquisition module, the receiving optical module, the process control module and the signal processing module are all arranged in the main chassis 1. The probe 2 is connected to the corresponding interfaces of the main chassis 1 through the transmitting optical fiber 21 and the acquisition optical fiber 22. The process chamber 3 and the display 4 are respectively connected to the corresponding interfaces on the main chassis 1 through cables.

[0118] In the process chamber 3 for epitaxial growth, there is a rotating stage 31. A plurality of the wafers 32 are placed on the upper surface of the stage 31 at equal intervals in a circumferential manner and rotate under the drive of the stage 31; the epitaxial growth layer grows on the surface of the wafer 32. The probe 2 is arranged directly above the process chamber 3 and receives the optical signal on the surface of the wafer 32 or the epitaxial growth layer in real time. The process information (stage rotation speed and wafer layout) is transmitted to the process control module in real time through a cable.

[0119] In this embodiment, the in-situ detection device has two detection scenarios. In the first scenario, the in-situ detection device detects the temperature on the surface of the wafer or the epitaxial growth layer. At this time, the signal emission end of the probe does not work, and the signal acquisition end receives the infrared radiation signal. After being processed by the signal processing module, the infrared radiation signal displays the temperature image of the surface of the wafer or the epitaxial growth layer on the display screen of the display.

[0120] In the second scenario, the in-situ detection device detects the reflected light signal of the epitaxial growth layer. At this time, the signal emission end of the probe works. The emission light signal transmitted by the signal emission module is emitted from the signal emission end to the wafer or the epitaxial growth layer, and after reflection, a reflected light signal is generated. The reflected light signal is received by the signal receiving end of the probe and is transmitted back to the main chassis through the acquisition optical fiber. Finally, after being processed by the signal processing module, the reflected light signal displays the image of the reflectivity on the display screen of the display. Further, in this embodiment, the magnitude of the reflectivity is reflected by the voltage value.

[0121] In summary, a method for regulating the gain of a photodetector, an optical detection module, and an in-situ detection device proposed by the present invention adjust the gain of the APD detector in the in-situ detection device in real time through a gain regulation model, so that there is the same non-linear relationship between the reverse bias voltage and the gain of the APD detector, and thus the APD detector is no longer affected by the ambient temperature. At the same time, since the APD detector has high sensitivity and can detect weak optical signals, the in-situ detection device in the present invention has a larger temperature measurement range. The present invention effectively overcomes various shortcomings in the prior art and has high industrial utilization value.

[0122] The above is only the specific implementation manner of the embodiments of the present application, but the protection scope of the embodiments of the present application is not limited thereto. Any changes or substitutions within the technical scope disclosed in the embodiments of the present application should be covered by the protection scope of the embodiments of the present application. Therefore, the protection scope of the embodiments of the present application should be subject to the protection scope of the claims.

Claims

1. A photodetector gain control method, characterized in that: include: Obtaining a set gain value of the photodetector and an ambient temperature of an environment where the photodetector is located; Inputting the ambient temperature and the set gain value into a gain control model to output a reverse bias voltage; The gain control model is constructed using a neural network algorithm and trained using sample ambient temperature, sample reverse bias voltage and sample gain value; Transmitting the reverse bias voltage to the photodetector to adjust the current gain value of the photodetector to return it to the set gain value; The gain control model includes: an input layer, an intermediate hidden layer and an output layer; The step of inputting the ambient temperature and the set gain value into a gain control model to output a reverse bias voltage includes: Inputting the ambient temperature and the set gain value into the input layer to obtain input layer data; The intermediate hidden layer processes the data of the input layer to obtain an intermediate value; The output layer processes the intermediate value to obtain the reverse bias voltage.

2. The photodetector gain control method according to claim 1, characterized in that: The training using the sample ambient temperature, the sample reverse bias voltage and the sample gain value includes: Constructing a data set with a plurality of known sets of sample ambient temperatures, sample reverse bias voltages and sample gain values; Dividing the data set into a training set and a validation set; using the training set to train the gain control model, and using the validation set to verify the accuracy of the gain control model; The stochastic gradient descent algorithm is used as the optimization algorithm in the training process to adjust the parameters of the gain control model.

3. An optical detection module, characterized in that: include: A photodetector comprising a variable gain unit having a set gain value; The photodetector is used to receive the optical signal and convert the optical signal into an amplified electrical signal according to the set gain value; The ambient temperature sensing module includes a temperature sensor and a microprocessor, wherein the temperature sensor is used to obtain ambient temperature data of the environment where the photodetector is located; the microprocessor is electrically connected to the temperature sensor, and the microprocessor is used to input the ambient temperature data and the set gain value into a gain control model to output a reverse bias voltage for controlling the photodetector, wherein the reverse bias voltage is used to adjust the current gain value to return it to the set gain value; The gain control model includes: Input layer, used to input ambient temperature and set gain value; An intermediate hidden layer, used for processing the data of the input layer to obtain an intermediate value; The output layer is used to process the intermediate value to obtain the reverse bias voltage.

4. An in-situ detection device, characterized in that: The optical detection module according to claim 3 further comprises: A signal collection module, disposed above the sample in the process chamber for epitaxial growth, for collecting optical signals from the surface of the sample; A receiving optical module, which is connected to the signal acquisition module through an optical fiber and is used to receive the optical signal and collimate, split and focus the optical signal; the optical detection module is arranged on the optical path of the optical signal and is used to receive the optical signal processed by the receiving optical module and convert the optical signal into an amplified electrical signal; Process control module, used to send process information; a signal processing module, electrically connected to the optical detection module and the process control module respectively, for receiving the electrical signal and the process information, and generating corresponding data information according to the process information and the electrical signal; The display module is electrically connected to the signal processing module and is used to receive and display the data information in real time.

5. The in-situ detection device according to claim 4, characterized in that: It also includes a signal transmission module, which is connected to the signal collection module through an optical fiber, and is used to generate an emission light signal, and the emission light signal irradiates the sample to generate a reflected light signal; The optical signal includes the reflected light signal and the infrared radiation signal of the sample surface.

6. The in-situ detection device according to claim 5, characterized in that: The signal transmission module comprises: A light source electronic module, used for generating the emission light signal; The transmitting optical module is arranged on the optical path of the transmitting optical signal, and is used for receiving the transmitting optical signal and performing collimation and focusing processing on the transmitting optical signal.

7. The in-situ detection device according to claim 6, characterized in that: The signal acquisition module includes a probe, the probe includes a signal transmitting end and a signal collecting end, and the optical fiber includes a transmitting optical fiber and a collecting optical fiber; The signal transmitting end of the probe is connected to the transmitting optical module through the transmitting optical fiber, and the receiving optical module is connected to the signal collecting end of the probe through the collecting optical fiber.

8. The in-situ detection device according to claim 7, characterized in that: The signal processing module comprises: A signal amplifier, used for receiving the electrical signal and amplifying the electrical signal; A signal converter, used for converting the electrical signal processed by the signal amplifier into a digital signal; The central processor is used to receive the electrical signal and the process information, generate corresponding data information according to the process information and the electrical signal, and transmit the data information to the display module for display.

9. The in-situ detection device according to claim 8, characterized in that: The display module includes a graphics processing system, which receives the data information and displays the data information as a curve image.

Images