Liquid particle counting device based on multi-detector fusion and detection method

Abstract

The invention discloses a liquid particle counting device based on multi-detector fusion and a detection method, and belongs to the field of liquid particle counting. The device comprises a polarization laser light source module, a sample flow and control module, a dual-channel optical collection module, a dual-mode detector detection module and a data processing fusion module. The SPAD array detection channel and the APD array detection channel are arranged, and output signals of the SPAD array detection channel and the APD array detection channel are processed and fused through the data processing fusion module, so that detection and particle size inversion of particles in a wide particle size range can be realized, the linear interval limitation of a single type of detector is broken through, and the detection accuracy is improved. And the particle size of the particles can be accurately detected in a wide-range manner.

Description

Technical Field

[0001] This invention relates to a liquid particle counting device and detection method based on multi-detector fusion, belonging to the field of liquid particle counting. Background Technology

[0002] Liquid particle counting technology is a key analytical tool in high-end industrial and scientific research fields such as semiconductor manufacturing, biomedicine, and nanomaterials science. Its core is to rapidly, non-destructively, and with high precision measure the particle size distribution and number concentration of suspended particles in liquids through the principle of optical scattering. Traditional light scattering particle counters typically use a single type of photodetector (such as a conventional avalanche photodiode (APD) or a single-photon avalanche diode (SPAD)) to receive the scattered light signal generated by the particles, and inversely determine the particle size through the theoretical or empirical relationship between light intensity and particle size.

[0003] However, liquid particle counters that use only one type of photodetector as described above face significant technical bottlenecks in practical applications: 1. Limited particle size detection range: Conventional avalanche photodiodes (APDs) have limited sensitivity and cannot effectively distinguish between signal and noise. For ultrafine particles with a diameter of less than 50nm, the scattered light signal is extremely weak and is often submerged in system noise (such as background light fluctuations, stray light from optical elements, and scattered light from water molecules). As a result, the detection limit is usually limited to the range of 50nm–200nm, which cannot meet the current detection requirements for particles smaller than 50nm. While single-photon avalanche diodes (SPADs) are suitable for detecting ultrafine particles, they have no ability to distinguish the strong scattered light from large particles, and therefore cannot meet the detection requirements for large-diameter particles. 2. Narrow dynamic range: Single-type detectors often have linear range limitations in signal response. For example, conventional avalanche photodiodes (APDs) can still maintain good linear response under strong light, but are not sensitive to extremely weak signals; while single-photon avalanche diodes (SPADs), although they have single-photon level detection capabilities, are prone to saturation due to the "dead time" effect under strong light signals, and lose counting accuracy. Although existing technologies attempt to extend the dynamic range through software algorithms or adjust the light intensity by adding attenuators in the optical path, these methods often introduce additional errors, reduce system stability, and cannot fundamentally solve the physical limitations of the detector itself.

[0004] Therefore, the industry urgently needs a liquid particle detection solution that combines low detection limit and wide range counting characteristics to break through the linear range limitation and achieve accurate particle size detection with wide range detection. Summary of the Invention

[0005] To address the limitations of existing liquid particle counters using a single type of detector, which suffer from limited particle size range and narrow dynamic range, this invention provides a liquid particle counting device and detection method based on multi-detector fusion. The technical solution is as follows: A liquid particle counting device based on multi-detector fusion, the device comprising: Polarized laser source module: used to generate a highly stable, linearly polarized laser beam, which is focused onto the detection area of ​​the flow cell; Sample flow and control module: including the flow cell, providing a pathway for the liquid to be tested, the measurement beam, and the scattered light from the particles, ensuring that the particles to be tested pass through the detection area and generate collectable scattered light; Dual-channel optical collection module: The dual-channel optical collection module includes a SPAD channel collection lens and an APD channel collection lens. A polarization optical path is provided perpendicular to the laser incident direction and the flow direction of the liquid to be measured. The SPAD channel collection lens and the APD channel collection lens are respectively located on both sides of the polarization optical path to converge the scattered light generated by the particles. Dual-mode detector detection module: The dual-mode detector detection module includes a SPAD array detection channel and an APD array detection channel. The SPAD array detection channel and the APD array detection channel are respectively set to the SPAD channel collection lens and the APD channel collection lens, and are used to collect and process the converged scattered light signals respectively. Data processing and fusion module: used to process and fuse the output signals of the SPAD array detection channel and the APD array detection channel to realize the detection and particle size inversion of particles with a wide particle size range.

[0006] Furthermore, the APD array detection channel includes an APD detector array, a transimpedance amplifier, an analog adder, a multi-stage amplifier, and an AD converter arranged sequentially. The APD detector array includes multiple APD detectors, each corresponding to one of the transimpedance amplifiers. The APD detectors collect scattered light signals and output analog current signals. The analog current signals are converted into voltage signals by their corresponding transimpedance amplifiers. The analog adder includes an RC coupling circuit, which can filter out DC background noise components in the preceding circuit. The analog adder with the RC coupling circuit collects the signals from each of the transimpedance amplifiers, removes DC current in real time, and accumulates them. The signal is then amplified by the multi-stage amplifier and converted from analog to digital by the AD converter to output a voltage signal. The peak value of this voltage signal represents the peak intensity signal of the scattered light. This signal is then input into the data processing and fusion module for processing.

[0007] Furthermore, the SPAD array detection channel includes a SPAD detector array, a transimpedance amplifier, a high-speed comparator, and a pulse shaping circuit. The SPAD detector array includes multiple SPAD detectors, which are configured one-to-one with the transimpedance amplifier, the high-speed comparator, and the pulse shaping circuit. The SPAD detectors collect scattered light signals and output current pulse signals. After preliminary conversion by the transimpedance amplifier, the current pulse signals are threshold-discriminated by the high-speed comparator and then generated into regular TTL pulses by the pulse shaping circuit. The TTL pulses are input to the data processing and fusion module for processing.

[0008] Furthermore, the SPAD detector array is also equipped with a quenching recovery circuit to ensure that the SPAD detector quickly terminates quenching after triggering an avalanche, with the total dead time of quenching-recovery within 15–20 ns.

[0009] Furthermore, the pixel size of the SPAD detector is smaller than that of the APD detector, and the spot size on its image plane is larger than that on the image plane of the APD detector.

[0010] Furthermore, the data processing fusion module includes a high-speed data processing unit and a digital signal processor. The high-speed data processing unit converts the TTL pulse signal into a photon counting signal and accumulates it to obtain the total number of photons scattered by the particle. The digital signal processor determines the particle size based on the photon counting signal of the SPAD detection channel and the peak pulse voltage of the APD detection channel.

[0011] A detection method for a liquid particle counting device based on multi-detector fusion, the detection method comprising: Step 1: The polarized laser source module provides a scattering source for the sample flow and control module. After the particle to be tested flows through the flow cell, the scattered light from the particle is converged by the two optical collection modules and then the scattered light signal of the same particle is synchronously collected by the SPAD array detection channel and the APD array detection channel, respectively. Step 2: The data is then fused by the data processing and fusion module to perform photon event detection and counting on the SPAD array detection channel signal to obtain the total photon count signal. The pulse height of the SPAD array detection channel signal is then analyzed to obtain the peak pulse voltage. Step 3: The data processing and fusion module fuses the total photon count signal and the pulse voltage peak value to determine whether they are small, medium, or large particles. Step 4: Based on the determination results, particle determination and particle size inversion are performed. For small particles and large particles, photon statistical comparison and pulse voltage comparison are used for particle size inversion, respectively. For particles in the middle region, dual-channel data fusion calibration comparison is performed for particle size inversion.

[0012] Furthermore, in step 3, the fusion process includes collaboratively determining the particle type based on whether the peak pulse voltage exceeds a first threshold and whether photon event clusters appear in the total photon count signal. If the peak value of the pulse voltage is lower than the first threshold and a cluster of photon events appears in the total photon count signal, it is determined to be a very small particle; If the peak value of the pulse voltage exceeds the second threshold, it is determined to be a large particle; If the peak value of the pulse voltage is between the first threshold and the second threshold, it is determined to be a particle in the intermediate region.

[0013] Furthermore, in step 4, the photon statistical comparison is performed by inverting the particle size information through analyzing the statistical model of the total number of photons or the photon arrival time interval within the cluster; the pulse voltage comparison is performed by inverting the particle size using the scattered light peak intensity signal and a pre-calibrated light pulse peak voltage-particle size lookup table.

[0014] Furthermore, in step 4, the dual-channel data fusion calibration comparison utilizes the extremely high sensitivity of the SPAD array probe channel signal data to assist in correcting the low-end nonlinearity error of the APD array probe channel signal, or utilizes the linear data of the APD array probe channel to verify the accuracy of the statistical model of the SPAD array probe channel signal data, thereby obtaining a more accurate particle size value.

[0015] The beneficial effects of this invention are: By setting up a dual-channel optical collection module, and corresponding SPAD and APD array detection channels with SPAD and APD channel collection lenses, the output signals of the SPAD and APD array detection channels are processed and fused by the data processing and fusion module. This enables the detection and particle size inversion of particles with a wide particle size range, breaking through the linear range limitation of a single type of detector and achieving accurate particle size detection with a wide range of detection capabilities. Attached Figure Description

[0016] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0017] Figure 1 This is a schematic diagram of the overall structure of Embodiment 1 of the present invention; Figure 2 This is Embodiment 1 of the present invention. Figure 1 A schematic diagram showing the specific structure of the medium polarization laser source module, the sample flow and control module, the dual-channel optical collection module, and the dual-mode detector fusion module; Figure 3 This is a schematic diagram of the particle detection area according to Embodiment 1 of the present invention; Figure 4 This is a schematic diagram of the coordinates of the Mie spherical scattering model according to Embodiment 1 of the present invention; Figure 5 This is a spatial distribution diagram of the scattered light from nanoparticles under linearly polarized light incident according to Embodiment 1 of the present invention; Figure 6 This is a graph showing the intensity of scattered light signals from particles of different sizes in Embodiment 1 of the present invention. Figure 7 This is a schematic diagram of the SPAD array detection channel and high-speed data processing unit according to Embodiment 1 of the present invention; Figure 8 This is a schematic diagram of the APD array detection channel and digital signal processor according to Embodiment 1 of the present invention; Figure 9 This is a schematic diagram of the particle determination and particle size inversion method of Embodiment 2 of the present invention; In the picture: 01. Incident laser; 02. Scattered light collection; 10. Polarized laser source module; 101. Semiconductor linearly polarized laser; 102. Laser focusing lens; 11. Sample flow and control module; 111. Sample liquid flow channel; 112. Laser channel; 113. Scattered light window; 114. Absorbing lens; 12. Dual-channel optical collection module; 121. SPAD channel collection lens; 122. APD channel collection lens; 13. Dual-mode detector detection module; 131. SPAD array detector; 132. APD array detector. Detailed Implementation

[0018] To make the objectives, technical solutions, and advantages of the present invention clearer, the embodiments of the present invention will be described in further detail below with reference to the accompanying drawings.

[0019] Example 1 This embodiment provides a liquid particle counting device based on multi-detector fusion, see [link to documentation]. Figure 1 , Figure 2 The device includes a polarized laser source module 10, a sample flow and control module 11, a dual-channel optical collection module 12, a dual-mode detector detection module 13, and a data processing and fusion module.

[0020] The polarized laser source module 10 is used to generate a highly stable, linearly polarized laser beam, which is focused onto the detection area of ​​the sample flow cell. The core of this module is to generate stable, pure, and well-focused linearly polarized excitation light, ensuring that the incident light is focused at the center of the flow cell to form the particle detection area. The polarized laser source module 10 includes a semiconductor linearly polarized laser 101 and a laser focusing lens 102. In the specific optical path layout, the semiconductor linearly polarized laser 101 is preferably a semiconductor linearly polarized laser with a center wavelength of 600 nm to 900 nm. Further, a collimated laser with a parallel beam output is preferred, or a shaping lens can be used to shape the output beam into a parallel beam. The output parallel beam is focused by the laser focusing lens 102 onto the center of the sample flow cell. By adjusting the position of the laser focusing lens 102, the position of the focusing point in the flow cell can be changed, thereby reducing the influence of manufacturing tolerances on the focusing position. Preferably, to extract the central intensity region of the laser and suppress stray light, an aperture (not shown in the figure) is added in front of the laser focusing lens 102 to block stray light.

[0021] The sample flow and control module 11 includes a precision flow cell designed to provide pathways for the test liquid flow, the measurement beam, and particle-scattered light, ensuring that particles pass through the detection area and generate scattered light. The flow cell is centered on the detection area, with the sample liquid flow channel 111, laser channel 112, and scattered light window 113 orthogonal to the detection center. The polarized laser source module 10 focuses the detection laser onto the detection area. The detection volume corresponding to the detection area is determined by the laser focusing diameter D, the magnification of the collection module, and the photosensitive size of the detector. Figure 3 As shown, after the particles pass through the detection area and generate scattered light, the scattered light is collected and focused by the collection module onto the photosensitive area of ​​the detector (focusing only on a small part of it). Assuming the length and width of the detector are L×H (L corresponds to the laser incident direction, and H corresponds to the flow channel direction of the flow cell), the size of the virtual image it forms in the detection area is calculated as follows (1): The sample flow and control module 11 includes a flow cell, which is designed to provide pathways for the liquid to be tested, the measurement beam, and the scattered light from the particles, ensuring that the particles pass through the detection area and generate scattered light. The flow cell is centered on the detection area, and the sample liquid channel 111, the laser channel 112, and the scattered light window 113 are orthogonal to the detection center.

[0022] Particles flowing through the sample liquid channel 111 generate scattered light after passing through the detection area. The scattered light can be collected by the dual-channel optical collection module 12 and then converged to the dual-mode detector detection module 13. To prevent the risk caused by laser emission and to ensure that the flow cell is not interfered with by external light, a light-absorbing lens 114 is installed at the rear of the laser channel 112.

[0023] Specifically, the dual-channel optical collection module 12 includes a SPAD channel collection lens and an APD channel collection lens. A polarization optical path is provided perpendicular to both the laser incident direction and the flow direction of the liquid to be measured. The SPAD and APD channel collection lenses are respectively positioned on opposite sides of the polarization optical path to converge the scattered light generated by the particles. Both the SPAD and APD channel collection lenses can employ high numerical aperture (greater than 1) collection objectives, which can be composed of multiple lenses. The polarized laser source module 10 focuses the detection laser onto the detection area. The detection volume corresponding to the detection area is jointly determined by the laser focusing diameter D, the magnification of the dual-channel optical collection module, and the photosensitive size of the detector (SPAD detector or APD detector). Figure 3 As shown, after particles pass through the detection area and generate scattered light, the scattered light is collected and focused by the optical collection module onto the photosensitive area of ​​the detector (focusing only on a small portion of it). Assuming the detector's dimensions are L×H (L corresponds to the laser incident direction, and H corresponds to the flow cell channel direction), the size of its virtual image in the detection area is... × Calculate as follows (1): (1) Where M represents the magnification of the dual-channel optical collection module (i.e., the collection objective of the corresponding channel). Specifically, It should be smaller than the laser focusing diameter during the design phase (e.g., This ensures that when the scattered light from the particle is focused onto the photosensitive surfaces of the APD and SPAD arrays by the collection module, the entire particle has entered the region of uniform laser power distribution. When the particle flows outside the virtual image range, its scattered light cannot be focused onto the photosensitive area of ​​the detector. The above only defines two scales of the detection area. When the virtual image of the photosensitive area of ​​the detector is moved back and forth along the optical axis to the edge of the laser focusing area, a complete detection area is formed. Figure 3 (See shaded area). It is worth noting that this design employs a sampling method for particles flowing through the flow cell; that is, the sampling area A on the flow cell cross-section is smaller than the flow cell cross-sectional area S. The final particle count output will take the sampling coefficient into account. Sample flow rate V.

[0024] According to the Mie spherical scattering model, for Figure 4 The coordinate graph shown shows that when the particle size is... d spherical particles with a strength of After being irradiated by linearly polarized incident light, its scattering angle is... azimuth ,distance r The intensity of the scattered light at a given location can be calculated using the following formula (2): (2) in It is a complex series generated during the derivation of the formula when solving Maxwell's differential equations, which is only related to the wavelength of the incident light. Particle size d and scattering angle This formula is relevant. It is used to calculate the particle's position relative to the incident light intensity. ,wavelength Below, any azimuth angle scattering angle The intensity of scattered light on Three-dimensional distribution (see) Figure 5 As shown), when linearly polarized light is incident (polarization direction along the x-direction), the particle is at an azimuth angle =90°, scattered light The maximum scattered light intensity occurs at a 90° angle, while the scattered light intensity at other receiving directions is lower. Therefore, it should be... =90° The SPAD channel and APD channel collecting lenses are designed at a 90° angle to receive scattered light from particles. This angle can obtain the maximum intensity of scattered light and at the same time better suppress background stray light.

[0025] In the direction of collection ( , The scattered light power of the particle can be expressed by the following integral equation (3): (3) Equation (3) above can be derived into equation (4): (4) Among them, variables θ The upper and lower limits of integration are controlled by the numerical aperture (NA) of the light-collecting lens. Calculate the scattered light signal intensity for each particle size by substituting relevant lens design parameters, see [reference needed]. Figure 6 As shown.

[0026] Specifically, the dual-mode detector detection module 13 includes a SPAD array detection channel and an APD array detection channel. The SPAD array detection channel and the APD array detection channel are respectively set to the SPAD channel collection lens and the APD channel collection lens, and are used to collect and process the converged scattered light signals respectively.

[0027] See Figure 7As shown, the APD array detection channel includes an APD detector array, a transimpedance amplifier, an analog adder with RC coupling circuitry, a multi-stage amplifier, and an AD converter arranged sequentially. The APD detector array includes multiple APD detectors (potentially dozens), each corresponding to a transimpedance amplifier. The APD detectors collect scattered light signals and output analog current signals proportional to the light intensity. The analog current signals are first converted into voltage signals by the corresponding transimpedance amplifiers. The signals are then amplified to a suitable level (e.g., mV) by the feedback resistor (not shown in the figure) of the transimpedance amplifiers. The signals from each channel are then filtered to remove DC background noise by the analog adder with RC coupling circuitry and accumulated in real time. They are then amplified in multiple stages by the multi-stage amplifiers to ensure the signal fully corresponds to the ADC range. Finally, the digital voltage signal output by the AD converter is input to the data processing and fusion module for processing. Preferably, since the scattered light intensity varies greatly between particles of different sizes, multiple amplifiers with different amplification factors can be used to ensure sufficient signal resolution. Figure 7 Only one link is drawn in the diagram.

[0028] The SPAD array detection channel includes a SPAD detector array, a transimpedance amplifier, a high-speed comparator, and a pulse shaping circuit. The SPAD detector array includes multiple SPAD detectors (which can be hundreds), and each detector is configured to correspond one-to-one with the transimpedance amplifier, the high-speed comparator, and the pulse shaping circuit. The SPAD detectors collect scattered light signals and output current pulse signals. After initial conversion by the transimpedance amplifier, the current pulse signals are threshold-discriminated by the high-speed comparator and then generated into regular TTL pulses by the pulse shaping circuit. The TTL pulses are then input to the data processing and fusion module for processing.

[0029] For the SPAD array detection channel, the core lies in designing a quenching and recovery circuit with an extremely short dead time to ensure its single-photon detection capability. Preferably, the quenching and recovery circuit is set in the SPAD detector array, and it adopts an active quenching scheme. Specifically, when an avalanche occurs in the SPAD detector due to photon incidence, a high-speed comparator quickly detects the sudden drop in cathode voltage and then drives a MOSFET switch to connect a low-impedance path in parallel across the SPAD detector, forcibly reducing the reverse bias voltage to quench the avalanche. After quenching, a constant current source composed of a high-speed operational amplifier restores the bias voltage to the operating point at a controllable rate, thereby controlling the total dead time of quenching-recovery to within 15–20 ns, ensuring detection efficiency under high event rates.

[0030] The scattered light from particles within the detection area converges onto the image plane of the SPAD or APD detector array, forming a spot with a diameter on the order of micrometers. For a SPAD detector array, each SPAD detector can only measure a single photon in a single quenching cycle. If a SPAD detector receives an excessive number of photons in a single quenching cycle, it is easy to miss photon counts. Therefore, preferably, the scattered light spot converged onto the SPAD array detector should be as large as possible without affecting the overall detection volume, to ensure that the scattered light from particles falling onto each SPAD detector is as small as possible. Furthermore, the SPAD detector array is preferably a product with small pixel size, large array size, and short quenching dead time. This ensures that when receiving scattered light from extremely small particles, the light signal received by each SPAD pixel unit does not cause the SPAD to remain in the quenching dead time for an extended period. The pixel size of the SPAD detector is smaller than that of the APD detector, and the spot size on its image plane is larger than that on the APD detector's image plane, ensuring that the particle size corresponding to SPAD detector saturation is relatively large, while simultaneously reducing the scattered light power received by the SPAD detector, thereby increasing the detection upper limit of the SPAD detector. Examples will be provided below to illustrate this further.

[0031] For example, for 75nm particle scattered light, according to Figure 6 It is known that the intensity of the scattered light from the collected particles is approximately 0.2 nW (theoretical calculation, neglecting propagation loss). Assuming the APD detector array is a 16-element linear array with a photosensitive area of ​​10 mm × 1 mm, and the APD channel collecting lens focuses the scattered light into a 5 µm diameter spot, since 5 µm is much smaller than the photosensitive size of a single APD detector, it will only hit one APD detector, ensuring sufficient response from that element. Subsequent filtering of the DC current and accumulation can effectively characterize the scattered light signal from the particles. For SPAD detector arrays, at a commonly used wavelength of 632 nm, 0.2 nW of scattered light corresponds to approximately [missing information - likely a number of photons per second]. Within a quenching time period of 20 ns, approximately 13 photons are generated. If the collected scattered light is focused onto a SPAD detector during this time, the measurement result will be significantly inaccurate due to the fact that most of the time is within the quenching dead time. Based on the above calculations and comparisons, a SPAD detector array with a pixel pitch of 10 µm can be selected. Using a SPAD channel collection lens, the collected scattered light is focused into a spot with a diameter of 100 µm. The number of pixels within this spot is approximately 69, which is sufficient to ensure that the SPAD detector accurately measures the number of photons generated by the scattered light, while also allowing for a margin of safety, which is ample considering propagation attenuation. Under this design, the accuracy of the measurement can be guaranteed by simultaneously considering the overlap of the virtual images formed by the photosensitive areas of the SPAD detector array and the APD detector array in the flow cell detection area.

[0032] The spot size is adjusted by adjusting the relative position of the channel collecting lens, or by adjusting the position of the detector array and the channel collecting lens, to ensure that the particle-scattered light in the detection area covers the sensitive area of ​​the detector array. All optical elements in both collecting optical paths are made of materials with low self-fluorescence and high transmittance. Preferably, to eliminate stray light, the entire dual-channel optical collecting module is enclosed in a light shield with its inner wall coated with high-absorption black paint.

[0033] Specifically, the data processing and fusion module includes a high-speed data processing unit and a digital signal processor. The high-speed data processing unit converts the TTL pulse signal output from the SPAD array detection channel into a photon counting signal and accumulates it to obtain the total number of photons scattered by the particle. The digital signal processor first obtains the peak value of the pulse voltage output from the APD array detection channel, and then determines the particle size based on the photon counting signal of the SPAD detection channel and the peak value of the pulse voltage of the APD detection channel.

[0034] Example 2 This embodiment provides a detection method for a liquid particle counting device based on multi-detector fusion. See [link to relevant documentation]. Figure 9 The detection method includes: Step 1: The polarized laser source module provides a scattering source for the sample flow and control module. After the particle to be tested flows through the flow cell, the scattered light from the particle is converged by the two optical collection modules and then the scattered light signal of the same particle is synchronously collected by the SPAD array detection channel and the APD array detection channel, respectively. Step 2: The data is then fused by the data processing and fusion module to perform photon event detection and counting on the SPAD array detection channel signal to obtain the total photon count signal. The pulse height of the SPAD array detection channel signal is then analyzed to obtain the peak pulse voltage. Step 3: The data processing and fusion module fuses the total photon count signal and the pulse voltage peak value to determine whether they are small, medium, or large particles. Step 4: Based on the determination results, particle determination and particle size inversion are performed. For small particles and large particles, particle size inversion is performed by photon statistical comparison and pulse voltage peak value, respectively. For particles in the middle region, particle size inversion is performed by dual-channel data fusion calibration comparison.

[0035] Specifically, in steps 3 and 4, the fusion processing and particle determination and particle size inversion methods are described in [reference needed]. Figure 9 As shown.

[0036] In the specific implementation of fusion processing, particle determination, and particle size inversion, the system, based on the synergistic characteristics of the dual-channel signals, achieves accurate identification and measurement of extremely small particles, large particles, and particles in the intermediate region through discrete hardware discrimination circuits and fusion processing algorithms. For the determination of extremely small particles (e.g., 20nm to 50nm), in terms of hardware, in the digital signal processor, the APD array detection channel signal is converted from analog to digital and then fed into a window comparator. When the signal continuously falls below a preset first low threshold (e.g., the digital signal processor corresponding to the APD array detection channel sets a 50nm detection threshold), it outputs an "APD not detected" flag. Specifically, it can also... Figure 8 A threshold comparison is added between the multi-stage amplifier and the AD converter in the illustrated link. When the signal is below this threshold, it indicates that the APD has not detected a particle pulse signal. Simultaneously, in the high-speed data processing unit, the photon event flow of the SPAD array detection channel is monitored in real time by the high-speed data processing unit's detection logic unit. This channel counts the number of photons within a sliding time window and compares it with the dynamically calculated water molecule scattering background noise threshold. If the number of photons continuously and significantly exceeds the threshold, it is determined that a "photon event cluster" has occurred and a trigger pulse is output. At this time, when the two conditions of "APD not detected" and "SPAD cluster triggering" are simultaneously satisfied in the co-triggered circuit, a "minimal particle event" flag is generated. Algorithmically, the digital signal processor then extracts the number of all photons arriving within the event time window and continues to monitor the photon data of the next time window until the photon number returns to the background level. The digital signal processor extracts the number of photons of the entire SPAD detector array during this period, subtracts the background noise data to obtain the total number of photons, and obtains the preliminary particle size by querying a "total photon number - particle size" range lookup table pre-established by standard nanoparticle calibration. The following section will provide a theoretical analysis of the relationship between "total photon count - particle size".

[0037] For a single photon, its energy The calculation is as shown in equation (5): (5) in, is Planck's constant. At the speed of light, Wavelength. In optical power P Below, photons per second: (6) The time it takes for a particle to pass through the detection area can be determined by the size of the detection area and the flow rate of the sample liquid. By substituting the collected scattered light intensity data corresponding to the particle size in the figure, the theoretical range of "total number of photons - particle size" can be obtained.

[0038] It is worth noting that the theoretical correspondence between "total photon count and particle size" mentioned above is a relatively ideal reference value. Actual data is affected by factors such as laser characteristics, lens parameters, attenuation, and SPAD response capabilities. Therefore, the background noise data and the "total photon count - particle size" range lookup table are not universally applicable standard data and tables. Instead, they are specific data and tables obtained by calibration and standardization using corresponding standard particles (e.g., 20nm, 30nm, 40nm, etc.) for each instrument. Furthermore, for instruments that have already been calibrated and standardized, timely calibration is necessary to ensure their counting accuracy. Table 1 provides a reference format for the "total photon count - particle size" range lookup table:

[0039] Table 1 "Total Photon Count - Particle Size" Range Lookup Table (Reference) For determining large particles (e.g., larger than 75nm), the hardware approach is the same as above: the APD array detection channel signal is converted from analog to digital and then fed into a window comparator. When the signal exceeds a preset second threshold (e.g., a 75nm detection threshold is set by the digital signal processor corresponding to the APD detection channel), it indicates "SPAD saturation." Specifically, it can also be done in... Figure 8 In the illustrated link, a second threshold comparison is added between the multi-stage amplifier and the AD converter. When the signal reaches this threshold, it indicates "SPAD saturation." The analog voltage signal of the APD array detection channel is captured and held by the peak voltage Vpeak of the scattered light pulse by the peak detection and hold circuit. After AD conversion, this voltage signal is compared with the set second threshold. When Vpeak exceeds the threshold, a "large particle event" is triggered. Algorithmically, the digital signal processor reads Vpeak and calls a lookup table of "light pulse peak voltage - particle size" calibrated with standard micron-sized particles. Through the lookup table, Vpeak is directly mapped to the particle size. The theoretical analysis of the correspondence between "light pulse peak voltage - particle size" will be carried out below.

[0040] In this invention, when the collected scattered light is focused onto the detection channel of the APD array, it will be converted into a current signal. The conversion capability is characterized by the responsivity R (A / W) of the APD detector. After the current signal of the APD detector is converted into a voltage signal through transimpedance amplification, it is then amplified and converted by multiple stages in the subsequent links to obtain the final pulse signal, which is the final optical pulse peak voltage Vpeak, which can be calculated by equation (7): (7) in, The collected scattered light power, For transimpedance amplification gain, This represents the link gain.

[0041] Similarly, the "peak voltage of optical pulse - particle size" range lookup table should be specific to each instrument, using data and tables obtained through calibration and extensibility testing with corresponding standard particles. Furthermore, for instruments that have already been calibrated and verified, timely calibration is necessary to ensure their counting accuracy. Table 2 provides a reference format for the "peak voltage of optical pulse - particle size" range lookup table:

[0042] Table 2. Reference Table for "Optical Pulse Peak Voltage - Particle Size" Range For particles in the intermediate region, in terms of hardware, when a particle event triggers the first threshold of the APD but does not trigger the second threshold, the system determines it as an "intermediate region particle event." A high-precision synchronous clock can be used to simultaneously latch the peak voltage Vpeak of the APD array detection channel and the photon data of the SPAD array detection channel. For particles with intermediate diameters (e.g., 50-75nm), their scattered light can generate a measurable but error-prone voltage signal peak (V) on the APD array detection channel and a countable but potentially saturated photon count (N) on the SPAD array detection channel. The relationship between a single signal and particle size may be ambiguous, but considering the two signals as a coordinate pair (V, N), it uniquely corresponds to a more definite particle size range in two-dimensional space. Through experiments with a large number of standard particles, we can draw a two-dimensional lookup table for this region. During subsequent measurements, simply locating the measured (V, N) on the table yields the particle size range. This enables smooth, accurate, and reliable particle size measurement across the entire range. Table 3 provides a reference form of the (V, N) range lookup table:

[0043] Table 3 (V, N) Range Lookup Table Reference Specifically, the setting of the first threshold depends on the detection capability of the APD array detection channel and the link noise level. Signals above this threshold can be completely detected by the APD array detection channel with sufficient margin. The setting of the second threshold depends on the saturation light intensity of the SPAD array detection channel. Signals below this threshold can be completely detected by the SPAD array detection channel with sufficient margin.

[0044] Beneficial effects: 1. Extremely low detection limit: The SPAD array detection channel has single-photon level detection sensitivity, which can effectively capture extremely weak scattered light generated by nanoparticles. Based on optimized optical collection and background suppression, the system can achieve stable detection and counting of particles below 50nm, filling the gap of traditional light scattering instruments in the detection of ultrafine particles. 2. Covering an ultra-wide dynamic range and achieving integrated measurement across the entire particle size range: Through the collaborative work of the SPAD array detection channel and the APD array detection channel, the system can simultaneously process single-photon events to intense light signals, with a dynamic range spanning 4–5 orders of magnitude. It can achieve full particle size coverage from nanometers to several micrometers without switching detectors or optical configurations, meeting the real-time monitoring needs of wide-distribution particle systems in fields such as semiconductors and pharmaceuticals. 3. Improve the measurement accuracy and reliability of the intermediate particle size region: For the region where traditional methods are prone to inaccuracies, the system uses dual-channel data fusion for cross-validation and correction. It utilizes the high sensitivity of the SPAD array detection channel to correct the error of the APD array detection channel in the low signal region, and uses the good linearity of the APD array detection channel to verify the reliability of the SPAD array detection channel photon statistical model at high count rates. Finally, it outputs more accurate particle size values ​​through an adaptive weighted fusion algorithm, reducing misjudgment and omission in the boundary region. 4. Enhance the robustness and anti-interference capability of the system: The independent acquisition and processing mechanism of the dual-channel optical collection module can resist the accidental noise or device drift of a single channel to a certain extent, deduct the background noise such as water molecule scattering in real time, and combine hardware threshold and algorithm discrimination to significantly improve the signal-to-noise ratio and detection stability. 5. Supports high-throughput, real-time online monitoring, facilitating system integration and automation: Using arrayed detectors and parallel signal processing architecture, it can achieve high-speed, continuous flow sample real-time processing and statistics. The modular design facilitates integration with existing fluid control systems, clean monitoring platforms or process lines, and supports long-term unattended operation. 5. Excellent scalability and adaptability: The pixel size, optical collection angle, signal processing parameters, etc. of the SPAD and APD detector arrays can be flexibly adjusted according to specific applications. This method can also be extended to multi-wavelength, multi-angle scattering measurement systems to further improve particle resolution in complex systems.

[0045] Some steps in the embodiments of the present invention can be implemented using software, and the corresponding software program can be stored in a readable storage medium, such as an optical disc or a hard disk.

[0046] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A liquid particle counting device based on multi-detector fusion, characterized in that, The device includes: Polarized laser source module: used to generate a highly stable, linearly polarized laser beam, which is focused onto the detection area of ​​the flow cell; Sample flow and control module: including the flow cell, providing a pathway for the liquid to be tested, the measurement beam, and the scattered light from the particles, ensuring that the particles to be tested pass through the detection area and generate collectable scattered light; Dual-channel optical collection module: The dual-channel optical collection module includes a SPAD channel collection lens and an APD channel collection lens. A polarization optical path is provided perpendicular to the laser incident direction and the flow direction of the liquid to be measured. The SPAD channel collection lens and the APD channel collection lens are respectively located on both sides of the polarization optical path to converge the scattered light generated by the particles. Dual-mode detector detection module: The dual-mode detector detection module includes a SPAD array detection channel and an APD array detection channel. The SPAD array detection channel and the APD array detection channel are respectively set to the SPAD channel collection lens and the APD channel collection lens, and are used to collect and process the converged scattered light signals respectively. Data processing and fusion module: used to process and fuse the output signals of the SPAD array detection channel and the APD array detection channel to realize the detection and particle size inversion of particles with a wide particle size range.

2. The liquid particle counting device based on multi-detector fusion according to claim 1, characterized in that, The APD array detection channel includes an APD detector array, a transimpedance amplifier, an analog adder, a multi-stage amplifier, and an AD converter arranged sequentially. The APD detector array includes multiple APD detectors, each corresponding to one of the transimpedance amplifiers. The APD detectors collect scattered light signals and output analog current signals. The analog current signals are converted into voltage signals by their corresponding transimpedance amplifiers. The analog adder includes an RC coupling circuit, which can filter out DC background noise components in the preceding circuit. The analog adder with the RC coupling circuit collects the signals from each of the transimpedance amplifiers, removes DC current in real time, and accumulates them. The signal is then amplified by the multi-stage amplifier and converted from analog to digital by the AD converter to output a voltage signal. The peak value of this voltage signal represents the peak intensity signal of the scattered light. This signal is then input to the data processing and fusion module for processing.

3. The liquid particle counting device based on multi-detector fusion according to claim 2, characterized in that, The SPAD array detection channel includes a SPAD detector array, a transimpedance amplifier, a high-speed comparator, and a pulse shaping circuit. The SPAD detector array includes multiple SPAD detectors, which are configured one-to-one with the transimpedance amplifier, the high-speed comparator, and the pulse shaping circuit. The SPAD detectors collect scattered light signals and output current pulse signals. After preliminary conversion by the transimpedance amplifier, the current pulse signals are threshold-discriminated by the high-speed comparator and then generated into regular TTL pulses by the pulse shaping circuit. The TTL pulses are input to the data processing and fusion module for processing.

4. A liquid particle counting device based on multi-detector fusion according to claim 3, characterized in that, The SPAD detector array is also equipped with a quenching and recovery circuit to ensure that the SPAD detector quickly terminates quenching after triggering an avalanche, with the total dead time of quenching-recovery within 15–20 ns.

5. A liquid particle counting device based on multi-detector fusion according to claim 3, characterized in that, The pixel size of the SPAD detector is smaller than that of the APD detector, and the spot size on its image plane is larger than that on the image plane of the APD detector.

6. A liquid particle counting device based on multi-detector fusion according to claim 3, characterized in that, The data processing and fusion module includes a high-speed data processing unit and a digital signal processor. The high-speed data processing unit converts the TTL pulse into a photon counting signal and accumulates it to obtain the total number of photons scattered by the particle. The digital signal processor determines the particle size based on the photon counting signal of the SPAD array detection channel and the peak pulse voltage of the APD array detection channel.

7. A detection method for a liquid particle counting device based on multi-detector fusion, wherein the liquid particle counting device based on multi-detector fusion according to any one of claims 1 to 6 is characterized in that, The detection method includes: Step 1: The polarized laser source module provides a scattering source for the sample flow and control module. After the particle to be tested flows through the flow cell, the scattered light from the particle is converged by the two optical collection modules and then the scattered light signal of the same particle is synchronously collected by the SPAD array detection channel and the APD array detection channel, respectively. Step 2: The data is then fused by the data processing and fusion module to perform photon event detection and counting on the signal of the SPAD array detection channel to obtain the total photon count signal. The pulse height of the signal of the SPAD array detection channel is then analyzed to obtain the peak pulse voltage. Step 3: The data processing and fusion module fuses the total photon count signal and the pulse voltage peak value to determine whether they are small, medium, or large particles. Step 4: Based on the determination results, particle determination and particle size inversion are performed. For small particles and large particles, photon statistical comparison and pulse voltage comparison are used for particle size inversion, respectively. For medium particles, dual-channel data fusion calibration comparison is performed for particle size inversion.

8. The detection method of a liquid particle counting device based on multi-detector fusion according to claim 7, characterized in that, In step 3, the fusion process includes collaboratively determining the particle type based on whether the peak pulse voltage exceeds a first threshold and whether photon event clusters appear in the total photon count signal. If the peak value of the pulse voltage is lower than the first threshold and a cluster of photon events appears in the total photon count signal, it is determined to be a very small particle; If the peak value of the pulse voltage exceeds the second threshold, it is determined to be a large particle; If the peak value of the pulse voltage is between the first threshold and the second threshold, it is determined to be a particle in the intermediate region.

9. The detection method of a liquid particle counting device based on multi-detector fusion according to claim 7, characterized in that, In step 4, the photon statistical comparison is used to invert the particle size information by analyzing the statistical model of the total number of photons or the photon arrival time interval within the cluster; the pulse voltage comparison is used to invert the particle size by using the scattered light peak intensity signal and according to a pre-calibrated light pulse peak voltage-particle size lookup table.

10. The detection method of a liquid particle counting device based on multi-detector fusion according to claim 7, characterized in that, In step 4, the dual-channel data fusion calibration comparison utilizes the extremely high sensitivity of the SPAD array probe channel signal data to assist in correcting the low-end nonlinearity error of the APD array probe channel signal, or utilizes the linear data of the APD array probe channel to verify the accuracy of the statistical model of the SPAD array probe channel signal data, thereby obtaining a more accurate particle size value.

Images