A method and apparatus for measuring band gap

Abstract

A kind of band gap measurement method and device, to be measured semiconductor sample is fixed with temperature sensor thermal coupling, resistance measurement probe is connected to general measurement interface, identify probe type and load matching measurement protocol and correction algorithm, apply constant current excitation, establish constant current voltage, temperature acquisition link, real-time acquisition voltage and conversion resistance, temperature data are synchronously collected, calculate temperature change rate and resistance fluctuation variance, determine measurement scene state and switch sampling period, adjust heating link to realize temperature closed-loop control, complete full-temperature zone scanning adaptation.In target temperature interval, at least two groups of different effective length total resistance are measured to obtain intrinsic resistance, obtain temperature intrinsic resistance value pair, slice is carried out to the synchronous acquisition of multiple physical quantity parameters, environmental error correction is carried out, linear relationship is constructed based on the corrected data, band gap is calculated based on the slope and output measurement report, which provides an efficient and reliable technical solution for the accurate measurement of semiconductor material band gap.

Description

Technical Field

[0001] This invention relates to the field of semiconductor material electrical parameter measurement technology, and in particular to a bandgap measurement method and apparatus that supports adaptive sampling, multi-probe compatibility, simultaneous measurement of multiple physical quantities and closed-loop temperature control. Background Technology

[0002] The bandgap width (Eg) is one of the important parameters characterizing the band structure of semiconductor materials. Its magnitude is closely related to the electrical and optical properties of the material and affects the performance characterization and parameter design of semiconductor materials in applications such as photoelectric detection, photovoltaic devices, and electronic devices. Therefore, accurate measurement of the bandgap width of semiconductor materials is of great significance.

[0003] Existing technologies for measuring the bandgap of semiconductors based on the resistance-temperature relationship have several technical problems. Firstly, traditional measurement devices have a fixed sampling rate, which cannot simultaneously capture transient characteristics during the initial heating phase and optimize power consumption during the steady-state phase. Furthermore, fixed-rate sampling leads to wasted communication bandwidth. Secondly, they only support single-mode alligator clip measurements, making them incompatible with different measurement scenarios such as the four-probe method and the van der Burg method, resulting in poor compatibility. Thirdly, they can only measure two parameters—resistance and temperature—offering limited functionality and failing to meet the needs of characterizing the multi-dimensional thermoelectric properties of semiconductor materials. Fourthly, temperature control relies solely on simple on / off control, lacking sufficient accuracy and a closed-loop correction mechanism based on the measurement state, which easily leads to temperature fluctuations affecting measurement accuracy.

[0004] Therefore, the existing technology has problems and needs further improvement and development. Summary of the Invention

[0005] (I) Purpose of the invention: In order to solve the problems existing in the prior art, the purpose of the present invention is to provide a method and device for measuring bandgap width, which can dynamically adjust the sampling frame period without changing the data bit length, balance the capture of measurement details and the occupation of communication bandwidth, be compatible with multiple measurement probes and automatically identify and match the corresponding correction algorithm, and expand the device from a single bandgap width measuring instrument to a comprehensive thermoelectric characteristic analysis platform for semiconductor materials.

[0006] (II) Technical Solution: To solve the above-mentioned technical problems, this technical solution provides a method for measuring the band gap width, including the following steps:

[0007] The semiconductor sample under test is thermally coupled and fixed to the temperature sensor. The resistance measurement probe is connected to the general measurement interface. The type of resistance measurement probe is identified and the matching measurement protocol and correction algorithm are loaded. A constant current excitation is applied to the semiconductor sample under test, and a constant current excitation and voltage acquisition measurement link and a temperature acquisition link are established.

[0008] The voltage data of the semiconductor sample under test is collected in real time by constant current excitation and voltage acquisition measurement link and converted into resistance data. At the same time, the temperature data of the semiconductor sample under test is collected in real time by temperature acquisition link. The temperature change rate and resistance fluctuation variance are calculated, the measurement scenario state is determined and the corresponding sampling period is switched. The heating link is adjusted synchronously to realize temperature closed-loop control and complete the dynamic adaptation of the full temperature range scanning process.

[0009] At each stable temperature point within the target temperature range, complete at least two sets of total resistance measurements with different effective lengths. Obtain the intrinsic resistance value of the semiconductor sample under test through difference calculation to obtain the temperature intrinsic resistance value pair. Simultaneously acquire multiple physical quantity parameters and dynamically slice the temperature intrinsic resistance value pair and multiple physical quantity parameters according to a fixed frame.

[0010] The system performs time-stamp alignment, quality screening, and environmental error correction on temperature intrinsic resistance pairs and multiple physical parameters. Based on the corrected temperature intrinsic resistance pairs, a linear relationship is constructed. The bandgap of the semiconductor sample under test is calculated according to the slope, and the final measurement report is output.

[0011] Preferably, step 1 includes:

[0012] Standardized pretreatment is performed on the semiconductor samples to be tested to remove surface oxide layers and contaminants;

[0013] The semiconductor sample under test is thermally coupled and fixed to the heating device by using high thermal conductivity silicone grease to establish a low-resistance continuous heat conduction path.

[0014] The temperature sensor probe is thermally coupled and fixed to the geometric center of the upper surface of the semiconductor sample under test to establish a temperature signal acquisition and transmission path.

[0015] The resistance measurement probe is physically mounted to the semiconductor sample under test, and a standardized electrical connection is made to the general measurement interface module.

[0016] Power on the low-voltage circuit, constant current excitation circuit, and heating circuit in stages, and simultaneously complete the basic initialization of all hardware peripherals and communication interfaces.

[0017] By outputting a standard detection excitation signal and acquiring probe response characteristics, the characteristic parameters of the resistance measurement probe are extracted and matched with the probe feature library to identify the type of resistance measurement probe.

[0018] Based on the type of resistance measurement probe, the system automatically retrieves and loads the matching measurement communication protocol and error correction algorithm to complete the configuration of measurement channel parameters.

[0019] The constant current excitation module is activated to output a constant current, and the current output accuracy is locked through real-time sampling and closed-loop calibration.

[0020] Construct constant current excitation and voltage acquisition measurement links, and temperature acquisition links, and verify the effectiveness and data stability of the links respectively.

[0021] Preferably, the types of resistance measurement probes identified include:

[0022] The main control microcontroller outputs a standard detection excitation signal to the resistance measurement probe through a DAC peripheral and synchronously acquires the response waveform of the resistance measurement probe through an ADC acquisition channel. Based on the response waveform, the core characteristic parameters of the resistance measurement probe are calculated. The core characteristic parameters include input impedance, amplitude-frequency characteristic attenuation, and signal rise time. The core characteristic parameters are then normalized and matched against a probe feature library. When the matching degree between the core characteristic parameters of the resistance measurement probe and the standard characteristic parameter range of a certain type of probe pre-stored in the probe feature library is greater than or equal to a first matching threshold, the probe is determined to be of the corresponding type.

[0023] Preferably, the resistance measurement probe includes clamp-on electrodes, four-probe probes, van der Burg probes, and non-contact induction probes.

[0024] Preferably, step 2 includes:

[0025] Start the constant current excitation and voltage acquisition measurement link and temperature acquisition link, initialize the relevant parameters according to the probe type, and start the ADC multi-channel continuous acquisition mode;

[0026] The resistance and temperature data of the semiconductor sample under test are simultaneously acquired through constant current excitation and voltage acquisition measurement links and temperature acquisition links, and then preprocessed.

[0027] Linear fitting is used to calculate the rate of temperature change, and the variance of resistance fluctuation is calculated using the variance formula.

[0028] The temperature change rate and resistance fluctuation variance are compared with preset thresholds to determine four measurement scenario states and complete the scenario state switching verification.

[0029] The sampling period is dynamically switched according to the scene status, and the temperature control execution module synchronously outputs the corresponding heating control signal to achieve closed-loop regulation.

[0030] Confirm the full temperature range scanning range, dynamically adapt the scene state, sampling period and temperature control mode according to temperature changes during the scanning process, and cyclically determine the temperature state of the semiconductor sample under test.

[0031] Preferably, determining the measurement scene state and switching the corresponding sampling period includes:

[0032] The calculated rate of temperature change and the variance of resistance fluctuation are compared with preset thresholds to determine whether it is a rapid temperature change state, a medium-speed temperature change state, a steady temperature state, or a natural cooling state.

[0033] When a rapid temperature change is detected, switch to a high-frequency sampling period; when a medium-speed temperature change is detected, switch to a normal sampling period; when a steady-temperature state or a natural cooling state is detected, switch to a low-frequency sampling period.

[0034] The switching process of the sampling period does not change the frame structure and effective data bit width of the measurement protocol.

[0035] Preferably, synchronous adjustment of the heating link to achieve closed-loop temperature control includes:

[0036] When a rapid temperature change is detected, the main control microcontroller controls the heating link to operate at full power; when a medium-speed temperature change is detected, the main control microcontroller outputs a PWM signal to linearly adjust the heating power; when a steady temperature state is detected, the main control microcontroller uses a PID algorithm to adjust the PWM duty cycle in real time to achieve high-precision constant temperature control of ±0.1℃; when a natural cooling state is detected, the main control microcontroller shuts off the heating output.

[0037] Preferably, step 3 includes:

[0038] Confirm the effectiveness of the steady-state test, calibrate the measurement parameters that match the probe type, initialize the acquisition of relevant parameters, and start the intrinsic resistance extraction and multi-physical quantity acquisition process;

[0039] Collect at least two sets of total resistance data under the same steady-state temperature, bind timestamps, and handle collection anomalies;

[0040] The corresponding correction algorithm is invoked to offset systematic errors through difference calculation, and the intrinsic resistance value of the sample is extracted and bound to temperature.

[0041] Simultaneously acquire multiple physical quantity parameters to complete outlier removal, range calibration, and time sequence binding;

[0042] Multi-source data slice encapsulation and CRC16 verification are completed according to the fixed frame length of the corresponding measurement protocol.

[0043] Preferably, extracting the intrinsic resistance value of the sample includes: obtaining at least two total resistance measurements by changing the clamping position to form different effective lengths at the same stable temperature point, and calculating the difference between the at least two total resistance measurements at the same temperature point to offset the contact resistance.

[0044] Preferably, the multi-physical parameters collected simultaneously include ambient humidity, sample clamping pressure, ambient temperature, and constant current excitation stability parameters;

[0045] The multiple physical quantity parameters and the intrinsic temperature resistance value are bound to the same timestamp, encapsulated with a fixed frame length, CRC checked, and then uploaded to the host computer.

[0046] Preferably, step 4 includes:

[0047] Match the corresponding multi-source environmental error according to the type of resistance measurement probe;

[0048] The error correction is performed sequentially step by step to eliminate environmental interference errors one by one and obtain the calibrated intrinsic resistance value.

[0049] The calibrated intrinsic resistance value is converted into intrinsic resistivity, and the bandgap value is solved by linear fitting.

[0050] The final measurement report is output in multiple formats.

[0051] Preferably, converting the calibrated intrinsic resistance value into intrinsic resistivity and solving for the bandgap value through linear fitting includes:

[0052] The corrected intrinsic resistance value is converted to intrinsic resistivity ρ=R. cal ×S / L, where S is the sample cross-sectional area and L is the effective measurement length, expressed as the reciprocal of the absolute temperature. The x-axis represents the natural logarithm of intrinsic resistivity. A linear fit is performed on the ordinate to obtain the fitting slope m, and the band gap width Eg = 2k·m is calculated, where k is the Boltzmann constant.

[0053] A bandgap width measuring device for performing the bandgap width measuring method, comprising a link construction unit, a sampling temperature control unit, a resistance measurement unit, and a processing and calculation unit;

[0054] The link construction unit thermally couples and fixes the semiconductor sample under test with the temperature sensor, connects the resistance measurement probe to the general measurement interface, identifies the type of the resistance measurement probe and loads the matching measurement protocol and correction algorithm, applies constant current excitation to the semiconductor sample under test, and establishes a constant current excitation and voltage acquisition measurement link and a temperature acquisition link.

[0055] The sampling temperature control unit collects the voltage data of the semiconductor sample under test in real time based on the constant current excitation and voltage acquisition measurement link and converts it into resistance data. Simultaneously, it collects the temperature data of the semiconductor sample under test in real time through the temperature acquisition link, calculates the temperature change rate and resistance fluctuation variance, determines the measurement scenario state and switches the corresponding sampling period, and synchronously adjusts the heating link to achieve temperature closed-loop control, thus completing the dynamic adaptation of the full temperature range scanning process.

[0056] The resistance measurement unit completes at least two sets of total resistance measurements with different effective lengths at each stable temperature point within the target temperature range. It obtains the intrinsic resistance value of the semiconductor sample under test through difference calculation, and obtains the temperature intrinsic resistance value pair. It also collects multiple physical quantity parameters simultaneously and dynamically slices the temperature intrinsic resistance value pair and the multiple physical quantity parameters according to a fixed frame.

[0057] The processing and calculation unit performs time-stamp alignment, quality screening, and environmental error correction on the temperature intrinsic resistance value pairs and multiple physical quantity parameters. Based on the corrected temperature intrinsic resistance value pairs, it constructs a linear relationship, calculates the bandgap of the semiconductor sample under test according to the slope, and outputs the final measurement report.

[0058] (III) Beneficial Effects: The present invention provides a method and apparatus for measuring bandgap width. First, at the same temperature point, two sets of total resistances with different effective lengths are obtained by changing the clamping position. The difference is then used to directly cancel out contact resistance terms unrelated to the effective length, thereby accurately extracting the intrinsic resistance value of the sample, eliminating contact resistance interference, and improving measurement accuracy and result consistency. Second, by calculating the temperature change rate and resistance fluctuation variance in real time, four measurement scenarios are dynamically determined, and the sampling period is automatically switched according to the scenario, providing a stable and controllable temperature environment for bandgap width measurement. Furthermore, by outputting a standard detection excitation signal and acquiring probe response characteristics, multiple measurement probes can be automatically identified, and matching measurement protocols and correction algorithms can be automatically loaded, significantly improving the compatibility and scenario adaptability of the device. Finally, strict timestamp matching is performed on the resistance and temperature data to ensure that each set of resistance values ​​corresponds to the temperature at its actual sampling time, reducing the interference of non-intrinsic conductivity ranges or noise data on the fitting results. Attached Figure Description

[0059] Figure 1 This is a flowchart of the steps of a bandgap width measurement method according to the present invention;

[0060] Figure 2 This is a schematic diagram of the structure of a bandgap width measuring device according to the present invention;

[0061] Figure 3 This is a connection structure diagram of a bandgap width measuring device according to a specific embodiment of the present invention;

[0062] Figure 4 This is a schematic diagram of the lnR–1 / T fitting results in an embodiment of the present invention.

[0063] Figure reference numerals: 2-constant current excitation module, 3-germanium block semiconductor, 4-main control microcontroller, 8-PTC heater, 9-K-type armored thermocouple, 10-signal converter. Detailed Implementation

[0064] The present invention will be further described in detail below with reference to preferred embodiments. More details are set forth in the following description in order to provide a full understanding of the present invention. However, the present invention can obviously be implemented in many other ways different from those described herein. Those skilled in the art can make similar extensions and derivations based on actual application situations without departing from the spirit of the present invention. Therefore, the scope of protection of the present invention should not be limited by the content of this specific embodiment.

[0065] The accompanying drawings are schematic diagrams of embodiments of the present invention. It should be noted that these drawings are for illustrative purposes only and are not drawn to scale, and should not be construed as limiting the actual scope of protection of the present invention.

[0066] A method for measuring bandgap width, such as Figure 1 As shown, the specific steps include:

[0067] Step 1: Complete the thermal coupling fixation of the semiconductor sample under test and the temperature sensor, connect the resistance measurement probe to the general measurement interface, identify the type of resistance measurement probe and load the matching measurement protocol and correction algorithm, apply constant current excitation to the semiconductor sample under test, and establish constant current excitation and voltage acquisition measurement link and temperature acquisition link.

[0068] Step 2: Based on the constant current excitation and voltage acquisition measurement link, the voltage data of the semiconductor sample under test is acquired in real time and the resistance data is converted. Simultaneously, the temperature data of the semiconductor sample under test is acquired in real time through the temperature acquisition link. The temperature change rate and resistance fluctuation variance are calculated, the measurement scenario state is determined and the corresponding sampling period is switched. The heating link is adjusted synchronously to realize temperature closed-loop control and complete the dynamic adaptation of the full temperature range scanning process.

[0069] Step 3: At each stable temperature point within the target temperature range, complete at least two sets of total resistance measurements with different effective lengths. Obtain the intrinsic resistance value of the semiconductor sample under test through difference calculation to obtain the temperature intrinsic resistance value pair. Simultaneously collect multiple physical quantity parameters and dynamically slice the temperature intrinsic resistance value pair and multiple physical quantity parameters according to a fixed frame.

[0070] Step 4: Perform time-stamp alignment, quality screening, and environmental error correction on the temperature intrinsic resistance pairs and multiple physical parameters. Based on the corrected temperature intrinsic resistance pairs, construct a linear relationship, calculate the bandgap of the semiconductor sample under test according to the slope, and output the final measurement report.

[0071] Step 1 specifically includes:

[0072] Step 101: Perform standardized pretreatment on the semiconductor sample to be tested to remove the surface oxide layer and contaminants.

[0073] Take the semiconductor sample to be tested, and use an analytical grade anhydrous ethanol-soaked microfiber dust-free wiping cloth to repeatedly wipe and clean the upper and lower heat-conducting surfaces, the two electrode contact surfaces in the length direction, and the effective measurement area of ​​the semiconductor sample. The number of wipes on a single side shall not be less than 3 times. Wear dust-free gloves throughout the process to avoid direct contact between fingers and the sample surface.

[0074] For semiconductor samples with a noticeable oxide layer on their surface, the end face of the sample is lightly polished in a single direction using 1500-grit metallographic sandpaper. After polishing, the polished area is cleaned again with anhydrous ethanol to remove polishing debris. After cleaning, the semiconductor sample is placed in a dust-free and dry environment to air dry naturally for at least 30 seconds, ensuring that there is no ethanol residue, dust, or oil on the surface of the sample. This eliminates random fluctuations in contact resistance caused by oxide layers, contaminants, and oil on the sample surface, ensuring the stability of electrical contact between the subsequent electrodes and the semiconductor sample, and reducing measurement system errors from the source.

[0075] The semiconductor sample to be tested in this invention is preferably a regular cuboid block with dimensions of 30mm×5mm×3mm, but can be adapted to other regular geometric sizes.

[0076] Step 102: Use high thermal conductivity silicone grease to thermally couple and fix the semiconductor sample to be tested to the heating device to establish a low-resistance continuous thermal conduction path.

[0077] After confirming that the heating surface of the heating device is clean and free of foreign objects, apply high thermal conductivity silicone grease evenly to the central area of ​​the heating surface, ensuring the application area perfectly matches the bottom dimensions of the semiconductor sample under test. Place the pre-treated semiconductor sample from step 101 horizontally on the grease-coated area of ​​the heating surface, ensuring complete contact between the bottom surface of the sample and the heating surface, without any gaps or offset. Apply uniform, light pressure to the sample to expel excess silicone grease and residual air gaps between the sample and the heating surface, ensuring the silicone grease layer is free of bubbles and breaks, forming a continuous heat conduction path. Simultaneously, use high-temperature resistant polyimide tape to further secure the sides of the sample to the edge of the heating surface, ensuring no relative displacement during heating and cooling, maintaining stable thermal coupling, and avoiding problems such as temperature measurement lag and uneven temperature distribution caused by poor thermal contact.

[0078] The heating device described in this invention is a PTC heater. The thickness of the high thermal conductivity silicone grease is 0.1 mm to 0.2 mm, and the thermal conductivity is ≥2.0 W / (m·K).

[0079] Step 103: Thermally couple and fix the temperature sensor probe to the geometric center of the upper surface of the semiconductor sample to be tested, and establish a temperature signal acquisition and transmission path.

[0080] Apply high thermal conductivity silicone grease to the geometric center of the surface of the semiconductor sample under test, completely embedding the temperature sensor probe end into the grease. Ensure that the metal wire of the temperature sensor probe end is in close contact with the surface of the semiconductor sample, without any gaps or air gaps. Use high-temperature resistant silicone to cure and fix the contact area between the temperature sensor probe end and the semiconductor sample under test, ensuring that the temperature sensor has no relative displacement throughout the measurement process and that the thermal coupling is stable. This enables accurate and real-time acquisition of the temperature in the core area of ​​the semiconductor sample under test, establishes a stable temperature signal transmission path, and avoids temperature data distortion caused by temperature measurement position deviation or poor thermal coupling.

[0081] Connect the positive and negative signal output lines of the temperature sensor to the corresponding thermocouple input terminals of the temperature acquisition module. Tighten the screws on the thermocouple input terminals to ensure a secure electrical connection without looseness or intermittent contact. Confirm that the signal output terminal of the temperature acquisition module and the input terminal of the signal converter are electrically connected. The USB port of the signal converter should be pre-connected to the corresponding port on the host computer.

[0082] The temperature sensor uses a type K armored thermocouple with accuracy class I and a temperature measurement range of -20℃ to 200℃. The exposed metal wire length at the thermocouple probe is 2mm to 3mm. The high-temperature resistant silicone sealant covers a 2mm area around the probe.

[0083] Step 104: Physically install the resistance measurement probe to the semiconductor sample under test, and make standardized electrical connections with the universal measurement interface module.

[0084] Select a resistance measurement probe according to the measurement requirements, and confirm that the measuring end of the probe is clean, free from oxidation, and undamaged. The types of resistance measurement probes include low-contact-resistance alligator clip electrodes, four-probe probes, van der Burg method probes, and non-contact inductive probes, among others.

[0085] When using alligator clip electrodes, two electrodes are clamped to the two end faces along the length of the semiconductor sample, with consistent clamping depth. The electrode clamping contact surfaces are fully in contact with the end faces of the semiconductor sample, and the clamping pressure is uniform and stable. When using a four-probe sensor, the four probe heads are pressed perpendicularly and equidistantly to the central flat area of ​​the upper surface of the semiconductor sample, with probe pressure conforming to the four-probe measurement standard, ensuring good contact between all probes and the sample surface. When using a van der Bauer method probe, four probes are symmetrically pressed to the four corners of the semiconductor sample, ensuring consistent contact area and contact pressure. When using a non-contact inductive probe, the probe sensing surface is kept at a fixed distance from the surface of the semiconductor sample, preferably 0.5 mm in this invention. The probe position is fixed by a positioning bracket to ensure that the distance remains unchanged throughout the measurement.

[0086] Connect the signal cable of the resistance measurement probe to the standardized pluggable terminals of the universal measurement interface module, and tighten the standardized pluggable terminals to ensure a secure electrical connection without crossing or loose connections. Confirm the internal signal terminals of the universal measurement interface module and pre-connect them to the ADC acquisition channel of the main control microcontroller and the output terminal of the constant current excitation module. Establish a stable electrical contact / inductive coupling path between the resistance measurement probe and the semiconductor sample under test, completing the physical connection between the resistance measurement probe and the hardware, providing a physical basis for subsequent adaptive identification and resistance measurement.

[0087] Step 105: Power on the devices in stages according to the order of low voltage, constant current excitation, and heating link, and simultaneously complete the basic initialization of all hardware peripherals and communication interfaces.

[0088] Strictly follow the sequence of low-voltage, constant current excitation, and heating link power-on in stages to avoid damage to precision data acquisition devices from power-on surges. First, connect a 5V regulated power supply to power the main control microcontroller, general measurement interface, and temperature acquisition module, confirming that the power indicator light is constantly on and there are no overcurrent or short-circuit alarms. Second, connect a 9V isolated DC power supply to power the constant current excitation module, confirming that the constant current excitation module's power indicator light is normal and there are no abnormal outputs. Third, connect a 12V DC power supply to power the relays and heating devices in the heating link, confirming that the heating link is in standby mode and heating is functioning correctly. This staged power-on process avoids hardware damage from power surges, completes the basic configuration of all hardware peripherals, and establishes the basic communication links between the main control microcontroller and each module and host computer, providing the hardware operating environment for subsequent adaptive operations.

[0089] The main control microcontroller powers on normally without alarms. After resetting, it performs basic hardware initialization. The main control microcontroller completes initialization without errors, communicates normally with the host computer via serial port, and all peripherals are in ready standby state. The overall completion time is ≤500ms.

[0090] Initialize the system clock, configuring the main frequency to 168MHz, and enable clocking for the ADC, GPIO, UART, timers, and DAC peripherals. When initializing the 12-bit ADC acquisition channel, configure the sampling time to 480 clock cycles, enable multi-channel scanning mode, and configure the reference voltage to a 3.3V precision reference source. When initializing the GPIO port, configure the constant current module enable pin and relay control pin to push-pull output mode, and configure the gain control pin of the general-purpose measurement interface to multiplexed output mode. When initializing the UART1 serial port, configure the baud rate to 115200, 8 data bits, 1 stop bit, and no parity bit, establishing a communication link with the host computer. When initializing the DAC peripheral, configure it to 12-bit voltage output mode to provide excitation signals for subsequent probe feature detection.

[0091] Step 106: The main control microcontroller outputs a standard detection excitation signal, collects probe response characteristics, extracts the characteristic parameters of the resistance measurement probe, matches them with the probe feature library, and identifies the type of resistance measurement probe.

[0092] The main control microcontroller outputs a standard detection excitation signal to the resistance measurement probe connected to the general measurement interface through the DAC peripheral. The standard detection excitation signal is a square wave signal with a fixed frequency of 1kHz, a peak-to-peak value of 50mV, and an output impedance of 50Ω. The signal has no DC bias and no spike interference.

[0093] The main control microcontroller synchronously acquires the response waveform of the resistance measurement probe to the standard detection excitation signal through the ADC acquisition channel, continuously acquiring waveform data for 100 cycles, and taking the average value as the valid response data. Based on the valid response data, the core characteristic parameters of the resistance measurement probe are calculated by comparing Ohm's law, waveform slope and amplitude, including input impedance, amplitude-frequency attenuation, signal rise time, and short-circuit / open-circuit response characteristics.

[0094] Specifically, after the standard detection excitation signal output stabilizes, the effective value V of the response voltage across the resistance measurement probe is acquired. rms With the effective value of the circuit excitation current I rms Input impedance Z = V rms / I rms Different resistance measurement probes have different input impedance ranges and are used for low contact resistance alligator clip electrodes, four-probe probes, van der Burg method probes, and non-contact induction probes.

[0095] When Z < 5Ω, the stable electrical contact / inductive coupling path between the resistance measurement probe and the semiconductor sample under test is determined to be a short circuit; when Z > 1MΩ, the stable electrical contact / inductive coupling path between the resistance measurement probe and the semiconductor sample under test is determined to be an open circuit; when 5Ω ≤ Z ≤ 1MΩ, the stable electrical contact / inductive coupling path between the resistance measurement probe and the semiconductor sample under test is determined to be a normal measurement probe characteristic.

[0096] The standard detection excitation signal amplitude V in Using this as a reference, the amplitude V of the resistance measurement probe output response is collected. out Calculate the amplitude-frequency characteristic attenuation A tt =20× This is used to distinguish between resistive probes and inductive probes.

[0097] The time difference Δt between the effective response data and the amplitude of 10% to 90% is used to obtain the signal rise time T. r =t 90% -t 10% This is used to distinguish between purely resistive contact probes and capacitive non-contact probes.

[0098] The calculated core characteristic parameters of the resistance measurement probe are compared with a pre-stored probe feature library in Flash memory using a normalized matching method. This library includes standard characteristic parameter ranges for clamp-on electrodes, four-probe probes, van der Burg probes, and non-contact sensing heads. When the matching degree between the core characteristic parameters of the resistance measurement probe and the pre-stored standard characteristic parameter range of a certain type of probe in the feature library is greater than or equal to a first matching threshold, the probe is identified as belonging to the corresponding type, and the identification result is output. When the matching degree is less than the first matching threshold, the normalized matching comparison is repeated 3 to 5 times. If all matching attempts fail, a probe identification failure alarm is sent to the host computer.

[0099] In this invention, the first matching threshold is set to 95%.

[0100] Step 107: The main control microcontroller automatically retrieves and loads the matching measurement communication protocol and error correction algorithm according to the type of resistance measurement probe, and completes the configuration of measurement channel parameters and communication handshake verification.

[0101] Based on the resistance measurement probe type identification result output in step 106, the main control microcontroller automatically retrieves the measurement communication protocol, data parsing rules, and correction compensation algorithm that are completely matched to the probe type from the protocol algorithm library in the on-chip Flash.

[0102] When the identified resistance measurement probe is a clamp-on electrode, the standard communication protocol for two-end resistance measurement and the contact resistance compensation algorithm with the same temperature difference are loaded, the ADC acquisition channel is configured as differential input mode, and the gain is configured as 1x.

[0103] The two-end resistance measurement standard communication protocol is a UART-based master-slave half-duplex fixed-frame-length communication protocol used to realize measurement control, data upload, and status interaction between a microcontroller and a host computer. The physical layer parameters of the two-end resistance measurement standard communication protocol are fixed at a baud rate of 115200bps, 8 data bits, 1 stop bit, and no parity bit. The frame structure adopts a fixed length and fixed effective data bit design, with a frame header of 0×AA containing resistance data segments, temperature data segments, and a parity segment. It supports adaptive sampling rate switching without changing the frame structure, ensuring reliable synchronous upload of resistance, temperature, and multiple physical quantity data. The same-temperature difference contact resistance compensation algorithm obtains two sets of total resistances R1 and R2 with different effective lengths at the same temperature point by changing the electrode clamping position. Using the difference operation ΔR=R1-R2, the contact resistance term unrelated to the effective length is mathematically canceled out, retaining only the intrinsic resistance-related component of the semiconductor sample under test, thereby eliminating the systematic error caused by the contact resistance in the bandgap calculation.

[0104] When the resistance measurement probe is identified as a four-probe probe, the four-probe resistivity measurement standard protocol, thin-layer resistivity correction formula, and probe spacing correction algorithm are loaded, and the ADC acquisition channel is configured as a 4-channel synchronous scanning mode with a gain of 10x.

[0105] The four-probe resistivity measurement standard protocol is a dedicated measurement communication protocol based on UART half-duplex and fixed frame length. It is used to realize excitation control, voltage acquisition, probe status synchronization, and resistivity data upload between the four-probe probe and the main control microcontroller. The four-probe resistivity measurement standard protocol adopts a fixed frame structure, which includes a frame header, channel identifier, four voltage data channels, probe spacing parameters, and check bits. It supports automatic adaptation to various configurations such as linear four-probe and square four-probe, and the frame length does not change with the sampling rate.

[0106] The thin-film resistivity correction formula is designed for semiconductor thin-film samples to eliminate systematic errors caused by thickness, boundaries, probe spacing, and the ratio of the sample size to the actual thickness. When the thickness d of the semiconductor sample under test is much smaller than the probe spacing s, the measured resistance is converted to the true resistivity: ρ = 2πs·(V / I)·C t Where ρ is the corrected resistivity; s is the probe center-to-center distance; V / I is the measured surface resistance; C t The thin-layer correction coefficient is determined by the ratio of sample thickness to probe spacing, d / s.

[0107] The probe spacing correction algorithm corrects the resistivity calculation results in real time based on the deviation between the actual probe spacing and the theoretical nominal value, eliminating machining and assembly errors. The correction coefficient K = s / s0, where s0 is the theoretical probe spacing pre-stored in the system, and s is the actual probe spacing obtained through calibration with a standard sample. Multiplying the corrected resistivity by the correction coefficient K yields the final true resistivity ρ. real =ρ×K.

[0108] When the resistance measurement probe is identified as a Van der Burg method probe, the Van der Burg method symmetric measurement protocol, geometric factor correction algorithm, and contact resistance correction algorithm are loaded, and the ADC acquisition channel is configured as a 4-channel cyclic switching mode with a gain of 10x.

[0109] The Van der Bauer method symmetric measurement protocol is a dedicated symmetric measurement communication protocol based on UART half-duplex and fixed frame length. It is used to achieve excitation current switching, voltage acquisition, electrode channel selection, symmetric measurement timing control, and data upload between the Van der Bauer four-electrode probe and the main control microcontroller. The Van der Bauer method symmetric measurement protocol adopts a symmetric measurement frame structure, including a frame header, electrode channel identifier, forward / reverse voltage data, temperature synchronization segment, and CRC check segment. It supports automatic polling measurement of the four electrodes, and changes in the sampling rate do not alter the frame format or the length of the effective data bits.

[0110] The geometric factor correction algorithm calculates the ratio of electrode eccentricity to boundary distance based on the actual geometry of the semiconductor sample under test, electrode position symmetry, and boundary distance. According to the sample shape and electrode distribution, the geometric correction factor G is obtained by looking up a table or calculation. The original measured resistivity is multiplied by the geometric correction factor G to obtain the corrected resistivity ρ. corrected =ρ measured ×G corrects the original measurements from the van der Burg method, eliminating systematic errors caused by non-ideal sample shape, electrode eccentricity, and boundary effects.

[0111] The contact resistance correction algorithm addresses the case where the contact resistance between the Van der Burg method electrode and the sample is not negligible. A forward excitation current I is applied between the same set of electrodes, and the total voltage V1, including the sample voltage and the contact resistance voltage drop, is measured. A reverse excitation current -I is applied, and the total voltage V2, including the difference between the sample voltage and the contact resistance voltage drop, is measured. The average of the forward and reverse voltages is then used to offset the contact resistance term V. sample =(V1-V2) / 2, the corrected V sample Determined solely by the intrinsic resistance of the sample, the contact resistance component is completely eliminated, allowing the extraction of the pure intrinsic voltage signal of the sample.

[0112] When the resistance measurement probe is identified as a non-contact induction probe, the non-contact induction measurement protocol, coupling distance error correction algorithm, and induction signal demodulation rules are loaded, and the ADC acquisition channel is configured as a single-ended high-speed sampling mode with a gain of 100 times.

[0113] The non-contact inductive measurement protocol is a dedicated inductive measurement communication protocol based on UART half-duplex and with a fixed frame length. It is used to realize excitation signal control, inductive signal acquisition, coupling status monitoring, resistivity data upload, and status interaction between the non-contact inductive probe and the main control unit. The non-contact inductive measurement protocol adopts a fixed frame structure, including a frame header, inductive signal data segment, coupling distance parameter segment, temperature synchronization segment, and CRC check segment. It supports adaptive sampling rate dynamic switching without changing the frame length and effective data bit width, ensuring stable, reliable, and synchronous upload of non-contact measurement data.

[0114] The coupling distance error correction algorithm is a correction algorithm that compensates the inductive measurement signal in real time based on the actual coupling distance between the non-contact inductive probe and the sample surface, eliminating the measurement deviation caused by the distance fluctuation. The system obtains the actual value of the coupling distance d in real time, compares it with the standard calibration distance d0 to obtain the distance deviation coefficient, corrects the measured inductive impedance according to the distance sensitivity calibration curve, and outputs the true equivalent resistance and resistivity data after eliminating the influence of the distance.

[0115] The demodulation rule for the inductive signal is a rule for detecting, amplifying, filtering, and extracting the effective value of the high-frequency modulated inductive signal output by the non-contact inductive probe. The fundamental amplitude signal related to the sample resistivity is extracted through synchronous detection, high-frequency carrier waves and environmental interference noise are filtered out, the AC inductive signal is converted into a DC equivalent voltage signal, and then converted into an equivalent resistance value, thus achieving stable calculation of the non-contact resistivity.

[0116] After the algorithm and protocol are loaded, the main control microcontroller automatically configures the gain, impedance matching, and sampling rate parameters of the measurement channel, and completes a communication handshake verification with the probe. If the handshake verification is successful, the loaded protocol, algorithm, and configuration parameters are synchronously uploaded to the host computer, completing the pre-configuration of the measurement mode; if the handshake verification fails, a protocol loading failure alarm is sent to the host computer.

[0117] Step 108: The main control microcontroller starts the constant current excitation module to output a constant current, and locks the current output accuracy through real-time sampling and closed-loop calibration.

[0118] The main control microcontroller outputs a high-level enable signal to the constant current excitation module to start the constant current excitation module. The constant current excitation module is based on an LM317 voltage regulator and a 1250Ω±0.1% low-temperature drift precision resistor to output a fixed 1.000mA constant DC current to the semiconductor sample under test, with a current ripple ≤10μA.

[0119] The main control microcontroller synchronously acquires the voltage across the sampling resistor in the constant current loop via an ADC acquisition channel, calculates the actual output current value in real time, and continuously acquires 20 sets of current data at a sampling frequency of 10Hz. The average value of the current data is calculated. When the deviation of the actual output current from the rated 1.000mA is ≤ ±0.05%, the current output is considered stable. When the deviation is > ±0.05%, the main control microcontroller automatically adjusts the fine-tuning resistor configuration of the constant current module to complete closed-loop calibration, with no more than 3 calibrations. After calibration, the constant current output parameters are locked, and the current stability verification result is synchronously uploaded to the host computer. If, after 3 calibrations, the deviation of the actual output current from the rated 1.000mA is still > ±0.05%, the constant current output is shut down, and a constant current anomaly alarm is sent to the host computer.

[0120] Step 109: The main control microcontroller constructs a constant current excitation and voltage acquisition measurement link and a temperature acquisition link, and performs link validity and data stability verification respectively.

[0121] A constant current excitation and voltage acquisition measurement link is constructed. The complete signal flow is as follows: constant current excitation module output, general measurement interface, resistance measurement probe, semiconductor sample under test, resistance measurement probe, general measurement interface, main control ADC acquisition channel, and main control data processing unit, forming a complete closed loop link.

[0122] The main control microcontroller collects the voltage signals across the semiconductor sample under test, calculates the initial resistance value of the semiconductor sample under test, continuously collects 10 sets of data, and calculates the data fluctuation variance. When the data fluctuation variance is ≤0.1%, the link is determined to be valid; when the data fluctuation variance is >0.1%, the link connection is rechecked and the verification is repeated.

[0123] A closed-loop temperature acquisition link is constructed, with the complete signal flow sequentially from the semiconductor sample under test, the K-type thermocouple, the temperature acquisition module, the signal converter, and the main control ADC acquisition channel / host computer data processing unit, forming a complete closed-loop link.

[0124] Collect the temperature data of the semiconductor sample under test at the current ambient temperature and compare it with the ambient temperature value of the standard thermometer. When the temperature deviation is ≤ ±0.2℃, the link is determined to be valid; when the temperature deviation is > ±0.2℃, re-check the thermocouple connection and cold junction compensation configuration, and repeat the verification.

[0125] Step 110: The main control microcontroller completes a comprehensive self-test. After all self-test items pass, it completes a handshake confirmation with the host computer to confirm the measurement ready state and enters the standby state for triggering measurement.

[0126] The main control microcontroller performs a comprehensive self-test across the entire chain, including hardware power supply self-test, probe identification and protocol loading self-test, constant current excitation self-test, voltage acquisition link self-test, temperature acquisition link self-test, and host computer communication self-test.

[0127] The hardware power supply self-test indicates that the output voltage deviation of all power modules is ≤±1%, with no overcurrent or overtemperature alarms. The probe identification and protocol loading self-test indicates that the probe type is correctly identified, the protocol and algorithm are loaded successfully, and the communication handshake is normal. The constant current excitation self-test indicates that the output current stability is ≤±0.05%, with no abnormal fluctuations. The voltage acquisition link self-test indicates that the signal-to-noise ratio of the acquisition channel is ≥60dB, and the data linearity meets the requirements. The temperature acquisition link self-test indicates that the temperature measurement deviation is ≤±0.2℃, and the data upload is normal. The host computer communication self-test indicates that the UART communication packet loss rate is 0, and the data uplink and downlink are normal.

[0128] Once all self-test items pass, the main control microcontroller sends a measurement ready signal to the host computer, simultaneously uploading the probe type, configuration parameters, and initial status data for this measurement to the host computer. The host computer receives the data and sends a confirmation signal. If any self-test fails, the main control microcontroller immediately terminates the process, shuts down the constant current excitation output, sends the corresponding fault code and alarm information to the host computer, and waits for manual troubleshooting before restarting from step 105. The entire process is completed when the host computer sends a confirmation signal, and the system enters a ready-to-trigger measurement state, awaiting the initiation of subsequent full-temperature range scanning and data acquisition.

[0129] Step 2 specifically includes:

[0130] Step 201: Start the constant current excitation and voltage acquisition measurement link and temperature acquisition link, initialize the relevant parameters according to the probe type, and start the ADC multi-channel continuous acquisition mode.

[0131] The main control microcontroller receives the measurement ready signal and probe configuration parameters from step 110. The probe configuration parameters include probe type, measurement protocol, and calibration algorithm. It starts the constant current excitation and voltage acquisition measurement link and temperature acquisition link, and confirms that the two links are in normal working condition and there are no abnormalities such as signal interruption or loose connection.

[0132] Based on the identified probe type, the acquisition parameters are initialized. Clamp-on electrodes and non-contact induction probes use a single-channel sampling mode, while four-probe probes and van der Bauer method probes use a multi-channel synchronous sampling mode. The resistance data acquisition resolution and temperature data acquisition resolution are both fixed at 16 bits for all probes, and the sampling reference voltage is a 3.3V precision reference source.

[0133] Initialize the sampling period baseline parameters, preset the high-frequency sampling period to 100ms, the normal sampling period to 500ms, and the low-frequency sampling period to 1s. Preset the temperature change rate threshold and the resistance fluctuation variance threshold, and initialize the basic parameters for temperature control closed-loop control. Start the ADC multi-channel continuous acquisition mode, configure the acquisition interrupt priority, and ensure the synchronization of resistance data and temperature data acquisition, with an acquisition delay ≤10ms.

[0134] The temperature change rate threshold is 5℃ / min or 1℃ / min. The resistance fluctuation variance threshold is preset according to the probe type, and ranges from 0.05% to 0.1%.

[0135] Step 202: Simultaneously acquire the resistance and temperature data of the semiconductor sample under test through the constant current excitation and voltage acquisition measurement link and the temperature acquisition link, and perform preprocessing.

[0136] Based on the constant current excitation and voltage acquisition measurement link, the ADC acquisition module acquires the voltage signal across the semiconductor sample under test in real time. The acquisition frequency is matched with the current sampling period. Five sets of voltage data are continuously acquired, and the average value is taken after removing outliers as the effective voltage value at the current moment. According to Ohm's law R=U / I, where U is the effective voltage value and I is the 1.000mA constant current output by the constant current excitation module, the total resistance value of the semiconductor sample under test at the current moment is calculated, and the resistance value and the corresponding acquisition timestamp are stored synchronously.

[0137] The initial default sampling period is 500ms.

[0138] The temperature acquisition module acquires the simulated temperature signal of the sample surface in real time through a K-type thermocouple. After processing by the signal conditioning circuit and cold junction compensation circuit, it is converted into a digital signal and uploaded to the ADC acquisition channel. It is acquired and stored synchronously with the resistance data. Similarly, five sets of temperature data are continuously acquired. After removing outliers using the 3σ principle, the average value is taken as the effective temperature value at the current moment and bound to the corresponding timestamp to ensure the temporal consistency between the resistance data and the temperature data.

[0139] The main control microcontroller performs preliminary moving average filtering on the collected resistance and temperature data. The filtering window consists of 5 sets of data to eliminate data fluctuations caused by environmental interference and signal noise, ensuring data stability.

[0140] The preprocessed effective resistance and effective temperature values ​​are uploaded to the host computer in real time according to the frame format of the corresponding measurement protocol. The upload interval is consistent with the sampling period. At the same time, they are backed up and stored in the Flash memory of the main control chip to prevent data loss.

[0141] Step 203: Calculate the temperature change rate using linear fitting and calculate the resistance fluctuation variance using the variance formula.

[0142] Based on the continuous and valid temperature data and corresponding timestamps collected in step 202, linear fitting is used to calculate the temperature change per unit time, i.e., the temperature change rate dT / dt.

[0143] Specifically, the most recent 10 sets of continuous temperature data are selected, and a linear equation for the temperature change over time, T=kt+b, is fitted. The rate of temperature change, dT / dt=60k, is calculated three times consecutively, and the average value is taken as the effective value of the rate of temperature change at the current moment.

[0144] Based on the collected continuous effective resistance data, the fluctuation variance σ of the resistance data is calculated. 2 It is used to characterize the stability of resistance data.

[0145] Specifically, select the most recent 10 consecutive sets of resistance data and calculate the average value R. avg Then calculate the variance σ. 2 =Σ(R i -R avg ) 2 / 10, where i is the i-th data set, i=1~10. Calculate 3 times consecutively and take the average value as the effective value of the resistance fluctuation variance at the current moment.

[0146] When -10℃ / min < temperature change rate < 20℃ / min and resistance fluctuation variance < 1%, the calculation results are considered reliable. When the temperature change rate > 20℃ / min or < -10℃ / min and / or resistance fluctuation variance > 1%, the calculation is considered abnormal, and data should be collected again and the calculation repeated.

[0147] Step 204: Compare the temperature change rate and resistance fluctuation variance with preset thresholds to determine the four measurement scenario states and complete the scenario state switching verification.

[0148] The main control microcontroller will use the temperature change rate dT / dt and resistance fluctuation variance σ calculated in step 203 to perform the calculation. 2 The current measurement scenario is compared with the temperature change rate threshold and resistance fluctuation variance threshold preset in step 201 to determine the current measurement scenario state, which is divided into four scenario states, including rapid temperature change state, medium-speed temperature change state, steady temperature state and natural cooling state.

[0149] When dT / dt > 5℃ / min, regardless of the resistance fluctuation variance, it is considered a rapid temperature change. At this time, the temperature of the semiconductor sample under test is in a rapid rise or fall phase, and the resistance data fluctuates significantly. When 1℃ / min < dT / dt ≤ 5℃ / min, it is considered a medium-speed temperature change. At this time, the sample temperature changes at a moderate rate, and the resistance data fluctuation is at a moderate level. When dT / dt ≤ 1℃ / min and the resistance fluctuation variance σ... 2 When the resistance fluctuation variance is less than the preset threshold, it is considered to be in a steady-temperature state. At this time, the sample temperature changes slowly, the resistance data is stable, and it meets the requirements for accurate measurement. When dT / dt < 0℃ / min and there is no heating input, it indicates that the temperature is continuously decreasing, and it is considered to be in a natural cooling state. At this time, the sample is in a natural cooling process and no heating control is required.

[0150] After determining the current measurement scene state, the main control microcontroller locks the scene state identifier and simultaneously uploads the scene state information to the host computer. At the same time, it sends scene state control commands to the adaptive sampling rate adjustment unit and the temperature control execution module. If the scene states are inconsistent for three consecutive determinations, it is considered a scene state switching transition phase. The control mode of the previous scene state is maintained until the same scene state is determined three times consecutively, at which point the corresponding control command is executed, avoiding system instability caused by frequent switching.

[0151] Step 205: Dynamically switch the sampling period according to the scene state, and the temperature control execution module synchronously outputs the corresponding heating control signal to realize closed-loop regulation.

[0152] Based on the measurement scenario state determined in step 204, the sampling temperature control unit receives control commands from the main control module and dynamically switches the sampling period. The switching process does not change the measurement protocol frame structure or the effective data bit width, ensuring that the host computer's parsing logic remains unchanged. The temperature control execution module receives control commands from the main control module and outputs corresponding heating control signals according to different scenario states, achieving closed-loop temperature control and ensuring that the sample temperature changes according to the preset process.

[0153] During temperature control adjustment, the temperature data of the semiconductor sample under test is fed back in real time. The main control microcontroller compares the actual temperature with the target temperature and calibrates the PWM duty cycle and relay on / off status in real time to ensure that the temperature control accuracy and temperature change trend meet the requirements. When the temperature deviation exceeds ±0.5℃, a temperature control abnormality alarm is triggered and simultaneously uploaded to the host computer.

[0154] Specifically, when the measurement scenario involves rapid temperature changes, a 100ms high-frequency sampling period is switched to increase the sampling frequency, capture transient changes in sample temperature and resistance, and avoid data loss. Simultaneously, the main control microcontroller outputs a fixed high or low level to control the relays to be fully on or off, enabling rapid temperature rise and fall of the semiconductor sample and shortening the temperature change time. When the measurement scenario involves medium-speed temperature changes, a 500ms conventional sampling period is switched to balance data acquisition details with communication bandwidth usage. The main control microcontroller outputs a linearly adjustable PWM signal to gradually adjust the relay on / off ratio, achieving linear and stable temperature changes and avoiding excessive temperature fluctuations. When the measurement scenario involves a steady-temperature state, a 1s low-frequency precision sampling period is switched to reduce data volume, system power consumption, and communication bandwidth usage while ensuring measurement accuracy. Based on the deviation between the current sample temperature and the preset target temperature range, the main control microcontroller uses a PID algorithm to adjust the PWM duty cycle in real time, controlling the relay on / off ratio to achieve high-precision constant temperature control of ±0.1℃, ensuring the sample temperature remains stable within the target range. When the measured scenario is in a natural cooling state, the system switches to a 1-second low-frequency power-saving sampling period, shuts down the heating output, and further reduces system power consumption. After the sampling period switch, the ADC acquisition frequency and data upload interval are adjusted synchronously to ensure that the sampling timing is consistent with the upload timing; the main control microcontroller outputs a low level to shut down the heating output, stop the PTC heater from working, and allow the semiconductor sample under test to cool down naturally, while reducing the sampling rate and saving system power consumption.

[0155] Step 206: Confirm the full temperature range scanning range. During the scanning process, dynamically adapt the scene state, sampling period and temperature control mode according to temperature changes, and cyclically determine the temperature state of the semiconductor sample under test.

[0156] The main control microcontroller pre-programs the full-temperature scanning range and defines the target temperature range. Based on the sampling cycle switching and temperature control closed-loop adjustment in step 205, the temperature of the semiconductor sample under test starts from room temperature and completes the full-temperature scanning according to the preset trend. During the scanning process, the scene state, sampling cycle, and temperature control mode are switched in real time according to temperature changes to achieve dynamic adaptation of the entire scanning process and ensure that the measurement data of different temperature zones meet the accuracy requirements. The main control microcontroller determines in real time whether the current temperature of the semiconductor sample under test has reached the end point of the full-temperature scanning or entered the target temperature range. If the temperature of the semiconductor sample under test has not entered the target temperature range and has not reached the scanning end point, it returns to step 202 and repeats the cycle of data acquisition, criterion calculation, scene state determination, sampling, and temperature control adjustment to continuously advance the full-temperature scanning. When the temperature of the semiconductor sample under test enters the target temperature range and is determined to be in a stable temperature state, the full-temperature scanning cycle is paused, and step 3 is entered to enter the stable temperature point measurement stage. When the temperature of the semiconductor sample under test reaches the end point of the full-temperature scanning, the full-temperature scanning is completed, the constant current excitation and heating link are turned off, and the process ends.

[0157] The full-temperature scanning range described in this invention is by default room temperature to 120°C, and can be adjusted via a host computer configuration. The target temperature range is by default 80°C to 100°C, which is the stable temperature measurement range.

[0158] Step 3 specifically includes:

[0159] Step 301: Confirm the effectiveness of the steady-state condition, calibrate the measurement parameters that match the probe type, initialize the acquisition of relevant parameters, and start the intrinsic resistance extraction and multi-physical quantity acquisition process.

[0160] The main control microcontroller receives the trigger signal for entering the steady-state state within the target temperature range. First, it confirms the validity of the current steady-state state by continuously collecting 10 sets of sample temperature data to verify that the temperature fluctuation is ≤ ±0.1℃, the temperature change rate dT / dt ≤ 1℃ / min, and the resistance fluctuation variance σ. 2 <The resistance fluctuation variance threshold preset for the corresponding probe. If all three conditions are met, it is determined to be a stable temperature state. Otherwise, return to step 205 to continue temperature control adjustment until a stable temperature state is reached.

[0161] Based on the probe type and calibration algorithm, the measurement parameters are calibrated. Specifically, for clamp-on electrodes, the effective lengths L1 and L2 and the sample cross-sectional area S parameters of the contact resistance compensation algorithm are calibrated. For four-probe probes, the probe spacing s and the thin-layer correction coefficient C are calibrated. t Parameters. Van der Burg method probe calibration geometric correction factor G, electrode symmetry parameters. Non-contact induction probe calibration coupling distance standard value d0 and sensitivity calibration curve parameters.

[0162] Set the number of resistance measurements, data acquisition interval, multi-physical quantity acquisition frequency, and data encapsulation frame format at the same stable temperature point. Start the intrinsic resistance extraction and multi-physical quantity acquisition process, send an acquisition enable command to the multi-physical quantity synchronous acquisition module, and confirm that the module is in a ready state.

[0163] Step 302: Collect no less than two sets of total resistance data under the same steady-state temperature, bind the timestamps, and handle any collection anomalies.

[0164] Based on the type of resistance measurement probe, the corresponding total resistance measurement is performed, ensuring that temperature fluctuations at the same stable temperature point are ≤±0.05℃. Specifically, for the clamp-type electrode, a constant current excitation current of 1.000mA is maintained. At the effective length L1, five sets of voltage data across the sample are continuously collected. Outliers are removed according to the 3σ principle, and the average value is calculated using Ohm's law R=U / I to obtain the first set of total resistance R1. Subsequently, while maintaining the clamping pressure and contact state, the electrode is moved to the effective length L2, and the same acquisition and calculation method is used to obtain the second set of total resistance R2. For the four-probe probe, a constant current excitation current is maintained. The probe excitation and voltage acquisition are controlled by the four-probe resistivity measurement standard protocol, and five sets of four-channel voltage data are continuously collected. After correction, the total resistance value at the current stable temperature point is calculated. This process is repeated once to obtain two sets of total resistance data, R1 and R2. The Van der Bauer method probe controls the switching of forward and reverse excitation of the electrodes through the Van der Bauer method symmetrical measurement protocol, acquiring two sets of forward and reverse voltage data. After processing by the contact resistance correction algorithm, two sets of total resistance data R1 and R2 are calculated. The non-contact induction probe maintains a constant coupling distance and controls the induction excitation and signal acquisition through a non-contact induction measurement protocol. After coupling distance error correction and induction signal demodulation, two sets of total resistance data R1 and R2 are obtained.

[0165] After each group of total resistance measurements is completed, the corresponding acquisition timestamp is recorded synchronously and bound to the current stable temperature point data to ensure time sequence consistency. If the fluctuation of a group of total resistance data exceeds the preset resistance fluctuation variance threshold, the data for that group is reacquired until the requirements are met; if three consecutive acquisitions are abnormal, a measurement abnormality alarm is sent to the host computer, the process is paused, and the process awaits investigation.

[0166] Step 303: Call the corresponding correction algorithm, cancel the system error through difference operation, extract the intrinsic resistance value of the sample and bind it to the temperature.

[0167] The contact resistance correction / compensation algorithm of the corresponding resistance measurement probe is invoked, and the difference calculation is performed based on the two sets of total resistance data R1 and R2 collected in step 302.

[0168] Specifically, the clamp-on electrode uses a contact resistance compensation algorithm based on temperature difference to calculate the difference ΔR = R1 - R2, offsetting the contact resistance component. Then, the intrinsic resistivity of the semiconductor sample under test is calculated using the formula ρ = ΔR·S / (L1 - L2), thus obtaining the intrinsic resistance value. The four-probe probe uses a thin-layer resistivity correction formula and a probe spacing correction algorithm to correct the two sets of total resistance data separately, taking the average as the intrinsic resistance value. The van der Bauer method probe performs geometric factor correction and contact resistance correction on the two sets of total resistance data separately, taking the average as the intrinsic resistance value. The non-contact inductive probe corrects the coupling spacing error on the two sets of total resistance data, taking the average as the intrinsic resistance value.

[0169] The extracted intrinsic resistance value is compared with the theoretical resistance range at the current stable temperature. If it exceeds the preset theoretical range, it is determined to be an extraction anomaly, and steps 302 to 303 are repeated until a qualified intrinsic resistance value is obtained. The extracted intrinsic resistance value is bound with the effective temperature value of the current stable temperature point and the acquisition timestamp to form a temperature-intrinsic resistance value pair, which is stored in the Flash memory of the main control chip, and a data validity identifier is marked.

[0170] The theoretical range is preset based on the sample material.

[0171] Step 304: Synchronously collect multiple physical quantity parameters, complete outlier removal, range calibration, and timing binding.

[0172] Three sets of data were continuously collected for each physical quantity parameter. Outliers were removed using the 3σ principle, and the average value was taken as the valid value of the multi-physical quantity parameter at the current moment. Based on the preset calibration curve, the range of the collected multi-physical quantity parameters was calibrated to eliminate the error of the acquisition module itself, and the parameter values ​​were converted into a standardized data format that matches the measurement protocol frame structure.

[0173] The multiple physical parameters include ambient humidity, sample clamping pressure, ambient temperature, and constant current excitation stability parameters.

[0174] The ambient humidity accuracy is ±1%RH, and the sampling range is 0–90%RH. The sample clamping pressure is only applicable to contact probes, with an accuracy of ±0.1N. The ambient temperature is subject to auxiliary calibration, with an accuracy of ±0.2℃. The constant current excitation stability parameter refers to the current fluctuation value.

[0175] The preprocessed valid multi-physical quantity parameters are bound to the timestamps of the corresponding temperature intrinsic resistance values ​​to ensure the timing consistency of multi-source data and avoid timing mismatches that could cause subsequent error correction to fail. Simultaneously, the bound multi-physical quantity parameters and temperature intrinsic resistance values ​​are backed up to the main control chip's internal Flash memory to prevent data loss.

[0176] Step 305: Complete the multi-source data slice encapsulation and CRC16 verification according to the fixed frame length of the corresponding measurement protocol.

[0177] The fixed frame length is consistent with the frame length when the sampling period is switched in step 2. In this invention, it is fixed at 16 bytes to ensure that the parsing logic of the host computer remains unchanged. The intrinsic temperature resistance value pairs and multiple physical quantity parameters are allocated to the frame structure according to the preset fields. Each frame contains a complete set of intrinsic temperature resistance value pairs and corresponding multiple physical quantity parameters.

[0178] The frame structure includes a frame header, slave address, function instruction code, temperature data segment, intrinsic resistance value segment, multiple physical quantity parameter segment, reserved bits, and CRC16 check bits. The check range is from the frame header to the reserved bits, ensuring the accuracy of data encapsulation.

[0179] After each frame of data is encapsulated, the data encapsulation and communication unit automatically performs a CRC16 check. If the check fails, the data frame is re-encapsulated. If the encapsulation check fails three times in a row, an encapsulation error alarm is sent to the host computer.

[0180] The data encapsulation and communication unit synchronously uploads the verified encapsulated data to the host computer via the UART serial port. The upload interval is consistent with the steady-state sampling period to ensure real-time data upload. After receiving the encapsulated data, the host computer verifies the frame format and CRC16 checksum. If the verification is successful, it sends an acknowledgment signal. If the main control module does not receive an acknowledgment signal within 100ms, it re-uploads the frame data, repeating the upload no more than 3 times. If no acknowledgment is received, an upload error alarm is sent.

[0181] Step 4 specifically includes:

[0182] Step 401: Match the corresponding multi-source environmental error according to the type of resistance measurement probe.

[0183] Based on the material of the semiconductor sample under test and the type of resistance measurement probe, the corresponding environmental influencing factors are matched.

[0184] Specifically, clamp-on, four-probe, and van der Burg method contact probes include those for humidity, pressure, ambient temperature, and constant current fluctuation. Non-contact sensing probes include those for humidity, ambient temperature, and constant current fluctuation.

[0185] Step 402: Perform error correction step by step in sequence to eliminate environmental interference errors one by one and obtain the calibrated intrinsic resistance value.

[0186] The order is as follows: ambient temperature interference correction, constant current excitation fluctuation correction, ambient humidity correction, and clamping pressure correction. For non-contact probes, clamping pressure correction is skipped.

[0187] Specifically, based on the difference ΔT between ambient temperature and sample temperature. envCalculate the resistance error ΔR caused by ambient temperature. env =k1×ΔT env Where k1 is the ambient temperature correction coefficient, based on the sample material preset. Subtract ΔR from the intrinsic resistance value R. env The resistance value R1 after eliminating the interference of ambient temperature is obtained.

[0188] Based on the constant current excitation fluctuation value ΔI, calculate the resistance error ΔR caused by the current fluctuation. I =R×(ΔI / I0), where I0 is the rated constant current of 1.000mA, and R1 minus ΔR I The resistance value R2 after eliminating current fluctuations is obtained.

[0189] Based on the ambient humidity value H, the humidity error ΔR is obtained by querying the humidity resistance error calibration curve. H Subtract ΔR from R2 H The resistance value R3 after eliminating humidity interference was obtained. The humidity resistance error calibration curve is the curve calibrated during the actual experiment.

[0190] Based on the clamping pressure value F, calculate the resistance error ΔR caused by pressure fluctuation. F =k2×(F0-F), where k2 is the pressure correction coefficient, F0 is the standard clamping pressure, and R3 is subtracted from ΔR. F The final calibrated intrinsic resistance value R after eliminating pressure interference is obtained. cal .

[0191] Step 403: Convert the calibrated intrinsic resistance value into intrinsic resistivity, and solve for the bandgap value through linear fitting.

[0192] The intrinsic resistance value R will be calibrated. cal Converted to intrinsic resistivity ρ=R cal ×S / L, where S is the sample cross-sectional area and L is the effective measurement length.

[0193] Calculate the bandgap value at the corresponding temperature: Where ρ is the intrinsic resistivity of the semiconductor sample under test; ρ0 is the inherent resistivity of the semiconductor sample under test, based on material presets; Eg is the bandgap value; and k is the Boltzmann constant, with a value of 1.38 × 10⁻⁶. -23 J / K; T is the absolute temperature of the semiconductor sample under test, in K.

[0194] Taking the natural logarithm of both sides of the formula, we get The bandgap width Eg is solved by linear fitting, so that... x-axis Using y as the ordinate, perform a linear fit on all valid data pairs to obtain the slope m of the fitted line. Calculate the bandgap width Eg = 2k·m, where k is the Boltzmann constant.

[0195] Step 404: Output the final measurement report in multiple formats.

[0196] The final measurement report is saved in Excel format. The final measurement report includes information such as the semiconductor sample under test, measurement parameters, calibration data, calculation process, final bandgap value, calculation error, and measurement time.

[0197] Users can export measurement reports to PDF or Excel format for easy archiving, analysis, and sharing.

[0198] A bandgap width measuring device is used to perform the bandgap width measuring method, such as... Figure 2 As shown, the system includes a link construction unit, a sampling and temperature control unit, a resistance measurement unit, and a processing and calculation unit. The link construction unit thermally couples and fixes the semiconductor sample under test (SUT) to the temperature sensor, connects the resistance measurement probe to a universal measurement interface, identifies the type of the resistance measurement probe and loads the matching measurement protocol and correction algorithm, applies a constant current excitation to the SUT, and establishes a constant current excitation and voltage acquisition measurement link, as well as a temperature acquisition link. The sampling and temperature control unit, based on the constant current excitation and voltage acquisition measurement link, acquires the voltage data of the SUT in real time and converts it into resistance data. Simultaneously, it acquires the temperature data of the SUT in real time through the temperature acquisition link, calculates the temperature change rate and resistance fluctuation variance, determines the measurement scenario state and switches the corresponding sampling period, and synchronously adjusts the heating link to achieve closed-loop temperature control, completing the dynamic adaptation of the full-temperature scanning process. The resistance measurement unit performs at least two sets of total resistance measurements at various stable temperature points within the target temperature range, using differential calculations to obtain the intrinsic resistance value of the semiconductor sample under test, resulting in a temperature-intrinsic resistance value pair. Simultaneously, it acquires multiple physical quantity parameters and dynamically slices the temperature-intrinsic resistance value pair and the multiple physical quantity parameters according to fixed frames. The processing and calculation unit performs timestamp alignment, quality screening, and environmental error correction on the temperature-intrinsic resistance value pair and the multiple physical quantity parameters. Based on the corrected temperature-intrinsic resistance value pair, it constructs a linear relationship, calculates the bandgap of the semiconductor sample under test based on the slope, and outputs the final measurement report.

[0199] The following detailed description is provided with reference to specific embodiments:

[0200] The method of this invention is used to measure the bandgap of germanium block semiconductors, such as... Figure 3 As shown, the sample to be tested is an intrinsic germanium block with a purity ≥99.99%, and its dimensions are 30mm × 5mm × 3mm, with a cross-sectional area S = 15mm². 2Use anhydrous ethanol to soak a microfiber lint-free wiping cloth and repeatedly wipe the upper and lower heat-conducting surfaces, the two end faces along the length direction, and the effective measurement area of ​​the germanium semiconductor 3, wiping each surface at least three times. For areas with oxide layers, lightly polish them in one direction with 1500-grit metallographic sandpaper, clean them again with anhydrous ethanol, and let them air dry naturally in a dust-free environment for at least 30 seconds.

[0201] Apply a uniform layer of high thermal conductivity silicone grease (≥2.0 W / (m·K)) to the center of the working surface of the PTC heater 8, with a thickness of 0.1–0.2 mm. Place the germanium semiconductor 3 horizontally on the grease-coated area, gently press to squeeze out excess grease and air gaps, and use high-temperature resistant polyimide tape for further fixation. Secure the probe end of a Class I precision K-type armored thermocouple 9 to the geometric center of the upper surface of the germanium block using high thermal conductivity silicone grease. The probe end wire length is 2–3 mm. Secure the surrounding 2 mm area with high-temperature resistant silicone to ensure stable thermal coupling. Connect the thermocouple signal line to the temperature acquisition module, and transmit the signal to the host computer via a USB 18745 signal converter 10.

[0202] Low-contact-resistance alligator clip electrodes are selected and clamped onto the two end faces along the length of the germanium semiconductor 3, with consistent clamping depth and complete contact between the contact surface and the end face. The probe signal line is connected to a general-purpose measurement interface module. The main control microcontroller 4 outputs a 1kHz, 5mV square wave detection signal through a DAC, acquires the probe response characteristics, calculates parameters such as input impedance, amplitude-frequency characteristic attenuation, and rise time, and matches them with the probe feature library to identify it as a clamp-on electrode with a matching degree ≥95%.

[0203] The main control microcontroller 4 automatically loads the standard communication protocol for two-end resistance measurement and the contact resistance compensation algorithm for temperature difference, configures the ADC to differential input mode with a gain of 1x, and completes communication handshake verification with the host computer. It then starts the constant current excitation module 2, outputting a constant current of 1.000mA. Real-time sampling and calibration via the ADC ensures the current deviation is ≤±0.05%, establishing the constant current excitation and voltage acquisition measurement link. After completing hardware power supply, probe identification, constant current excitation, voltage / temperature acquisition link, and host computer communication self-tests, and entering the ready-to-trigger measurement state after all tests are passed, the system proceeds to the ready-to-trigger measurement state.

[0204] The high-frequency sampling period is set to 100ms, the normal sampling period to 500ms, and the low-frequency sampling period to 1s; the preset temperature change rate thresholds are 5℃ / min and 1℃ / min, and the resistance fluctuation variance threshold is 0.1%. The ADC multi-channel continuous acquisition mode is activated, with an acquisition delay ≤10ms. Initially, a normal sampling period of 500ms is used to simultaneously acquire voltage and temperature data across the germanium semiconductor. Five sets of data are continuously acquired per group, outliers are removed, and the average is taken to calculate the total resistance value. After moving average filtering, the data is uploaded to the host computer and backed up locally.

[0205] The master control single-chip microcomputer calculates the temperature change rate dT / dt and the resistance fluctuation variance σ in real time 2 . When dT / dt > 5℃ / min, it is determined as a rapid temperature change state, and the sampling period is switched to 100 ms high-frequency sampling, and the PTC heater operates at full power. When 1 < dT / dt ≤ 5℃ / min, it is determined as a medium-speed temperature change state, and the 500 ms conventional sampling period is maintained, and the heating is linearly adjusted by PWM. When dT / dt ≤ 1℃ / min and σ 2 < 0.1%, it is determined as a stable temperature state, and the sampling period is switched to 1 s low-frequency sampling, and the PID control realizes a constant temperature of ±0.1℃

[0206] The full temperature range scanning range is set from room temperature to 120℃, and the target temperature range is 80℃ - 100℃. During the heating process, the scene state and the sampling period are dynamically switched according to the temperature change until entering the target temperature range and stabilizing in the stable temperature state

[0207] Continuously collect 10 groups of temperature data. After meeting the requirements of temperature fluctuation ≤ ±0.1℃, dT / dt ≤ 1℃ / min, and resistance fluctuation variance < 0.1%, enter the stable temperature point measurement process

[0208] At the stable temperature point T = 85.0℃, the electrode is clamped at the effective length L1 = 25 mm as the first measurement position, and 5 groups of voltages are continuously collected. The total resistance R1 = 12.345Ω is calculated. The electrode is moved to the effective length L2 = 20 mm as the second measurement position. Keeping the temperature unchanged, the total resistance R2 = 10.123Ω is collected and calculated

[0209] Adopt the same temperature difference contact resistance compensation algorithm , the effective length difference ΔL = L1 - L2 = 5 mm, and the sample cross-sectional area S = 15 mm 2 , calculate the intrinsic resistivity , the intrinsic resistance value at this stable temperature point is or directly use ρ for subsequent fitting

[0210] Simultaneously collect parameters such as environmental humidity 45%RH, environmental temperature 24.5℃, and constant current excitation stability, with current fluctuation ≤ 0.02%. Bind the time stamps to the resistance and temperature data, encapsulate them in a 16-byte fixed frame structure, and upload them to the host computer after adding CRC16 checksum. In the target temperature range of 80℃ - 100℃, repeat the above process every 2 - 3℃, and collect the intrinsic resistance values and corresponding temperatures of 8 - 10 stable temperature points in total

[0211] The host computer matches the resistance data uploaded by the single-chip microcomputer with the temperature data uploaded by the temperature acquisition module according to the time stamp, forms a one-to-one corresponding temperature intrinsic resistance value pair, and eliminates the abnormally fluctuating points

[0212] The ambient temperature interference correction, constant current excitation fluctuation correction, and ambient humidity correction are performed sequentially. For contact probes, clamping pressure correction is also performed to obtain the calibrated intrinsic resistance value R. cal The temperature T at each steady-state point is compared with the corresponding intrinsic resistance R. cal Convert to and Data, perform linear fitting, The fitting slope is m = 0.6329 eV / 2k. B Then the bandwidth ,like Figure 4 As shown.

[0213] Output an Excel-formatted measurement report, including sample information, measurement parameters, data at each steady-state temperature, fitted curve, bandgap width value, and fit confidence level.

[0214] In this embodiment, the intrinsic bandgap of germanium at around 85°C was measured to be 0.6329 eV, which is in high agreement with the literature value and the relative error is <0.2%, verifying the accuracy of the method and the effectiveness of contact resistance compensation.

[0215] This invention provides a method and apparatus for measuring bandgap width. First, at the same temperature point, two sets of total resistances with different effective lengths are obtained by changing the clamping position. The difference is then used to directly cancel out contact resistance terms unrelated to the effective length, thereby accurately extracting the intrinsic resistance value of the sample, eliminating contact resistance interference, and improving measurement accuracy and result consistency. Second, by calculating the temperature change rate and resistance fluctuation variance in real time, four measurement scenarios are dynamically determined, and the sampling period is automatically switched according to the scenario, providing a stable and controllable temperature environment for bandgap width measurement. Third, by outputting a standard detection excitation signal and acquiring probe response characteristics, multiple measurement probes can be automatically identified, and matching measurement protocols and correction algorithms can be automatically loaded, significantly improving the compatibility and scenario adaptability of the apparatus. Finally, strict timestamp matching is performed on the resistance and temperature data to ensure that each set of resistance values ​​corresponds to the temperature at its actual sampling time, reducing interference from non-intrinsic conductivity ranges or noise data on the fitting results.

[0216] In summary, this invention systematically solves the prominent problems in existing bandgap measurement technologies, such as contact resistance interference, the contradiction between sampling efficiency and accuracy, poor scenario adaptability, and single measurement dimension, at both the principle and engineering implementation levels. It provides an efficient, reliable, and easy-to-promote technical solution for the accurate measurement of the bandgap of semiconductor materials.

[0217] The above description illustrates preferred embodiments of the present invention and helps those skilled in the art to more fully understand the technical solution of the present invention. However, these embodiments are merely illustrative and should not be construed as limiting the specific implementation of the present invention to these embodiments. For those skilled in the art, several simple deductions and modifications can be made without departing from the inventive concept, and all such modifications should be considered within the protection scope of the present invention.

Claims

1. A method for measuring band gap width, characterized in that, Includes the following steps: The semiconductor sample under test is thermally coupled and fixed to the temperature sensor. The resistance measurement probe is connected to the general measurement interface. The type of resistance measurement probe is identified and the matching measurement protocol and correction algorithm are loaded. A constant current excitation is applied to the semiconductor sample under test, and a constant current excitation and voltage acquisition measurement link and a temperature acquisition link are established. The voltage data of the semiconductor sample under test is collected in real time by constant current excitation and voltage acquisition measurement link and converted into resistance data. At the same time, the temperature data of the semiconductor sample under test is collected in real time by temperature acquisition link. The temperature change rate and resistance fluctuation variance are calculated, the measurement scenario state is determined and the corresponding sampling period is switched. The heating link is adjusted synchronously to realize temperature closed-loop control and complete the dynamic adaptation of the full temperature range scanning process. At each stable temperature point within the target temperature range, complete at least two sets of total resistance measurements with different effective lengths. Obtain the intrinsic resistance value of the semiconductor sample under test through difference calculation to obtain the temperature intrinsic resistance value pair. Simultaneously acquire multiple physical quantity parameters and dynamically slice the temperature intrinsic resistance value pair and multiple physical quantity parameters according to a fixed frame. The system performs time-stamp alignment, quality screening, and environmental error correction on temperature intrinsic resistance pairs and multiple physical parameters. Based on the corrected temperature intrinsic resistance pairs, a linear relationship is constructed. The bandgap of the semiconductor sample under test is calculated according to the slope, and the final measurement report is output.

2. The method for measuring band gap width according to claim 1, characterized in that, The process of thermally coupling and fixing the semiconductor sample under test with the temperature sensor, connecting the resistance measurement probe to a universal measurement interface, identifying the type of the resistance measurement probe and loading a matching measurement protocol and calibration algorithm, applying a constant current excitation to the semiconductor sample under test, and establishing a constant current excitation and voltage acquisition measurement link and a temperature acquisition link includes: Standardized pretreatment is performed on the semiconductor samples to be tested to remove surface oxide layers and contaminants; The semiconductor sample under test is thermally coupled and fixed to the heating device by using high thermal conductivity silicone grease to establish a low-resistance continuous heat conduction path. The temperature sensor probe is thermally coupled and fixed to the geometric center of the upper surface of the semiconductor sample under test to establish a temperature signal acquisition and transmission path. The resistance measurement probe is physically mounted to the semiconductor sample under test, and a standardized electrical connection is made to the general measurement interface module. Power on the low-voltage circuit, constant current excitation circuit, and heating circuit in stages, and simultaneously complete the basic initialization of all hardware peripherals and communication interfaces. By outputting a standard detection excitation signal and acquiring probe response characteristics, the characteristic parameters of the resistance measurement probe are extracted and matched with the probe feature library to identify the type of resistance measurement probe. Based on the type of resistance measurement probe, the system automatically retrieves and loads the matching measurement communication protocol and error correction algorithm to complete the configuration of measurement channel parameters. The constant current excitation module is activated to output a constant current, and the current output accuracy is locked through real-time sampling and closed-loop calibration. Construct constant current excitation and voltage acquisition measurement links, and temperature acquisition links, and verify the effectiveness and data stability of the links respectively.

3. The method for measuring the band gap width according to claim 2, characterized in that, The specific types of resistance measurement probes include: The main control microcontroller outputs a standard detection excitation signal to the resistance measurement probe through a DAC peripheral and synchronously acquires the response waveform of the resistance measurement probe through an ADC acquisition channel. Based on the response waveform, the core characteristic parameters of the resistance measurement probe are calculated. The core characteristic parameters include input impedance, amplitude-frequency characteristic attenuation, and signal rise time. The core characteristic parameters are then normalized and matched against a probe feature library. When the matching degree between the core characteristic parameters of the resistance measurement probe and the standard characteristic parameter range of a certain type of probe pre-stored in the probe feature library is greater than or equal to a first matching threshold, the probe is determined to be of the corresponding type.

4. The method for measuring the band gap width according to claim 3, characterized in that, The types of resistance measurement probes include clamp-on electrodes, four-probe probes, van der Burg method probes, and non-contact induction probes.

5. The method for measuring the band gap width according to claim 1, characterized in that, The constant current excitation and voltage acquisition measurement link acquires the voltage data of the semiconductor sample under test in real time and converts it into resistance data. Simultaneously, the temperature data of the semiconductor sample under test is acquired in real time through the temperature acquisition link. The temperature change rate and resistance fluctuation variance are calculated, the measurement scenario state is determined and the corresponding sampling period is switched. The heating link is adjusted synchronously to achieve closed-loop temperature control, and the dynamic adaptation of the full temperature range scanning process is completed, including: Start the constant current excitation and voltage acquisition measurement link and temperature acquisition link, initialize the relevant parameters according to the probe type, and start the ADC multi-channel continuous acquisition mode; The resistance and temperature data of the semiconductor sample under test are simultaneously acquired through constant current excitation and voltage acquisition measurement links and temperature acquisition links, and then preprocessed. Linear fitting is used to calculate the rate of temperature change, and the variance of resistance fluctuation is calculated using the variance formula. The temperature change rate and resistance fluctuation variance are compared with preset thresholds to determine four measurement scenario states and complete the scenario state switching verification. The sampling period is dynamically switched according to the scene status, and the temperature control execution module synchronously outputs the corresponding heating control signal to achieve closed-loop regulation. Confirm the full temperature range scanning range, dynamically adapt the scene state, sampling period and temperature control mode according to temperature changes during the scanning process, and cyclically determine the temperature state of the semiconductor sample under test.

6. The method for measuring the band gap width according to claim 5, characterized in that, Determining the measurement scene state and switching the corresponding sampling period includes: The calculated rate of temperature change and the variance of resistance fluctuation are compared with preset thresholds to determine whether it is a rapid temperature change state, a medium-speed temperature change state, a steady temperature state, or a natural cooling state. When a rapid temperature change is detected, switch to a high-frequency sampling period; when a medium-speed temperature change is detected, switch to a normal sampling period; when a steady-temperature state or a natural cooling state is detected, switch to a low-frequency sampling period. The switching process of the sampling period does not change the frame structure and effective data bit width of the measurement protocol.

7. The method for measuring the band gap width according to claim 1, characterized in that, The process involves performing at least two sets of total resistance measurements at each stable temperature point within the target temperature range, obtaining the intrinsic resistance value of the semiconductor sample under test through difference calculation, obtaining temperature intrinsic resistance value pairs, simultaneously acquiring multiple physical quantity parameters, and dynamically slicing the temperature intrinsic resistance value pairs and multiple physical quantity parameters according to fixed frames, including: Confirm the effectiveness of the steady-state test, calibrate the measurement parameters that match the probe type, initialize the acquisition of relevant parameters, and start the intrinsic resistance extraction and multi-physical quantity acquisition process; Collect at least two sets of total resistance data under the same steady-state temperature, bind timestamps, and handle collection anomalies; The corresponding correction algorithm is invoked to offset systematic errors through difference calculation, and the intrinsic resistance value of the sample is extracted and bound to temperature. Simultaneously collect multiple physical quantity parameters to complete outlier removal, range calibration, and timing binding; Multi-source data slice encapsulation and CRC16 verification are completed according to the fixed frame length of the corresponding measurement protocol.

8. The method for measuring the band gap width according to claim 7, characterized in that, Extracting the intrinsic resistance value of the sample includes: obtaining at least two total resistance measurements by changing the clamping position to form different effective lengths at the same stable temperature point, and calculating the difference between the at least two total resistance measurements at the same temperature point to offset the contact resistance.

9. A method for measuring the band gap width according to claim 8, characterized in that, The synchronously acquired physical parameters include ambient humidity, sample clamping pressure, ambient temperature, and constant current excitation stability parameters. The multiple physical quantity parameters and the intrinsic temperature resistance value are bound to the same timestamp, and after being encapsulated with a fixed frame length, they are subjected to CRC verification before being uploaded to the host computer.

10. The method for measuring band gap width according to claim 1, characterized in that, The process involves aligning temperature intrinsic resistance pairs and multiple physical parameters with timestamps, performing quality screening and environmental error correction, constructing a linear relationship based on the corrected temperature intrinsic resistance pairs, calculating the bandgap of the semiconductor sample under test based on the slope, and outputting a final measurement report including: Match the corresponding multi-source environmental error according to the type of resistance measurement probe; The error correction is performed sequentially step by step to eliminate environmental interference errors one by one and obtain the calibrated intrinsic resistance value. The calibrated intrinsic resistance value is converted into intrinsic resistivity, and the bandgap value is solved by linear fitting. The final measurement report is output in multiple formats.

11. The method for measuring band gap width according to claim 10, characterized in that, Converting the calibrated intrinsic resistance value to intrinsic resistivity, and then solving for the bandgap value through linear fitting includes: The corrected intrinsic resistance value is converted to intrinsic resistivity ρ=R. cal ×S / L, where S is the sample cross-sectional area and L is the effective measurement length, expressed as the reciprocal of the absolute temperature. The x-axis represents the natural logarithm of intrinsic resistivity. A linear fit is performed on the ordinate to obtain the fitting slope m, and the band gap width Eg = 2k·m is calculated, where k is the Boltzmann constant.

12. A bandgap width measuring device, used to perform the bandgap width measuring method, characterized in that, It includes a link construction unit, a sampling temperature control unit, a resistance measurement unit, and a processing and calculation unit; The link construction unit thermally couples and fixes the semiconductor sample under test with the temperature sensor, connects the resistance measurement probe to the general measurement interface, identifies the type of the resistance measurement probe and loads the matching measurement protocol and correction algorithm, applies constant current excitation to the semiconductor sample under test, and establishes a constant current excitation and voltage acquisition measurement link and a temperature acquisition link. The sampling temperature control unit collects the voltage data of the semiconductor sample under test in real time based on the constant current excitation and voltage acquisition measurement link and converts it into resistance data. Simultaneously, it collects the temperature data of the semiconductor sample under test in real time through the temperature acquisition link, calculates the temperature change rate and resistance fluctuation variance, determines the measurement scenario state and switches the corresponding sampling period, and synchronously adjusts the heating link to achieve temperature closed-loop control, thus completing the dynamic adaptation of the full temperature range scanning process. The resistance measurement unit completes at least two sets of total resistance measurements with different effective lengths at each stable temperature point within the target temperature range. It obtains the intrinsic resistance value of the semiconductor sample under test through difference calculation, and obtains the temperature intrinsic resistance value pair. It also collects multiple physical quantity parameters simultaneously and dynamically slices the temperature intrinsic resistance value pair and the multiple physical quantity parameters according to a fixed frame. The processing and calculation unit performs time-stamp alignment, quality screening, and environmental error correction on the temperature intrinsic resistance value pairs and multiple physical quantity parameters. Based on the corrected temperature intrinsic resistance value pairs, it constructs a linear relationship, calculates the bandgap of the semiconductor sample under test according to the slope, and outputs the final measurement report.

Images