A Non-destructive Method and System for the Detection of Trace Anthocyanins in Fermented Fruit Wine Based on Dual-Wavelength Photothermal Deflection

Abstract

This invention belongs to the field of photoelectric detection and food engineering technology, and relates to a non-destructive detection method and system for trace anthocyanins in fermented fruit wine based on dual-wavelength photothermal deflection. This invention achieves non-destructive, in-situ, and online monitoring of core trace components in dark, complex wines: It employs purely physical photothermal deflection spectroscopy, eliminating the need for destructive pretreatment such as large-scale dilution, high-speed centrifugation, or organic reagent extraction of the raw wine. The system can be directly connected to the side line of the fermentation tank or the bottling pipeline for continuous sample injection, significantly overcoming the shortcomings of traditional high-performance liquid chromatography (HPLC) such as detection lag, cumbersome operation, and the tendency of traditional absorbance methods to cause photodetector saturation blind zones.

Description

Technical Field

[0001] This invention belongs to the field of optoelectronic detection and food engineering technology, and relates to a non-destructive detection method and system for trace anthocyanins in fermented fruit wine based on dual-wavelength photothermal deflection. Specifically, it relates to a non-destructive measurement method and system that utilizes dual-wavelength modulation to excite the photothermal lens effect to monitor the specific photothermal conversion of target molecules, thereby accurately assessing the concentration of effective components in fermented fruit wine under strong background interference. Background Technology

[0002] Mulberry wine is a dark-colored berry fermented wine, rich in natural antioxidants, particularly anthocyanins such as cyanidin-3-glucoside. These are the core trace components that give it its high health value and characteristic color. Rapid and accurate monitoring of anthocyanin concentration changes throughout fermentation, aging, and shelf life is crucial for assessing product quality, optimizing process parameters, and preventing counterfeiting and ensuring traceability. However, the matrix of this type of wine is extremely complex. Not only is the original liquid very dark in color, ranging from deep purplish-red to blackish-purple, with extremely low light transmittance, but it is also rich in pectin, protein, and plant polysaccharides, making it highly susceptible to forming microscopic colloidal suspensions, resulting in strong background light scattering characteristics.

[0003] Existing detection methods have significant limitations when targeting trace amounts of active ingredients in dark-colored, highly turbid wines:

[0004] (1) Ultraviolet-visible spectrophotometry (such as the classic pH differential method): This method relies on the intensity of transmitted light penetrating the sample to calculate absorbance. For dark-colored raw wines, the probe light is extremely difficult to penetrate, which can easily cause the detector signal to overflow. At the same time, the suspended colloids in the wine will induce strong Rayleigh scattering and Mie scattering, which seriously interfere with the optical measurement baseline. In order to force the measurement, the sample must be subjected to destructive pretreatment such as large-scale dilution, high-speed centrifugation, or even demulsification. This is not only cumbersome and time-consuming, but also directly destroys the original microenvironment of trace substances in the wine, which can easily introduce huge systematic errors.

[0005] (2) High-performance liquid chromatography (HPLC) and mass spectrometry: Although they have extremely high detection accuracy and specificity, these methods are highly dependent on expensive large-scale laboratory instruments and have long single analysis cycles. Furthermore, chromatography requires extremely rigorous sample pretreatment before injection, such as organic reagent extraction and solid-phase extraction column purification, which consumes a large amount of chemical reagents. It is a typical offline, destructive, and high-cost detection method, which cannot meet the needs of rapid, in-situ, or online monitoring in production sites.

[0006] (3) Conventional fluorescence spectroscopy: Although the sensitivity is higher than that of ordinary absorption spectroscopy, there are a large number of competitive absorption substances and other polyphenolic byproducts in complex liquid systems such as mulberry wine. These substances can cause severe internal filtration effects and background fluorescence signal overlap, which can mask or quench the true signal of trace anthocyanins, making it difficult to achieve accurate anti-interference quantification.

[0007] Therefore, there is an urgent need for an innovative detection technology that can ignore the dark background of the wine, resist the interference of light scattering in high turbidity, and eliminate the need for complex pretreatment, so as to achieve high-precision quantification of trace anthocyanins in complex matrices quickly, in situ, and non-destructively.

[0008] Furthermore, although photothermal deflection spectroscopy has been applied in the detection of solid-state optical thin films or high-purity transparent liquids, traditional single-beam photothermal deflection systems cannot be directly applied to dark-colored fermented wines. Due to the strong infrared overtone absorption of polar solvents such as water and ethanol in the wine, and the substrate endothermic drift caused by high concentrations of suspended pectin, the deflection signal excited by a single wavelength is mixed with a large background thermal lensing response. Simultaneously, natural thermal convection and fermentation microbubbles within the traditional static detection chamber instantly disrupt the local refractive index gradient field. Therefore, it is necessary to develop novel optical modulation decoupling mechanisms and hydrodynamic steady-state structures to extract the true response of trace target molecules in complex, highly turbid liquid systems. Summary of the Invention

[0009] The technical problem to be solved by the present invention is to overcome the shortcomings of the existing technology, namely, the fact that conventional ultraviolet-visible spectroscopy is prone to detection blind zone for dark fermented wines and is seriously affected by light scattering interference from suspended colloids, and that chromatographic methods are cumbersome and require destructive pretreatment. The present invention proposes a non-destructive quantitative detection method and system based on the dual-wavelength alternating modulation photothermal deflection effect.

[0010] This invention utilizes a dual-wavelength pump light source (characteristic excitation light and reference light) to alternately irradiate the in-situ wine liquid, exciting target molecules to generate a specific micro-region thermal gradient field. Unlike traditional methods that rely on adding optical reflection cavities to amplify signals, this invention innovatively employs a deep hardware-software collaborative architecture: it utilizes a high-frequency complementary PWM sequence output by an advanced timer in the underlying microcontroller to directly drive a dual-wavelength pump laser for time-division multiplexing; coupled with precise laminar flow control (Re < 50) of an orthogonal optical fluid control cell, it maintains a very weak thermal lens steady state in dynamic liquid flow; finally, through multi-stage active analog filtering and dual-channel synchronous oversampling, it executes a digitally locked amplification (DLIA) algorithm in the digital kernel. This system, through electronic decoupling and frequency-domain phase-sensitive detection, suppresses common-mode thermal background interference from non-pigment suspended matter and the base liquid, establishing an absolute quantitative relationship between pure light spot deflection and trace anthocyanin concentration at extremely low signal-to-noise ratios, achieving in-situ, rapid, and high-precision determination of core effective components in dark, highly turbid, and complex wines.

[0011] The core principle of the trace anthocyanin non-destructive detection system based on dual-wavelength photothermal deflection provided by this invention is as follows: This invention is based on the "dual-wavelength differential photothermal deflection effect (Mirage Effect)," overcoming the physical limitation of conventional single-wavelength photothermal deflection technology, which cannot distinguish between the target signal and background absorption. When a reference light (e.g., 808 nm near-infrared light) at the absorption trough of the target molecule is used to irradiate the in-situ dark wine liquid, the matrix (water, ethanol, and colloidal suspensions) and anthocyanins in the wine liquid only undergo weak absorption in this wavelength band. When a very weak probe light orthogonal to it passes through this region, only a small and constant background spatial deflection is generated. When the excitation light (e.g., 532 nm green light) matching the characteristic absorption peak of anthocyanins is switched to irradiate the same region, the trace anthocyanins strongly absorb specific light energy and undergo non-radiative relaxation, instantly generating a significant local temperature rise (ΔT). Due to the change in the refractive index of the liquid with temperature, a specific refractive index gradient field is formed locally (i.e., the "thermal lensing" effect), causing the probe light passing through this region to undergo significant angular deflection and spot displacement. By alternately modulating the dual-wavelength pump light, and calculating the physical spatial differential displacement of the probe spot in the two states ( This difference can significantly suppress thermal interference from the high turbidity of the wine itself and the matrix background, and its magnitude is highly correlated with the anthocyanin concentration in the wine. The ingenuity of this invention lies in its non-reliance on the physical extension of the spatial optical path, instead utilizing the principle of common-mode cancellation by calculating the minute spatial deflection displacement difference between adjacent states on the time axis to suppress the significant thermal background drift introduced by the high turbidity of the wine itself and matrix changes at the physical source. This difference is strictly monotonic only with the concentration of the excited target molecules, thus achieving specific quantification in highly interfering fluids.

[0012] To solve the above-mentioned technical problems, the technical solution adopted by the present invention is as follows:

[0013] This invention discloses a non-destructive detection system for trace anthocyanins in fermented fruit wine based on dual-wavelength photothermal deflection, including a dual-wavelength pump light source, a dichroic mirror beam combiner, an orthogonal optical fluidic cell, a detection light source, and a position-sensitive detector.

[0014] in,

[0015] The output end of the dual-wavelength pump light source is coupled to the input end of the dichroic mirror combiner in the spatial optical path; a square vertical laminar flow channel is formed inside the orthogonal optical fluid control cell, with the direction of the fermentation fruit wine fluid flow in the square vertical laminar flow channel as the Z-axis, and the output end of the dichroic mirror combiner is orthogonally aligned with the X-axis side of the square vertical laminar flow channel in the spatial optical path; the probe beam output by the probe light source is orthogonally aligned with the Y-axis side of the square vertical laminar flow channel in the orthogonal optical fluid control cell in the spatial optical path, and the probe beam transmitted from the orthogonal optical fluid control cell illuminates the photosensitive target surface of the position-sensitive detector in the far-field optical path.

[0016] In some embodiments, the non-destructive testing system for trace anthocyanins in fermented fruit wine based on dual-wavelength photothermal deflection further includes a main control unit, a modulation drive circuit, a constant current injection pump, a signal conditioning module, and an analog-to-digital converter.

[0017] in,

[0018] The control signal output terminal of the main control unit is electrically connected to the input terminal of the modulation drive circuit; the power output terminal of the modulation drive circuit is electrically connected to the input terminal of the dual-wavelength pump light source; the communication control terminal of the main control unit is electrically connected to the constant current injection pump, the probe light source and the analog-to-digital converter respectively.

[0019] The constant flow injection pump is connected to the bottom liquid inlet of the orthogonal optical fluidic cell via a fluid microtube; the current signal output of the position-sensitive detector is electrically connected to the input of the signal conditioning module; the voltage output of the signal conditioning module is electrically connected to the analog input of the analog-to-digital converter; and the digital signal output of the analog-to-digital converter is closed-loop connected back to the digital input bus of the main control unit.

[0020] In some embodiments, the orthogonal optical fluid control cell is constructed with an insulating and heat-resistant material as its main body. A square vertical laminar flow channel is formed inside the main body, with a liquid inlet at the bottom and a liquid outlet at the top. A thermally conductive metal temperature control jacket is tightly fitted to the outer side of the square vertical laminar flow channel. The orthogonal optical fluid control cell is equipped with a closed-loop temperature control stabilization module, which integrates a TEC semiconductor temperature control chip, a liquid cooling circulation chamber, and a PID precision temperature controller. The TEC semiconductor temperature control chip is tightly fitted to the outer flat wall of the temperature control jacket. The temperature sensor of the PID precision temperature controller is attached to the outer wall of the temperature control jacket adjacent to the square vertical laminar flow channel, specifically downstream of the flow direction of the fermented fruit wine fluid. The liquid cooling circulation chamber is installed on the back of the TEC semiconductor temperature control chip. The side wall of the liquid cooling circulation chamber has a coolant inlet and a coolant outlet.

[0021] In some embodiments, with the direction of fluid flow in the square vertical laminar flow channel as the Z-axis, a Y-axis probe window and a Y-axis receiver probe window are provided in the Y-axis direction of the square vertical laminar flow channel, which are arranged opposite to each other on both sides of the square vertical laminar flow channel and the windows are parallel to each other; with the direction of fluid flow in the square vertical laminar flow channel as the Z-axis, a pump light incident window is provided in the X-axis direction of the square vertical laminar flow channel, so that the pump beam after beam combining can be incident on one side into the square vertical laminar flow channel; the probe window, the receiver probe window and the pump light incident window are all made of quartz.

[0022] In some embodiments, on the transmission paths of the X-axis pump light and the Y-axis probe light corresponding to the square vertical laminar flow channel, the temperature control jacket and the main body are respectively provided with through mechanical light-transmitting holes, and the quartz light windows are respectively embedded and sealed inside the mechanical light-transmitting holes to form the light-transmitting sidewalls of the square vertical laminar flow channel.

[0023] In some embodiments, the pump light output from the dual-wavelength pump light source is spatially collinearly combined by a dichroic mirror combiner to form a pump beam with completely overlapping spatial paths. This combined pump beam enters the square vertical laminar flow channel through the pump light incident window. The probe beam output from the probe light source is orthogonally aligned with the Y-axis probe light window of the orthogonal optical fluid control cell in the spatial optical path, passes through the square vertical laminar flow channel, and is transmitted out from the relatively parallel Y-axis receiving probe light window. The probe beam transmitted from the orthogonal optical fluid control cell illuminates the photosensitive target surface of the position-sensitive detector in the far-field optical path. In the square vertical laminar flow channel, the pump beam output from the dual-wavelength pump light source and the probe beam output from the probe light source intersect and are orthogonal in space, and the intersection point is the interaction center.

[0024] In some embodiments, the main control unit outputs a square wave pulse width modulation signal through a modulation driving circuit to perform time-division multiplexing driving on the dual-wavelength pump light source, so that the dual-wavelength pump light source alternately outputs anthocyanin characteristic excitation light or reference light; the wavelength of the anthocyanin characteristic excitation light is 500 nm to 550 nm; the wavelength of the reference light is 780 nm to 1100 nm.

[0025] In some embodiments, the wavelength of the anthocyanin characteristic excitation light is 532 nm or 520 nm.

[0026] In some embodiments, the reference light is 808 nm or 1064 nm.

[0027] In some embodiments, the wavelength of the probe light output by the probe light source is 630 nm to 1600 nm.

[0028] In some embodiments, the wavelength of the detection light output by the detection light source is 1310 nm, 1550 nm, or 632.8 nm.

[0029] In some embodiments, the dual-wavelength pump source can be replaced with a tunable laser; the position-sensitive detector can be replaced with a four-quadrant photodiode or a linear CCD / CMOS array camera; and the constant-flow injection pump can be replaced with an injection pump or a pneumatically driven microfluidic system.

[0030] In some embodiments, the system further includes a control device for generating and sending control commands; the main control unit is communicatively connected to the control device and is used to control the execution device according to the received control commands; a transimpedance amplifier is built into the signal conditioning module; and a digital lock-in amplifier is built into the digital signal processing module of the main control unit.

[0031] In some embodiments, the current signal output terminal of the position-sensitive detector is connected to the input terminal of the transimpedance amplifier; the output terminal of the transimpedance amplifier is connected to the analog input terminal of the analog-to-digital converter; the digital signal output terminal of the analog-to-digital converter is connected to the input terminal of the digital lock-in amplifier; and the output terminal of the digital lock-in amplifier is communicatively connected to the control device.

[0032] Furthermore, the present invention discloses a method for non-destructive testing of trace anthocyanins in fermented fruit wine based on the above-mentioned non-destructive testing system for trace anthocyanins in fermented fruit wine using dual-wavelength photothermal deflection, comprising the following steps: system initialization and benchmark calibration: establishing the thermodynamic steady state and hydrodynamic steady state inside the square vertical laminar flow channel, and establishing the initial physical coordinate benchmark zero point of the light spot on the photosensitive target surface of the position-sensitive detector;

[0033] After system initialization and reference calibration, the system is started. The dual-wavelength pump light source is based on a preset modulation frequency. Alternating outputs of anthocyanin-specific excitation light or reference light are directed into the interaction center of a square vertical laminar flow channel. Simultaneously, a probe light source is activated, and as the probe beam passes over the interaction center, a position-sensitive detector captures the real-time displacement of the light spot. The displacement of the light spot is then pre-processed to obtain the physical spatial differential displacement. The concentration of anthocyanins in the current fermented fruit wine was calculated using the sensitivity mapping concentration equation.

[0034] In some embodiments, the modulation frequency It is 10 Hz-1 kHz.

[0035] In some embodiments, the modulation frequency The range is 10 Hz to 200 Hz.

[0036] In some embodiments, the modulation frequency The frequency range is 10 Hz to 50 Hz.

[0037] In some embodiments, the modulation frequency It is 20 Hz.

[0038] In some embodiments, the ;in, Total deflection shift resulting from the combined effects of anthocyanin-specific absorption and basal absorption, unit: μm; Background deflection displacement caused solely by the base of the fermented fruit wine, in μm.

[0039] In some embodiments, the sensitivity mapping concentration equation is: ;in, The overall photothermal response sensitivity coefficient of the system, in mg / (L·μm); Zero-point calibration intercept of the system, unit: mg / L.

[0040] In some embodiments, the preprocessing includes: real-time capture of the spot displacement by a position-sensitive detector, converting it into photocurrent; converting the photocurrent into a voltage signal, sampling to obtain a discrete voltage sequence, and normalizing and reconstructing the spot position; and using differential computation to modulate the frequency. Using a reference frame, the digital lock-in amplifier performs digital phase-sensitive detection on the reconstructed spot position values. By calculating the magnitude of the orthogonal vector, the characteristic amplitude at a specific frequency is extracted. Based on the extracted characteristic amplitude, the actual physical spatial differential displacement is calculated using the Fourier reconstruction formula. .

[0041] In some embodiments, the preprocessing includes: a position-sensitive detector capturing the displacement of the light spot in real time and converting it into photocurrent; a transimpedance amplifier converting the photocurrent into a voltage signal; an analog-to-digital converter synchronously sampling the photocurrent to obtain a discrete voltage sequence; and normalizing and reconstructing the light spot position; and using the differential calculation of the digital signal processing module inside the main control unit to modulate the frequency. Using a reference frame, the digital lock-in amplifier performs digital phase-sensitive detection on the reconstructed spot position values. By calculating the magnitude of the orthogonal vector, the characteristic amplitude at a specific frequency is extracted. Based on the extracted characteristic amplitude, the actual physical spatial differential displacement is calculated using the Fourier reconstruction formula. .

[0042] In some embodiments, when the absolute value of the real-time temperature deviation inside the square vertical laminar flow channel is ≤0.1℃ and is maintained for 60 s, the system determines that the thermodynamic steady state has been established; when the Reynolds number of the fermented fruit wine fluid inside the square vertical laminar flow channel is <50, the system determines that the hydrodynamic steady state has been established.

[0043] In some embodiments, a non-destructive detection method for trace anthocyanins in fermented fruit wine based on dual-wavelength photothermal deflection includes the following steps:

[0044] Step S1: System initialization and benchmark calibration: Pump the fermented fruit wine into the orthogonal photofluidic cell, start the temperature control stabilization module, and establish the thermodynamic steady state inside the square vertical laminar flow channel; control the injection flow rate of the fermented fruit wine to establish the hydrodynamic steady state inside the square vertical laminar flow channel; turn on the detection light source separately, collect the initial photocurrent output from both ends of the position sensitive detector, and establish the initial physical coordinate reference zero point of the light spot on the photosensitive target surface of the position sensitive detector;

[0045] Step S2: Start the system. The main control unit outputs a square wave pulse width modulation signal to drive the dual-wavelength pump light source in time division multiplexing. This causes the dual-wavelength pump light source to alternately output anthocyanin characteristic excitation light or reference light, which is then injected into the interaction center of the square vertical laminar flow channel to generate a dynamic refractive index gradient field.

[0046] Step S3: Synchronously turn on the detection light source. The detection beam passes through the dynamic refractive index gradient field in the interaction center and is periodically deflected. The position-sensitive detector captures the displacement of the light spot in real time and converts it into photocurrent. The transimpedance amplifier converts the photocurrent into a voltage signal. The analog-to-digital converter performs synchronous sampling to obtain a discrete voltage sequence and normalizes and reconstructs the position of the light spot.

[0047] Step S4: Utilize the differential calculation of the digital signal processing module inside the main control unit to modulate the frequency. Using the digital lock-in amplifier as a reference frame, the digital phase-sensitive detector performs digital phase-sensitive detection on the reconstructed spot position value, and extracts the characteristic amplitude at a specific frequency by calculating the magnitude of the orthogonal vector;

[0048] Step S5: Calculate the actual physical space differential displacement using the Fourier reconstruction formula based on the extracted feature amplitudes. Subsequently, the sensitivity mapping concentration equation pre-stored in the system is substituted. The concentration of anthocyanins in the current fermented fruit wine is calculated.

[0049] In some embodiments, in step S1: establishing a thermodynamic steady state inside a square vertical laminar flow channel, over time... Inside, the system's real-time temperature Controlled by a PID closed-loop algorithm, its control output law is expressed as:

[0050]

[0051] in:

[0052] : Continuous time variable during system initialization and temperature control establishment phase (unit: seconds, setting the initial time of temperature control stabilization module startup as...) =0);

[0053] Real-time output power control quantity of the temperature control stabilization module (unit: W);

[0054] Real-time temperature deviation is defined as follows: (Unit: °C)

[0055] : The set constant temperature reference target temperature (engineering setting is 25.0 ℃);

[0056] Temperature of the thermal jacket of the orthogonal optofluidic cell, collected in real time by the PT100 platinum resistance temperature sensor (unit: °C).

[0057] The integral time variable represents the time from start time 0 to the current time. Historical moments in the process;

[0058] At a historical moment Corresponding instantaneous temperature deviation (unit: °C);

[0059] Integral term in the formula It represents the sum of historical temperature deviations accumulated during system operation, used to eliminate steady-state errors in isothermal control;

[0060] , , These are the proportional, integral, and differential gain coefficients, respectively.

[0061] when The thermodynamic steady state is determined to be established after the temperature is ≤0.1℃ and maintained for 60 s.

[0062] In some embodiments, in step S1: establishing a steady state of fluid dynamics inside a square vertical laminar flow channel, the dimensionless Reynolds number inside the square vertical laminar flow channel ( The following relationship must be satisfied:

[0063]

[0064] : Fluid Reynolds number (dimensionless);

[0065] The average density of the fruit wine being tested (unit: kg / m³) 3 );

[0066] Average flow velocity of fruit wine liquid in the fluid channel of the orthogonal optical fluid control cell (unit: m / s).

[0067] Hydraulic diameter of the fluid channel in the orthogonal photofluidic cell (unit: m); for a square cross-section with side length a (2 mm × 2 mm). ;

[0068] Dynamic viscosity of fruit wine (unit: Pa·s);

[0069] The critical Reynolds number at which a fluid transitions from laminar to transitional flow.

[0070] In some embodiments, in step S1: the initial physical coordinate reference zero point of the light spot on the photosensitive target surface of the position-sensitive detector is established, and the initial photocurrent output from both ends of the position-sensitive detector is collected. and Calculate the initial reference geometric center position of the light spot at this time. :

[0071]

[0072] in:

[0073] Without thermal lensing effect, the initial physical coordinate reference zero point (unit: μm) of the probe spot on the PSD target surface is used as the basis for subsequent calculation of physical space differential displacement. Absolute zero point calibration reference;

[0074] Effective physical length of the photosensitive target surface of a position-sensitive detector (PSD) (unit: μm);

[0075] , : Initial substrate photocurrent at both ends of the PSD (unit: μA, including system dark current and dark noise background) read by the transimpedance amplifier (TIA) under calibration conditions; where subscripts "1" and "2" represent the physical signal output electrodes at both ends of the one-dimensional position-sensitive detector (PSD); subscript "0" represents the system in the initial baseline (Zero-baseline) calibration state without dual-wavelength pump source excitation.

[0076] In some embodiments, step S2: Time-division multiplexing is applied to the dual-wavelength pump light source to alternately output anthocyanin characteristic excitation light or reference light; the modulation frequency is set to... The adjustable range is 10 Hz to 1 kHz, and the modulation period is... During this period, the driving function of the dual-wavelength pump source... Defined as:

[0077]

[0078]

[0079] in:

[0080] : Switching parameters of the laser with characteristic excitation wavelength of anthocyanins (1 for constant power on, 0 for off);

[0081] : Switching state parameters of the reference wavelength laser (1 for constant power on, 0 for off);

[0082] Modulation period sequence number ( =0,1,2,…);

[0083] : Continuous time variable during the formal testing phase of the system (unit: s).

[0084] In the first half of the modulation cycle ( The pump light, driven by the characteristic excitation wavelength of anthocyanins, is focused at the interaction center of the orthogonal photofluidic cell. Trace amounts of anthocyanins in the fruit wine (mainly existing as flavonoid cations) strongly absorb photon energy in this wavelength range and release localized heat through non-radiative transitions. This results in a localized temperature rise. The instantaneous change in refractive index of the liquid is caused by the following equation:

[0085]

[0086] in:

[0087] : The vertical physical distance (in meters) from any point in the wine being tested to the central axis of the pump beam.

[0088] : Continuous time variables during the formal testing phase of the system (unit: s);

[0089] In state A, the radial distance from the center of the pump beam is The instantaneous change in refractive index at a point (dimensionless).

[0090] Thermo-optic coefficient (usually a negative value, unit: K) is the thermal-optic coefficient of the fruit wine matrix. -1 );

[0091] Total absorption coefficient of fruit wine at the characteristic excitation wavelength of anthocyanins (including characteristic anthocyanin absorption and background absorption, unit: m)-1 );

[0092] : The characteristic excitation wavelength of anthocyanins and the incident light power of the pump light (unit: W).

[0093] Fruit wine liquid density (unit: kg / m³) 3 );

[0094] Specific heat capacity of fruit wine under constant pressure (unit: J / (kg·K));

[0095] Thermal diffusivity of fruit wine (unit: m) 2 / s);

[0096] The heat conduction distribution function is related to the spatial distribution and relaxation time of the Gaussian beam. For a fundamental Gaussian-excited beam, its specific mathematical expansion is:

[0097]

[0098] in:

[0099] : Continuous time variables during the formal testing phase of the system (unit: s);

[0100] : The vertical physical distance (in meters) from any point in the wine being tested to the central axis of the pump beam.

[0101] : Beam waist radius of the pump beam at the interaction center (unit: m);

[0102] The characteristic relaxation time of a thermal lens (unit: s) is expressed as follows: This is used to characterize the timescale of local heat diffusion and dissipation into the surrounding area; among which, Thermal diffusivity of fruit wine (unit: m) 2 / s).

[0103] In the second half of the modulation cycle ( The pump light is switched to the reference wavelength. Since anthocyanins are at their absorption trough in this band, their absorption contribution is significantly lower than the characteristic excitation wavelength. The light energy in this band is mainly weakly absorbed by water, ethanol, and pectin suspensions (matrix components) in the fruit wine. The resulting instantaneous refractive index change distribution equation is:

[0104]

[0105] in:

[0106] In state B, the radial distance from the pump beam center is The instantaneous change in refractive index at a given point represents the pure "substrate interference background".

[0107] : The vertical physical distance (in meters) from any point in the wine being tested to the central axis of the pump beam.

[0108] : Continuous time variables during the formal testing phase of the system (unit: s);

[0109] Thermo-optic coefficient of fruit wine matrix (unit: K) -1 );

[0110] Absorption coefficient of fruit wine at the reference wavelength (representing only the background absorption characteristics of the fruit wine matrix, unit: m) -1 );

[0111] : Incident light power of pump light at the reference wavelength (unit: W);

[0112] Fruit wine liquid density (unit: kg / m³) 3 );

[0113] Specific heat capacity of fruit wine under constant pressure (unit: J / (kg·K));

[0114] Thermal diffusivity of fruit wine (unit: m) 2 / s);

[0115] : The heat conduction distribution function related to the spatial distribution and relaxation time of the Gaussian beam (its specific mathematical expansion and parameter definitions are the same as those above).

[0116] In some embodiments, step S3: synchronously turn on the detection light source, the detection beam passes through the dynamic refractive index gradient field in the interaction center, and undergoes periodic deflection, the detection beam deflection angle is measured. The integral equation can be expressed as:

[0117]

[0118] in:

[0119] : Continuous time variables during the formal testing phase of the system (unit: s);

[0120] : The instantaneous deflection angle of the probe beam after passing through the thermal lens region;

[0121] Fruit wine liquid at the reference temperature The initial background refractive index;

[0122] Thermo-optic coefficient of fruit wine matrix (unit: K) -1 );

[0123] The effective optical path length of the pump light and probe light that orthogonally overlap within the orthogonal optical flow control cell channel (due to channel limitations, the maximum is no more than 2 × 10⁻⁶). -3 m);

[0124] Local temperature gradient component along the direction perpendicular to the probe light (X-axis direction) (unit: K / m).

[0125] in, : Represents the spatial coordinate axis (unit: m) along the propagation direction of the continuous probe beam; : Represents the spatial coordinate axis (unit: m) that is perpendicular to the probe beam and along the propagation direction of the pump dual-wavelength light source excitation beam.

[0126] In some embodiments, step S3: the position-sensitive detector captures the displacement of the light spot in real time and converts it into photocurrent; the transimpedance amplifier converts the photocurrent into a voltage signal, the analog-to-digital converter performs synchronous sampling to obtain a discrete voltage sequence, and the light spot position is normalized and reconstructed. The specific parameter processing steps are as follows:

[0127] According to the principle of small angle approximation, a slight deflection angle Macroscopic physical displacement magnified on the PSD target surface Combined with the reference zero point Instantaneous absolute coordinates The mapping equation is:

[0128]

[0129] in:

[0130] : Continuous time variables during the formal testing phase of the system (unit: s);

[0131] : Absolute center position of the probe spot on the PSD target surface in the instantaneous state (unit: μm);

[0132] : The initial physical coordinate reference zero point of the probe spot on the PSD target surface when there is no thermal lensing effect (unit: μm);

[0133] : Instantaneous relative physical displacement of the probe spot due to the thermal lensing effect (unit: μm);

[0134] : The far-field physical flight distance from the center of the orthogonal interaction zone within the orthogonal optical fluidic cell channel to the PSD target surface (i.e., the length of the optical magnification arm, unit: μm).

[0135] : The instantaneous deflection angle of the probe beam after passing through the thermal lens region;

[0136] Therefore, under time-division multiplexing modulation, the position of the light spot on the PSD target surface will follow the modulation frequency. The periodic spatial jumps correspond to the positions in state A. Position in state B .

[0137] When a light spot illuminates the PSD target surface, it excites photogenerated carriers proportional to the light energy. Based on the transverse photoelectric effect, the instantaneous photocurrent output across the PSD... , With the position of the light spot They exhibit a strict proportional relationship. The photocurrent output equation is:

[0138]

[0139]

[0140] in:

[0141] : Continuous time variables during the formal testing phase of the system (unit: s);

[0142] , : These are the instantaneous photocurrents output across the first and second output electrodes of the PSD (unit: A);

[0143] The total photocurrent generated when the probe light from the probe light source is incident on the PSD target surface ( (Unit: A);

[0144] : Absolute center position of the probe spot on the PSD target surface in the instantaneous state (unit: μm);

[0145] The effective physical length of the photosensitive target surface of a position-sensitive detector (PSD) (unit: μm, needs to be compared with...) (The dimensions must remain consistent).

[0146] To convert the extremely weak photocurrent into a voltage signal that can be acquired by a high-precision analog-to-digital converter (ADC) without introducing additional phase delay and thermal noise, the system incorporates a transimpedance amplifier (TIA) network based on a precision operational amplifier. Its photoelectric signal conditioning and conversion model is as follows:

[0147]

[0148]

[0149] in:

[0150] : Continuous time variables during the formal testing phase of the system (unit: s);

[0151] , : These are the first and second instantaneous output voltage signals (unit: V) obtained after the photocurrents of the first and second output electrodes of the PSD are converted by the TIA network and then filtered by a low-bias active filter.

[0152] , : These are the instantaneous photocurrents output across the first and second output electrodes of the PSD (unit: A);

[0153] : Feedback resistor value of the transimpedance amplifier (unit: Ω);

[0154] : The reference bias voltage (unit: V) set to match the single-supply sampling range of the back-end ADC.

[0155] In some embodiments, step S4: using the differential calculation of the digital signal processing module inside the main control unit to modulate the frequency. Using the digital lock-in amplifier as a reference frame, digital phase-sensitive detection is performed on the reconstructed spot position values. By calculating the magnitude of the orthogonal vector, the characteristic amplitude at a specific frequency is extracted. The specific processing is as follows;

[0156] Synchronous oversampling and instantaneous position normalization reconstruction: Enable the dual-channel synchronous oversampling function of the ADC, based on the sampling period. (must meet) (Nyquist and oversampling conditions) Synchronous acquisition of discrete voltage sequences and To eliminate common-mode interference caused by absolute fluctuations in light source intensity (such as power drift in continuous probe lasers), the system first performs normalized reconstruction of the spot position and calculates the... Instantaneous original physical coordinates of each sampling point :

[0157]

[0158] in:

[0159] : No. The original one-dimensional position reconstruction value of the light spot on the PSD target surface at each sampling time (including all background noise and the real deflection signal, unit: μm);

[0160] , : in the At each discrete sampling moment, the high-precision ADC outputs the first instantaneous continuous voltage signal from the front-end signal conditioning module. With the second instantaneous continuous voltage signal The corresponding instantaneous equivalent voltage digital quantity obtained after synchronous digital conversion;

[0161] The digital value of the reference bias voltage of the hardware front-end TIA circuit;

[0162] The effective physical length of the photosensitive target surface of a position-sensitive detector (PSD) (unit: µm, needs to be compared with...) (The dimensions must remain consistent).

[0163] Digital Lock-In Amplification (DLIA) and Quadrature Demodulation: Reconstructed It contains the effects of fluid perturbations. Low-frequency noise and high-frequency white noise. This is because the actual photothermal deflection signal is strictly controlled by the modulation frequency. The system internally generates a pair of digital orthogonal reference frames that are strictly in sync with the excitation signal, and performs phase-sensitive detection. Within one solution cycle (including...) Within each sampling point, calculate the in-phase component. Orthogonal components :

[0164]

[0165]

[0166] in:

[0167] : No. The original one-dimensional position reconstruction value of the light spot on the PSD target surface at each sampling time (including all background noise and the real deflection signal, unit: μm);

[0168] Time-division multiplexing modulation frequency of dual-wavelength pump light source;

[0169] : Discrete sampling time interval of ADC (unit: s);

[0170] : Fixed phase delay (unit: rad) caused by the system's photothermal response and electronic links.

[0171] By calculating the magnitude of orthogonal vectors, the amplitude of weak signals at specific frequencies can be extracted. :

[0172] .

[0173] In some embodiments, step S5: Calculate the actual physical space differential displacement using the Fourier reconstruction formula based on the extracted feature amplitudes. Subsequently, the sensitivity mapping concentration equation pre-stored in the system is substituted. The concentration of anthocyanins in the current fermented fruit wine was calculated, and the specific processing was as follows:

[0174] Pure differential displacement equivalent reconstruction: Since the dual-wavelength alternating illumination is a square wave driven signal, the target displacement signal in the time domain exhibits an amplitude of The square wave transition. Extracted by DLIA. In reality, it is the amplitude of the fundamental wave (first harmonic) of the square wave. According to the Fourier series expansion, the system will... Reverse reconstruction into actual physical space differential displacement :

[0175]

[0176] in:

[0177] Physical space differential displacement (unit: μm) after eliminating all common-mode interference;

[0178] : Total deflection shift (unit: μm) resulting from the combined effects of anthocyanin-specific absorption and basal absorption under state A.

[0179] : Background deflection displacement (unit: μm) caused solely by the wine base (water, ethanol, suspended pectin, etc.) under state B.

[0180] This differential subtraction method completely eliminates the fixed thermal lensing effect caused by the high turbidity and dark background of the fruit wine itself. It becomes a physical quantity driven purely by "anthocyanin concentration".

[0181] Target component concentration mapping: The system retrieves a pre-calibrated concentration evolution function from non-volatile memory. Under the low concentration range and linear heat conduction model, physical space differential displacement... The actual concentration of anthocyanins in the wine The correlation is highly linear; in the medium-to-high concentration range, a polynomial correction coefficient is introduced for fitting.

[0182]

[0183] in:

[0184] : The current concentration of anthocyanins in the fruit wine (unit: mg / L);

[0185] The overall photothermal response sensitivity coefficient of the system (unit: mg / (L·μm)) is determined by hardware parameters such as pump light power, fluid specific heat capacity and optical magnification arm length.

[0186] Physical space differential displacement (unit: μm) after eliminating all common-mode interference;

[0187] Zero-point calibration intercept of the system (unit: mg / L).

[0188] In some embodiments, the algorithm based on digital lock-in amplifier and differential calculation can be replaced by a frequency domain feature peak extraction algorithm based on fast Fourier transform.

[0189] The hardware physical connections, nested algorithms, and their functions within the dual-wavelength photothermal deflection-based non-destructive testing system for trace anthocyanins in fermented fruit wine are as follows:

[0190] (1) Hardware physical connection relationships (including electrical, optical and fluid dynamic connections):

[0191] Control and Drive Link: The control signal output of the main control unit is electrically connected to the input of the modulation drive circuit; the power output of the modulation drive circuit is electrically connected to the input of the dual-wavelength pump light source. Simultaneously, the communication control terminal of the main control unit is electrically connected to the constant current sample pump, the probe light source, and the high-precision analog-to-digital converter (ADC).

[0192] Optical excitation and detection link: The output end of the dual-wavelength pump source is coupled to the input end of the dichroic mirror combiner in the spatial optical path; the output end of the dichroic mirror combiner is orthogonally aligned with the X-axis pump light incident window of the orthogonal optical fluid control cell in the spatial optical path. Specifically, the physical optical path and emission direction are as follows: the 780nm~1100nm and 500nm~550nm laser beams output from the dual-wavelength pump source are spatially collinearly combined by the dichroic mirror combiner to form a single pump beam with completely overlapping spatial paths. This combined pump beam is uniformly incident into the interaction center of the orthogonal optical fluid control cell along the X-axis direction. The 630 nm to 1600 nm probe beam output by the probe light source is orthogonally aligned with the Y-axis probe light window of the orthogonal optical fluid control cell in the spatial optical path. It is transmitted from the relatively parallel Y-axis receiving probe light window and the probe beam transmitted from the orthogonal optical fluid control cell illuminates the photosensitive target surface of the position-sensitive detector (PSD) in the far-field optical path.

[0193] Flow control and signal acquisition link: The constant flow injection pump is connected to the bottom liquid inlet of the orthogonal photofluidic cell via a fluid microtube, forming a laminar flow injection channel for the fruit wine to be tested. The current signal output of the position sensitive detector (PSD) is electrically connected to the input of the signal conditioning module; the voltage output of the signal conditioning module is electrically connected to the analog input of the high-precision analog-to-digital converter (ADC) 10; the digital signal output of the high-precision ADC is loop-connected back to the digital input bus of the main control unit.

[0194] (2) The algorithm nested within the hardware of the micro-anthocyanin nondestructive testing system based on dual-wavelength photothermal deflection and its function:

[0195] The core control unit (STM32 microcontroller) of this system internally integrates the following two core software algorithm modules to achieve coordinated control of the hardware and signal processing:

[0196] Module A: High-Frequency Time Division Multiplexing (TDM) PWM Timing Generation Program

[0197] Function: Running in the advanced timer of the main control unit, it is responsible for generating two strictly complementary square wave pulse width modulation (PWM) signals. This software directly determines the physical actions of the system on the time axis, controlling the dual-wavelength pump light source to perform microsecond-level seamless alternating excitation of 780nm~1100 nm and 500 nm~550 nm lasers, isolating spectral overlap interference from the source.

[0198] Module B: Digital Lock-in Amplifier (DLIA) and Differential Calculation Algorithm

[0199] Function: Running within the digital signal processing (DSP) core of the main control unit. It receives the raw discrete digital sequence from the high-precision analog-to-digital converter (ADC), which contains a large amount of low-frequency environmental thermal disturbances and high-frequency circuit white noise. Internally, it generates a digital orthogonal reference frame in phase and frequency with the excitation light for phase-sensitive detection. The algorithm extracts the weak thermal deflection displacement signal submerged in strong noise, performs differential subtraction, and finally maps and calculates the anthocyanin concentration.

[0200] Beneficial effects:

[0201] This invention achieves high-precision, non-destructive quantitative detection of trace anthocyanins in dark, highly turbid, and complex fermented wines (such as mulberry wine) through a comprehensive architecture of dual-wavelength time-division multiplexing alternating pumping, orthogonal laminar microfluidic design, and digital phase-sensitive differential computation. It has the following advantages and positive effects:

[0202] (1) Achieving non-destructive, in-situ, and online monitoring of core trace components in dark and complex wines: This invention employs purely physical photothermal deflection spectroscopy, eliminating the need for destructive pretreatment of the raw wine, such as large-scale dilution, high-speed centrifugation, or organic reagent extraction. The system can be directly connected to the side line of the fermentation tank or the bottling pipeline for continuous sample injection, significantly overcoming the shortcomings of traditional high-performance liquid chromatography (HPLC) in terms of detection lag, cumbersome operation, and the tendency of traditional absorbance methods to cause photodetector saturation blind zones.

[0203] (2) Suppressing strong light scattering interference from highly turbid matrices at the physical source by utilizing photothermal conversion mechanism: Addressing the issue that fruit wines rich in pectin, polysaccharides, and other macromolecules are prone to forming colloidal suspensions, this invention does not measure transmitted light intensity, but rather measures the refractive index gradient induced by the endothermic reaction of the target molecules (thermal lensing effect). Since the probe light is only sensitive to spatial deflection caused by local temperature rise, it is immune to Rayleigh and Mie scattering interference from macroscopic particles, greatly improving the penetration and effective resolution of optical signals in complex liquid matrices.

[0204] (3) Based on the dual-wavelength differential modulation and solution model, specific stripping and absolute quantification of target molecules are achieved: This invention innovatively introduces 780 nm~1100 nm near-infrared light as a background reference light that is not absorbed by anthocyanins, and alternates it with 500 nm~550 nm characteristic excitation light through high-frequency time-division multiplexing. By calculating the physical spatial differential shift induced by the two ( This method perfectly eliminates the basic thermodynamic background interference from water, ethanol, and non-target background pigments in the wine, achieving specific separation and weighted calibration of specific target signals in multi-component composite systems, and significantly improving the qualitative and quantitative reliability of concentration mapping.

[0205] (4) Combining orthogonal optical fluidic cell and precision temperature control design significantly improves the thermodynamic stability and repeatability of the measurement results: The photothermal deflection signal is extremely sensitive to environmental disturbances. This invention constructs a 2 mm × 2 mm orthogonal optical window microfluidic cell (orthogonal optical fluidic cell) with bottom inlet and top outlet, and supplements it with a semiconductor PID temperature control jacket, effectively eliminating bubble retention and fluid turbulence. This design locks the Reynolds number and environmental temperature drift of the detected microenvironment at an extremely low level, ensuring that the extremely weak refractive index gradient field will not be dispersed by natural convection, thus guaranteeing the stability and consistency of long-term measured data.

[0206] (5) Constructing a high signal-to-noise ratio weak signal conditioning and digital detection architecture to minimize system noise and achieve extremely low detection limits: For micrometer-level probe spot displacement, this invention employs an ultra-low bias precision transimpedance amplifier (TIA) network in the analog domain to losslessly convert weak differential photocurrent, and introduces a digital lock-in amplification (DLIA) algorithm in the digital domain that is in phase with the modulation frequency. This hardware-software co-operated narrowband detection mechanism significantly suppresses ambient stray light and... The wideband noise of the circuit gives the system extremely high sensitivity, enabling it to accurately track the early degradation trend of trace anthocyanins during shelf life.

[0207] (6) High system integration, strong application scenario generalization ability, and no chemical reagents required for operation: The all-solid-state laser, orthogonal optical fluidic cell, main control unit and sensing module of this invention can all be highly integrated and miniaturized. The entire detection process does not consume any chemical auxiliary reagents, which not only avoids secondary pollution, but also the detection logic and hardware architecture can be seamlessly transferred to the online quality monitoring and anti-counterfeiting traceability of a variety of high-value-added, dark-colored liquid phase foods such as black goji berry wine, blueberry wine and wine, which has excellent prospects for industrial engineering applications.

[0208] In summary, this invention innovatively applies photothermal deflection spectroscopy to liquid-phase food engineering by reducing the dimensionality of the technology. By cleverly combining differential modulation decoupling mechanism and microfluidic structure, it solves the industry problem of in-situ photometric quantitative detection of dark and highly turbid wines, significantly improving the specificity, sensitivity and real-time performance of the detection of target active ingredients. Attached Figure Description

[0209] The present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments, and the advantages of the present invention in the above and / or other aspects will become clearer.

[0210] Figure 1 This is a block diagram of the overall structure of a trace anthocyanin nondestructive testing system based on dual-wavelength photothermal deflection.

[0211] Figure 2 This is a schematic diagram of the temperature control jacket and internal microfluidic cell structure design for an orthogonal optical fluidic cell.

[0212] Figure 3 This is a cross-sectional view (XZ plane) of the center of the orthogonal optical fluidic cell.

[0213] Figure 4 This is a cross-sectional view (YZ plane) of the center of the orthogonal optical fluidic cell.

[0214] Figure 5 This is a schematic diagram of the interaction region of the orthogonal optical fluidic cell viewed from above (XY plane).

[0215] Figure 6 This is a typical representative graph of the dynamic change monitoring curve of anthocyanin concentration over time. Detailed Implementation

[0216] The present invention can be better understood from the following embodiments. However, those skilled in the art will readily understand that the descriptions in the embodiments are for illustrative purposes only and should not, and will not, limit the invention as detailed in the claims.

[0217] Example 1: In-situ dynamic non-destructive monitoring of trace anthocyanins in dark, highly turbid mulberry fermented wine

[0218] This embodiment takes mulberry fermented wine as an example to conduct in-situ dynamic non-destructive monitoring of trace anthocyanins in the wine.

[0219] Based on the photothermal deflection detection system constructed in this invention, the core effective component (cyanidin-3-glucoside, C3G) in mulberry wine during the late fermentation and shelf life is detected in situ without damage. This verifies the effectiveness of dual-wavelength time multiplexing modulation, orthogonal microfluidic physical isolation, and digital phase-sensitive differential calculation model in dark, highly scattering complex liquid phase systems.

[0220] (1) System composition

[0221] The schematic diagram of the temperature control jacket and internal microfluidic cell structure of the orthogonal photofluidic cell used in the embodiment is shown below. Figure 2 As shown; the central cross-sectional view (XZ plane) of the orthogonal optical fluidic cell is as follows. Figure 3 As shown; the central cross-sectional view (YZ plane) of the orthogonal optical fluidic cell is as follows. Figure 4 As shown; a schematic diagram of the interaction region of the orthogonal optical fluidic cell viewed from above (XY plane) is shown below. Figure 5 As shown.

[0222] The main framework of the orthogonal photofluidic cell 5 is constructed using PEEK (polyetheretherketone) material to create an insulating and heat-resistant body (utilizing its extremely low thermal conductivity to isolate environmental thermal disturbances and resist wine corrosion). Inside this body, a 2 mm × 2 mm square vertical laminar flow channel is formed, with a liquid inlet at the bottom and a liquid outlet at the top. The orthogonal photofluidic cell drives the wine to a stable laminar flow with a 2 mm × 2 mm cross-section, eliminating dead zones caused by scattering of bubbles and suspended matter. The flow rate must be controlled within a range that does not disrupt the dominant thermal diffusion mechanism to avoid damage to the thermal lens structure caused by fluid disturbances.

[0223] The square, vertical laminar flow channel is tightly fitted with a temperature control jacket made of a high thermal conductivity metal. To achieve extremely high-precision control of the micro-perturbation thermal field within the mulberry wine fluid, the orthogonal optical fluid control cell is equipped with a closed-loop temperature control stabilization module. This module integrates a Peltier semiconductor temperature control chip (TEC), a liquid cooling circulation chamber, and a PID precision temperature controller. The closed-loop temperature control stabilization module precisely locks in the ambient substrate temperature, creating an ideal thermodynamic detection microenvironment free from thermal drift.

[0224] Bidirectional high-precision temperature control: The TEC semiconductor temperature control chip is tightly attached to the outer flat wall of the temperature control jacket. The temperature sensor of the PID precision temperature controller (such as a PT100 platinum resistance temperature sensor) is attached to the outer wall of the temperature control jacket adjacent to the square vertical laminar flow channel, and is specifically arranged above the optical window (i.e., the downstream thermal equilibrium steady-state zone in the direction of fluid flow), without directly intruding into the square vertical laminar flow channel to prevent disruption of microscale laminar flow. Based on the real-time boundary temperature of the temperature control jacket fed back by the temperature sensor, the PID precision temperature controller dynamically changes the current direction to drive the TEC to perform high-frequency bidirectional heating or cooling compensation, thereby strictly locking the fluid temperature inside the square vertical laminar flow channel at the reference point (control accuracy can reach ±0.1℃).

[0225] Hot-end waste heat dissipation: The liquid cooling circulation chamber (water cooling head) is installed on the back of the TEC semiconductor temperature control chip (i.e. heat dissipation end). The side wall of the liquid cooling circulation chamber is provided with a coolant inlet and a coolant outlet for introducing external circulating cold water. Its sole function is to continuously remove the large amount of waste heat generated on the back of the TEC semiconductor temperature control chip during intense temperature control operation, ensuring the efficient operation of the Peltier effect of the TEC semiconductor temperature control chip.

[0226] With the direction of fluid flow in the square vertical laminar flow channel as the Z-axis, the square vertical laminar flow channel is provided with a Y-axis detection light window (JGS3 infrared quartz light window) and a Y-axis receiving detection light window (JGS3 infrared quartz light window) in the Y-axis direction. The two are arranged opposite each other on both sides of the square vertical laminar flow channel, and the windows are parallel to each other.

[0227] With the direction of fluid flow in the square vertical laminar flow channel as the Z-axis, a pump light incident window (JGS1 quartz window) is provided in the X-axis direction of the square vertical laminar flow channel, allowing the pump beam after beam combining to be incident on one side into the square vertical laminar flow channel.

[0228] To ensure unobstructed penetration of the orthogonal optical path, the aforementioned high thermal conductivity metal temperature control jacket and PEEK insulating body are respectively provided with through mechanical light-transmitting holes along the transmission paths of the X-axis pump light and Y-axis probe light. Each quartz optical window is embedded and sealed inside these mechanical light-transmitting holes, forming the light-transmitting sidewalls of the fluid channel. Simultaneously, the external TEC semiconductor temperature control chip and liquid cooling circulation chamber are physically positioned to strictly avoid the aforementioned cross-shaped orthogonal optical path. Thus, the closed-loop temperature control stabilization module, while efficiently conducting heat transfer around the square vertical laminar flow channel, does not interfere with the orthogonal optical probe path in three-dimensional space.

[0229] The overall structural block diagram of the trace anthocyanin nondestructive testing system based on dual-wavelength photothermal deflection used in the following embodiments is shown below. Figure 1 As shown, the physical hardware connections (including electrical, optical, and fluid dynamic connections) are as follows:

[0230] The control signal output terminal of the main control unit 1 is electrically connected to the input terminal of the modulation drive circuit 2; the power output terminal of the modulation drive circuit 2 is electrically connected to the input terminal of the dual-wavelength pump light source 3; the light output terminal of the dual-wavelength pump light source 3 is coupled to the input terminal of the dichroic mirror combiner 4 in the spatial optical path; the output terminal of the dichroic mirror combiner 4 is orthogonally aligned with the X-axis pump light incident window of the orthogonal optical fluid control cell 5 in the spatial optical path.

[0231] The specific physical optical path and emission direction of the dual-wavelength pump light are as follows: the 808nm laser beam or 532nm laser beam output by the dual-wavelength pump light source 3 is spatially collinearly combined by the dichroic mirror beam combiner 4 to form a pump beam with completely overlapping spatial paths. The combined pump beam is uniformly injected into the interaction center of the orthogonal optical fluid control cell 5 along the X-axis direction of the orthogonal optical fluid control cell 5.

[0232] Among them, the dual-wavelength pump source outputs dual-wavelength pump light (targeted excitation and background reference), which alternately irradiates to generate micro-area thermal lenses. The main control unit outputs time-division multiplexed control signals. The pump light source is driven alternately by frequency.

[0233] The communication control terminal of the main control unit 1 is electrically connected to the constant flow injection pump 6, the detection light source 7, and the high-precision analog-to-digital converter (ADC) 10, respectively. The constant flow injection pump 6 is connected to the bottom liquid inlet of the orthogonal optical fluid control cell 5 through a fluid microtube to form a laminar flow injection channel for the mulberry wine to be tested. The 1310 nm detection beam output by the detection light source 7 is orthogonally aligned with the Y-axis detection light window of the orthogonal optical fluid control cell 5 in the spatial optical path. It is transmitted from the relatively parallel Y-axis receiving detection light window and the detection beam of the orthogonal optical fluid control cell 5 is irradiated on the photosensitive target surface of the position-sensitive detector (PSD) 8 in the far-field optical path.

[0234] Among them, the probe light source provides a stable orthogonal probe optical path to map the deflection angle caused by the local refractive index gradient. .

[0235] The current signal output of the position sensitive detector (PSD) 8 is electrically connected to the input of the signal conditioning module 9; the voltage output of the signal conditioning module 9 is electrically connected to the analog input of the high-precision analog-to-digital converter (ADC) 10; and the digital signal output of the high-precision analog-to-digital converter 10 is closed-loop connected back to the digital input bus of the main control unit 1.

[0236] Among them, the position-sensitive detector converts the micro-radian level deflection angle into the micrometer level spatial displacement through the optical magnification arm; the signal conditioning module converts the extremely weak photocurrent (pA to µA level) into a voltage signal with a high signal-to-noise ratio and filters out high-frequency hardware noise.

[0237] (2) Assembly of experimental apparatus

[0238] Orthogonal Optical Fluidic Cell 5: The main body is insulated and heat-resistant with PEEK material, housing a 2 mm × 2 mm square vertical laminar flow channel. The Z-axis is defined by the direction of fluid flow within the square vertical laminar flow channel, and the X-axis is equipped with a pump light incident window (JGS1 quartz window for pump light incident). The Y-axis of the square vertical laminar flow channel features a Y-axis detection light window (JGS3 infrared quartz window for detection light transmission) and a Y-axis receiving detection light window (JGS3 infrared quartz window for detection light transmission), arranged opposite each other on both sides of the square vertical laminar flow channel, with the windows parallel to each other. A high thermal conductivity metal temperature control jacket is tightly fitted to the outside of the square vertical laminar flow channel, supplemented by a Peltier semiconductor temperature control chip (TEC), a liquid cooling circulation chamber, and a PID precision temperature controller (control accuracy ±0.1℃) to construct a closed-loop temperature control stabilization module.

[0239] Microfluidic injection system: A constant flow injection pump 6 (such as a constant flow micro peristaltic pump with adjustable flow rate from 0.1 mL / min to 5.0 mL / min) is used to continuously pump mulberry wine into the bottom liquid inlet of the orthogonal photofluidic cell 5.

[0240] Dual-wavelength pump source 3 and beam combining unit: includes a 532 nm solid-state laser (characteristic excitation) and an 808 nm laser diode (reference). The output beams of the two are collinearly combined by a dichroic mirror beam combiner 4 and orthogonally aligned and input to the X-axis pump light incident window of the orthogonal optical fluid control cell 5.

[0241] The detection light source 7 is a 1310 nm continuous wave fiber laser, which is aligned with the Y-axis detection light window of the orthogonal optical fluid control cell 5. The detection light beam transmitted from the relatively parallel Y-axis receiving detection light window illuminates the photosensitive target surface of the position-sensitive detector 8 in the far-field optical path.

[0242] Photoelectric sensing and conditioning: A one-dimensional silicon-based position-sensitive detector (PSD) 8 is configured on the far-field optical path, and its output is connected to a signal conditioning module 9 (including a transimpedance amplifier TIA based on OPA380 and a low-pass filter circuit of OPA2188).

[0243] Data processing and control module: includes main control unit 1 (such as STM32 microcontroller), modulation drive circuit 2, and high-precision analog-to-digital converter (ADC) 10. Main control unit 1 outputs PWM to drive dual-wavelength pump light source 3 through modulation drive circuit 2; the signal from signal conditioning module 9 is sampled by high-precision ADC 10 and transmitted to main control unit 1 and host computer to complete digital lock-in amplification and mapping calculation.

[0244] (3) Experimental sample: The pure juice fermented mulberry wine was taken from the 60th day of constant temperature aging. The wine was dark purple-black and had very low light transmittance (<2%). It contained a large amount of unsettled pectin and plant polysaccharides to form colloidal turbidity. The initial pH value was 3.5. The total concentration of its targeted anthocyanins was determined to be about 185 mg / L by offline liquid chromatography (HPLC).

[0245] (4) Detection method and process

[0246] Step 1: System initialization and fluid steady-state establishment

[0247] After the system starts up, in order to eliminate the interference of ambient temperature drift and fluid turbulence on the optically induced refractive index gradient field, it is necessary to establish the initial reference states in thermodynamics, fluid mechanics, and optics in sequence. The specific steps and parameter definitions are as follows:

[0248] Step 1.1: Construction of Steady-State Background Temperature Field

[0249] The main control unit activates the temperature control jacket (a temperature control stabilization module constructed based on a TEC semiconductor temperature control module, a PID precision temperature controller, a PT100 platinum resistance temperature sensor, and a liquid-cooled circulation chamber), locking the internal ambient temperature of the orthogonal optical fluidic cell to a preset reference point. In time... Inside, the system's real-time temperature Controlled by a PID closed-loop algorithm, its control output law is expressed as:

[0250]

[0251] in:

[0252] : Continuous time variable during system initialization and temperature control establishment phase (unit: seconds, setting the initial time of temperature control stabilization module startup as...) );

[0253] Real-time output power control quantity of the temperature control stabilization module (unit: W);

[0254] Real-time temperature deviation is defined as follows: (Unit: °C)

[0255] : The set constant temperature reference target temperature (engineering setting is 25.0 ℃);

[0256] : The temperature of the thermally conductive jacket of the orthogonal optical fluidic cell, acquired in real time by the PT100 platinum resistance temperature sensor (unit: °C); To avoid probe intrusion and disruption of the 2 mm × 2 mm microscale laminar steady state inside the flow channel, the sensor is attached to the metal thermally conductive wall adjacent to the flow channel, and the boundary temperature of the fluid inside the cavity is approximated by thermal conduction balance.

[0257] The integral time variable represents the time from start time 0 to the current time. Historical moments in the process;

[0258] At a historical moment Corresponding instantaneous temperature deviation (unit: °C);

[0259] Integral term in the formula It represents the sum of historical temperature deviations accumulated during system operation, used to eliminate steady-state errors in isothermal control;

[0260] , , These are the proportional, integral, and differential gain coefficients, respectively.

[0261] when The thermodynamic steady state is determined to be established after the temperature is ≤0.1℃ and maintained for 60 s.

[0262] Step 1.2: Laminar flow injection and hydrodynamic steady state

[0263] A high-precision micro-peristaltic pump is activated to inject the dark-colored fruit wine to be tested into the liquid inlet at the bottom of the orthogonal photofluidic cell. To prevent fluid turbulence from disrupting the microscopic "thermal lens" structure excited by the laser in subsequent steps, the flow rate must be strictly controlled. This ensures that the wine within the fluid channel is in an ideal laminar flow state. The dimensionless Reynolds number within the channel (…) The following relationship must be satisfied:

[0264]

[0265] : Fluid Reynolds number (dimensionless);

[0266] The average density of the fruit wine being tested (unit: kg / m³) 3 The average density of mulberry wine is approximately taken as 1015 kg / m³. 3 );

[0267] Average flow velocity of fruit wine liquid in the fluid channel of the orthogonal optical fluid control cell (unit: m / s).

[0268] Hydraulic diameter of the fluid channel in the orthogonal photofluidic cell (unit: m); for a square cross-section with side length a (2 mm × 2 mm). ;

[0269] Dynamic viscosity of fruit wine (unit: Pa·s);

[0270] The critical Reynolds number (usually taken as 2300) is the number at which a fluid transitions from laminar to transitional flow.

[0271] To ensure extremely low disturbance, the injection fluid dynamics parameters of this system are set as shown in Table 1 below:

[0272] Table 1. System Injection Fluid Dynamics Parameter Settings

[0273]

[0274] Step 1.3: Initial collimation and baseline calibration of the probe optical path

[0275] Under the conditions of thermodynamic and hydrodynamic steady state and without activating any dual-wavelength pump source (i.e., without introducing external photothermal excitation), the detection source (a 1310 nm continuous probe laser) is activated alone. The direction of fluid flow within the square vertical laminar flow channel of the orthogonal optical fluid control cell is taken as the Z-axis. A Y-axis detection window and a Y-axis receiving detection window are positioned opposite each other on either side of the square vertical laminar flow channel, with the windows parallel to each other. The detection beam orthogonally penetrates the Y-axis detection window of the orthogonal optical fluid control cell, passes over the central isothermal flowing fruit wine liquid, and then passes through the Y-axis receiving detection window, finally projecting onto the photosensitive target surface of the position-sensitive detector (PSD) at the far end. At this time, the system performs baseline calibration and collects the initial photocurrent output from both ends of the PSD. and Calculate the initial reference geometric center position of the light spot at this time. :

[0276]

[0277] in:

[0278] Without thermal lensing effect, the initial physical coordinate reference zero point (unit: μm) of the probe spot on the PSD target surface is used as the basis for subsequent calculation of physical space differential displacement. Absolute zero point calibration reference;

[0279] Effective physical length of the photosensitive target surface of a position-sensitive detector (PSD) (unit: μm);

[0280] , : Initial substrate photocurrent at both ends of the PSD (unit: μA, including system dark current and dark noise background) read by the transimpedance amplifier (TIA) under calibration conditions; where subscripts "1" and "2" represent the physical signal output electrodes at both ends of the one-dimensional position-sensitive detector (PSD); subscript "0" represents the system in the initial baseline (Zero-baseline) calibration state without dual-wavelength pump source excitation.

[0281] Step 2: Time-division multiplexing dual-wavelength alternating pump

[0282] After establishing the aforementioned fluid dynamics and thermodynamic baseline steady state, the system enters the core photothermal excitation stage. Utilizing the precise high-frequency timing logic output by the main control unit (STM32 microcontroller), the characteristic excitation light source and the reference light source are alternately driven to periodically induce a localized temperature field distribution within the micro-nano space of the fruit wine, thereby forming a dynamic refractive index gradient field (i.e., a "thermal lens"). Taking mulberry wine as an example, the characteristic excitation light is 532 nm green light, and the reference light is 808 nm near-infrared light. This wavelength parameter is also applicable to other fermented fruit wines containing anthocyanins. The specific steps and parameter definitions are as follows:

[0283] Step 2.1: Generation of high-frequency time-division multiplexed modulation sequence

[0284] The system's main control chip (STM32 microcontroller) uses an advanced timer module to generate strictly complementary square-wave pulse width modulation (PWM) signals to drive the dual-wavelength pump laser (dual-wavelength pump source) using time-division multiplexing (TDM). The modulation frequency is set to... The adjustable range is 10 Hz to 1 kHz, and the modulation period is... During this period, the driving function of the dual-wavelength pump source... Defined as:

[0285]

[0286]

[0287] in:

[0288] : Switching parameters of the 532 nm green laser (1 for constant power on, 0 for off);

[0289] : Switching parameters of the 808 nm near-infrared laser (1 for constant power on, 0 for off);

[0290] Modulation period sequence number ( =0,1,2,…);

[0291] : Continuous time variable during the formal testing phase of the system (unit: s).

[0292] This alternating modulation mechanism ensures that at any given time, only one wavelength of light energy is injected into the interaction center of the orthogonal optical fluidic cell, thus physically isolating spectral overlap interference.

[0293] Step 2.2: State A (targeted excited state) – 532 nm induced characteristic thermal lens

[0294] In the first half of the modulation cycle ( The 532 nm pump light is focused at the interaction center of the orthogonal photofluidic cell. Trace amounts of anthocyanins in the mulberry wine (mainly existing as flavonoid cations) strongly absorb photon energy in this wavelength band and release localized heat through non-radiative transitions. This results in a localized temperature rise. The instantaneous change in refractive index of the liquid is caused by the following equation:

[0295]

[0296] in:

[0297] The vertical physical distance (in meters) between any point in the wine being tested and the central axis of the pump beam (532 nm or 808 nm laser).

[0298] : Continuous time variables during the formal testing phase of the system (unit: s);

[0299] In state A, the radial distance from the center of the pump beam is The instantaneous change in refractive index at a point (dimensionless).

[0300] Thermo-optic coefficient (usually a negative value, unit: K) is the thermal-optic coefficient of the fruit wine matrix. -1 );

[0301] Total absorption coefficient of fruit wine at 532 nm (including anthocyanin characteristic absorption and background absorption, unit: m). -1 );

[0302] Incident light power of 532 nm pump light (unit: W);

[0303] Fruit wine liquid density (unit: kg / m³) 3 );

[0304] Specific heat capacity of fruit wine under constant pressure (unit: J / (kg·K));

[0305] Thermal diffusivity of fruit wine (unit: m) 2 / s);

[0306] The heat conduction distribution function is related to the spatial distribution and relaxation time of the Gaussian beam. For a fundamental Gaussian-excited beam, its specific mathematical expansion is:

[0307]

[0308] in:

[0309] : Continuous time variables during the formal testing phase of the system (unit: s);

[0310] The vertical physical distance (in meters) between any point in the wine being tested and the central axis of the pump beam (532 nm or 808 nm laser).

[0311] : Beam waist radius (m) of the 532 nm pump beam at the interaction center;

[0312] The characteristic relaxation time of a thermal lens (unit: s) is expressed as follows: This is used to characterize the timescale of local heat diffusion and dissipation into the surrounding area; among which, Thermal diffusivity of fruit wine (unit: m) 2 / s).

[0313] Step 2.3: State B (background reference state) – 808 nm induced substrate thermal lensing

[0314] In the second half of the modulation cycle ( The light is switched to 808 nm near-infrared pump light. Since anthocyanins are at their absorption trough in this band, their absorption contribution is significantly lower than the characteristic excitation wavelength. The light energy in this band is mainly weakly absorbed by the water, ethanol, and pectin suspension (matrix component) in the mulberry wine. The resulting instantaneous refractive index change distribution equation is:

[0315]

[0316] in:

[0317] In state B, the radial distance from the pump beam center is The instantaneous change in refractive index at a given point represents the pure "substrate interference background".

[0318] The vertical physical distance (in meters) between any point in the wine being tested and the central axis of the pump beam (532 nm or 808 nm laser).

[0319] : Continuous time variables during the formal testing phase of the system (unit: s);

[0320] Thermo-optic coefficient of fruit wine matrix (unit: K) -1 );

[0321] Absorption coefficient of fruit wine at 808 nm (represents only the background absorption characteristics of the fruit wine matrix, unit: m) -1 );

[0322] Incident light power of 808 nm pump light (unit: W);

[0323] Fruit wine liquid density (unit: kg / m³) 3 );

[0324] Specific heat capacity of fruit wine under constant pressure (unit: J / (kg·K));

[0325] Thermal diffusivity of fruit wine (unit: m) 2 / s);

[0326] : The heat conduction distribution function related to the spatial distribution and relaxation time of the Gaussian beam (its specific mathematical expansion and parameter definition are the same as in step 2.2 above).

[0327] To ensure energy equivalence, the system hardware is pre-calibrated to guarantee the actual optical power injected into the orthogonal optical fluidized cell. .

[0328] To clearly define the physical variable mapping relationship during this time-division multiplexing phase, the system's dual-wavelength photothermal response allocation parameters are shown in Table 2 below:

[0329] Table 2 System Dual-Wavelength Photothermal Response Assignment Parameters

[0330]

[0331] Step 3: Orthogonal deflection and photoelectric conversion

[0332] Simultaneously with the alternating pump light inducing a dynamic refractive index gradient field (thermal lens), the system utilizes orthogonal optical paths and a high-sensitivity position-sensitive detector (PSD) to convert microscopic thermodynamic changes into macroscopic optical spatial displacements, and ultimately demodulates them into electrical signals. The specific sensing steps and physical model are as follows:

[0333] Step 3.1: Continuously detect the orthogonal transmission and spatial deflection of light

[0334] With the probe light source (1310 nm continuous probe laser) kept on, the beam in this band penetrates orthogonally along the Y-axis through the central interaction region of the orthogonal optical fluidic cell. Since 1310 nm has good transmittance in mulberry wine and relatively weak additional photothermal effects, it serves as an ideal "probe." When passing over the temperature gradient field excited by alternating 532 nm / 808 nm pump light, according to Snell's Law and its continuum corollary, the beam will deflect towards the region with higher refractive index (i.e., the lower-temperature external region). Probe beam deflection angle. The integral equation can be expressed as:

[0335]

[0336] in:

[0337] : Continuous time variables during the formal testing phase of the system (unit: s);

[0338] : The instantaneous deflection angle of the probe beam after passing through the thermal lens region;

[0339] Fruit wine liquid at the reference temperature The initial background refractive index;

[0340] Thermo-optic coefficient of fruit wine matrix (unit: K) -1 );

[0341] The effective optical path length of the pump light and probe light that orthogonally overlap within the orthogonal optical flow control cell channel (due to channel limitations, the maximum is no more than 2 × 10⁻⁶). -3 m);

[0342] Local temperature gradient component along the direction perpendicular to the probe light (X-axis direction) (unit: K / m).

[0343] in, : Spatial coordinate axis (unit: m) representing the direction of propagation of the continuous probe beam (1310 nm); : Represents the spatial coordinate axis (unit: m) perpendicular to the probe beam and along the propagation direction of the pump dual-wavelength excitation beam (532 nm / 808 nm).

[0344] Step 3.2: Far-field spot displacement mapping

[0345] The deflected probe beam exits the Y-axis receiving probe window of the orthogonal optofluidic cell, travels a fixed geometric distance in free space, and is then projected onto the photosensitive target surface of the far-field position-sensitive detector (PSD). Based on the small-angle approximation principle, a slight deflection angle... Macroscopic physical displacement magnified on the PSD target surface Combined with the reference zero point in "Step 1.3" Instantaneous absolute coordinates The mapping equation is:

[0346]

[0347] in:

[0348] : Continuous time variables during the formal testing phase of the system (unit: s);

[0349] : Absolute center position of the probe spot on the PSD target surface in the instantaneous state (unit: μm);

[0350] : The initial physical coordinate reference zero point of the probe spot on the PSD target surface when there is no thermal lensing effect (unit: μm);

[0351] : Instantaneous relative physical displacement of the probe spot due to the thermal lensing effect (unit: μm);

[0352] The far-field physical flight distance from the center of the orthogonal interaction zone within the orthogonal optical fluidic cell channel to the PSD target surface (i.e., the length of the optical magnification arm, which is usually 100 mm-150 mm in engineering design to ensure sufficient displacement magnification, unit: μm).

[0353] : The instantaneous deflection angle of the probe beam after passing through the thermal lens region;

[0354] Therefore, under time-division multiplexing modulation, the position of the light spot on the PSD target surface will follow the modulation frequency. The periodic spatial jumps correspond to the positions in state A. Position in state B .

[0355] Step 3.3: PSD photoelectric conversion and weak signal transimpedance conditioning

[0356] When a light spot illuminates the PSD target surface, it excites photogenerated carriers proportional to the light energy. Based on the transverse photoelectric effect, the instantaneous photocurrent output across the PSD... , With the position of the light spot They exhibit a strict proportional relationship. The photocurrent output equation is:

[0357]

[0358]

[0359] in:

[0360] : Continuous time variables during the formal testing phase of the system (unit: s);

[0361] , : These are the instantaneous photocurrents output across the first and second output electrodes of the PSD (unit: A, typically nA to µA in actual engineering).

[0362] The total photocurrent generated when the probe light from the probe light source is incident on the PSD target surface ( (Unit: A);

[0363] : Absolute center position of the probe spot on the PSD target surface in the instantaneous state (unit: μm);

[0364] The effective physical length of the photosensitive target surface of a position-sensitive detector (PSD) (unit: μm, needs to be compared with...) (The dimensions must remain consistent).

[0365] To convert the extremely weak photocurrent into a voltage signal that can be acquired by a high-precision analog-to-digital converter (ADC) without introducing additional phase delay and thermal noise, the system incorporates a transimpedance amplifier (TIA) network built using a precision operational amplifier (e.g., an OPA380 chip). Its photoelectric signal conditioning and conversion model is as follows:

[0366]

[0367]

[0368] in:

[0369] : Continuous time variables during the formal testing phase of the system (unit: s);

[0370] , : These are the first and second instantaneous output voltage signals (unit: V) obtained after the photocurrents of the first and second output electrodes of the PSD are converted by the TIA network and then filtered by a low-bias active filter (such as a low-pass filter circuit composed of OPA2188).

[0371] , : These are the instantaneous photocurrents output across the first and second output electrodes of the PSD (unit: A, typically nA to µA in actual engineering).

[0372] The feedback resistor value of the transimpedance amplifier (unit: Ω) (determines the transimpedance gain of the system; a typical value is 10). 5 Ω-10 6 Ω);

[0373] : The reference bias voltage (unit: V) set to match the single-supply sampling range of the back-end ADC.

[0374] To clarify the hardware link characteristics of photoelectric conversion and displacement amplification, the sensing and conditioning parameters of the system are configured as shown in Table 3 below:

[0375] Table 3 System Sensing and Conditioning Parameter Configuration

[0376]

[0377] Step 4: Digital phase-sensitive detection and differential calculation

[0378] The system digitizes the front-end analog voltage signal using a high-precision analog-to-digital converter (ADC) and then sends it to the digital signal processing (DSP) module of the main control chip in the main control unit. Utilizing digital lock-in amplification (DLIA) and differential decomposition algorithms, it accurately extracts pure physical quantities relevant only to the target molecule from the raw data, which is often obscured by environmental thermal drift and broadband circuit noise. The specific decomposition model is as follows:

[0379] Step 4.1: Synchronous oversampling and instantaneous position normalization reconstruction

[0380] Enable the dual-channel synchronous oversampling function of the ADC, based on the sampling period. (must meet) (Nyquist and oversampling conditions) Synchronous acquisition of discrete voltage sequences and To eliminate common-mode interference caused by absolute fluctuations in light source intensity (such as power drift in continuous probe lasers), the system first performs normalized reconstruction of the spot position and calculates the... Instantaneous original physical coordinates of each sampling point :

[0381]

[0382] in:

[0383] : No. The original one-dimensional position reconstruction value of the light spot on the PSD target surface at each sampling time (including all background noise and the real deflection signal, unit: μm);

[0384] , : in the At each discrete sampling moment, the high-precision ADC outputs the first instantaneous continuous voltage signal from the front-end signal conditioning module. With the second instantaneous continuous voltage signal The corresponding instantaneous equivalent voltage digital quantity obtained after synchronous digital conversion;

[0385] The digital value of the reference bias voltage of the hardware front-end TIA circuit;

[0386] The effective physical length of the photosensitive target surface of a position-sensitive detector (PSD) (unit: µm, needs to be compared with...) (The dimensions must remain consistent).

[0387] Step 4.2: Digital Lock-in Amplification (DLIA) and Quadrature Demodulation

[0388] Reconstructed It contains the effects of fluid perturbations. Low-frequency noise and high-frequency white noise. This is because the actual photothermal deflection signal is strictly controlled by the modulation frequency. The system internally generates a pair of digital orthogonal reference frames that are strictly in sync with the excitation signal, and performs phase-sensitive detection. Within one solution cycle (including...) Within each sampling point, calculate the in-phase component. Orthogonal components :

[0389]

[0390]

[0391] in:

[0392] : No. The original one-dimensional position reconstruction value of the light spot on the PSD target surface at each sampling time (including all background noise and the real deflection signal, unit: μm);

[0393] Time-division multiplexing modulation frequency of dual-wavelength pump light source;

[0394] : Discrete sampling time interval of ADC (unit: s);

[0395] : Fixed phase delay (unit: rad) caused by the system's photothermal response and electronic links, which can be automatically compensated through hardware precalibration.

[0396] By calculating the magnitude of orthogonal vectors, the amplitude of weak signals at specific frequencies can be extracted. :

[0397]

[0398] Step 4.3: Pure differential displacement equivalent reconstruction

[0399] Since the alternating dual-wavelength illumination is driven by a square wave, the target displacement signal in the time domain exhibits an amplitude of The square wave transition. Extracted by DLIA. In reality, it is the amplitude of the fundamental wave (first harmonic) of the square wave. According to the Fourier series expansion, the system will... Reverse reconstruction into actual physical space differential displacement :

[0400]

[0401] in:

[0402] Physical space differential displacement (unit: μm) after eliminating all common-mode interference;

[0403] : Total deflection shift (unit: μm) resulting from the combined effects of anthocyanin-specific absorption and basal absorption under state A.

[0404] : Background deflection displacement (unit: μm) caused solely by the wine base (water, ethanol, suspended pectin, etc.) under state B.

[0405] This differential subtraction method completely eliminates the fixed thermal lensing effect caused by the high turbidity and dark background of the mulberry wine itself. It becomes a physical quantity driven purely by "anthocyanin concentration".

[0406] Step 4.4: Target component concentration mapping

[0407] The system retrieves a pre-calibrated concentration evolution function from non-volatile memory. Under the low concentration range and a linear heat conduction model, the physical space differential displacement... The actual concentration of anthocyanins in the wine The correlation is highly linear; in the medium-to-high concentration range, a polynomial correction coefficient is introduced for fitting.

[0408]

[0409] in:

[0410] : The current concentration of anthocyanins in the fruit wine (unit: mg / L);

[0411] The overall photothermal response sensitivity coefficient of the system (unit: mg / (L·μm)) is determined by hardware parameters such as pump light power, fluid specific heat capacity and optical magnification arm length.

[0412] Physical space differential displacement (unit: μm) after eliminating all common-mode interference;

[0413] Zero-point calibration intercept of the system (unit: mg / L).

[0414] To clearly illustrate the data flow and anti-interference dimensionality reduction process throughout the entire digital signal processing stage, the evolution of algorithm processing parameters is shown in Table 4 below:

[0415] Table 4. Evolution of Algorithm Processing Parameters in the Digital Signal Processing Stage

[0416]

[0417] (5) Setting core physical parameters

[0418] Fluid dynamics settings: The peristaltic pump flow rate is set to 1.0 mL / min to ensure that the Reynolds number Re < 50 in the square vertical laminar flow channel of the orthogonal optical fluid control cell 5, thus forming absolute laminar flow.

[0419] Thermodynamic settings: The target temperature of the TEC-based temperature control and stabilization module is locked at 25.0℃.

[0420] Optical power and modulation settings: The in-cell power of both the 532 nm and 808 nm lasers was precisely calibrated to 50 mW; the alternating square wave modulation frequency was set to... =20 Hz (modulation period) =50 ms).

[0421] (6) Test steps

[0422] Step S1: System Initialization and Reference Calibration

[0423] The mulberry wine sample to be tested was introduced into the microfluidic sample introduction system and pumped into the orthogonal photofluidic cell. A TEC-based temperature control and stabilization module was activated to establish a constant thermodynamic substrate. A 1310 nm continuous-wave fiber laser was turned on separately, and the initial dark background and static photocurrent at both ends of the PSD were collected to establish the initial physical coordinate reference zero point of the laser spot. .

[0424] Step S2: Dual-wavelength time-division multiplexing induction and physical separation

[0425] First half-cycle (characteristic excitation, 0-25 ms): The STM32 outputs a PWM signal to activate the 532 nm laser. Anthocyanins in the wine target and absorb green light, undergoing a non-radiative transition, resulting in a momentary local temperature rise. The thermal lens formed at this time contains "anthocyanin response + wine matrix background response".

[0426] Second half-cycle (reference, 25-50 ms): 532 nm laser off, seamlessly switching to 808 nm laser on. Because anthocyanins do not absorb near-infrared light, a localized temperature rise occurs at this time. This is caused solely by absorption of the wine's matrix (water, ethanol, pectin suspension).

[0427] Additional technical effects: By using time-division multiplexing for alternating dual-wavelength irradiation, effective "decoupling" of the target pigment response and the interference of the highly turbid background is achieved in both the time domain and physical space.

[0428] Step S3: Orthogonal deflection and synchronous dynamic data acquisition

[0429] The system performs continuous detection: a 1310 nm probe beam glides across an alternating refractive index gradient field, causing periodic deflection. A far-field PSD sensor captures the microscopic displacement oscillations of the beam spot in real time, converting them into differential photocurrent. A TIA conditioning circuit converts this into a high signal-to-noise ratio voltage signal, and a dual-channel ADC performs synchronous high-frequency oversampling at a sampling rate of 10 kHz to acquire a discrete voltage sequence containing the fundamental frequency and broadband noise.

[0430] Step S4: Extracting Digital Locked Magnification (DLIA) Parameters

[0431] Utilizing the computing power of the STM32's internal DSP, at a modulation frequency of 20 Hz Using a reference frame, digital phase-sensitive detection is performed on the ADC-acquired sequence. The signals are then multiplied by the in-phase reference signal. Orthogonal reference signal After low-pass filtering, the in-phase component is extracted. Orthogonal components Then, the fundamental characteristic amplitude, unaffected by low-frequency fluid perturbations and high-frequency circuit shot noise, is calculated. .

[0432] Step S5: Substrate decoupling and concentration correction

[0433] Based on the extracted feature amplitudes, the microcontroller calculates the actual physical space differential displacement using the Fourier reconstruction formula. Core logic: Utilize the "pure background thermal lens signal" of state B to calibrate the "mixed thermal lens signal" of state A in real time, and execute... This difference completely eliminates the baseline drift caused by color depth and pectin scattering. Then, it is substituted into the system's pre-stored sensitivity mapping equation. The current concentration of anthocyanins is calculated.

[0434] Step S6: Generation of dynamic concentration evolution map

[0435] The continuously calculated concentration data is transmitted in real time to a host computer (control device) to plot a dynamic monitoring curve with "time-concentration-temperature" as the coordinate system. The graph clearly shows the non-linear concentration decay trajectory of anthocyanins due to oxidative degradation during the accelerated aging experiment, triggering a quality warning when the concentration falls below a threshold. A typical example of anthocyanin concentration dynamic change monitoring curves over time is shown in the figure below. Figure 6 As shown.

[0436] (7) Effect verification and comparison

[0437] To verify the effectiveness of the method of this invention, the original wine sample of this embodiment was tested directly using a traditional ultraviolet-visible spectrophotometer (pH differential method) without any pretreatment, and the transmittance absorbance was compared. The results showed that:

[0438] Anti-turbidity interference capability: Traditional optical methods suffer from severe light scattering due to suspended pectin, resulting in severe transmission baseline drift and the detector often being in a blind zone overflow state, unable to output effective data; the photothermal differential method of this invention ignores scattering, and the baseline drift amplitude is reduced by nearly 95%.

[0439] Improved measurement accuracy: In 10 consecutive repeated measurements, the relative standard deviation (RSD) of the traditional direct measurement method is as high as 22.4%, while the RSD of the method of this invention is only 2.1%.

[0440] Non-destructive in-situ advantages: Traditional methods require 10 times the volume of buffer solution for demulsification, dilution, and centrifugation, taking more than 40 minutes; this invention can achieve online detection of pure stock solution in seconds.

[0441] The above results demonstrate that the present invention has significant technical effects in suppressing interference from dark, highly turbid matrices and extracting specific trace pigment signals.

[0442] Example 2: Alternative Solution

[0443] (1) The dual-wavelength pump source can be replaced by a wavelength-tunable laser (such as an OPO laser), or by other laser combinations with equivalent absorption differences, such as 520 nm (characteristic excitation) and 1064 nm (reference).

[0444] (2) The far-field spot position sensitive detector (PSD) can be replaced by a four-quadrant photodiode (QPD) or a high frame rate linear CCD / CMOS array camera.

[0445] (3) The microfluidic injection power unit (constant flow injection pump) can be replaced with a high-precision pulsation-free injection pump (SyringePump) or a pneumatically driven microfluidic system to further reduce fluid pulsation noise.

[0446] (4) The processing body (PEEK) of the orthogonal optical fluid cell can be replaced with a PDMS (polydimethylsiloxane) microfluidic chip bonded to JGS quartz glass, and its temperature control stabilization module can also be replaced with an external high-precision constant temperature water bath circulation system.

[0447] (5) The digital lock-in amplifier (DLIA) solution algorithm after transimpedance amplification can be replaced by a frequency domain feature peak extraction algorithm based on fast Fourier transform (FFT), or a pure hardware architecture analog lock-in amplifier circuit can be directly adopted.

[0448] (6) The continuous probe light source (1310 nm) can be replaced with a 1550 nm continuous wave fiber laser or a helium-neon (He-Ne) laser (632.8 nm, under specific low-concentration matrix).

[0449] This invention provides a method and system for the non-destructive detection of trace anthocyanins in fermented fruit wine based on dual-wavelength photothermal deflection. Many methods and approaches exist for implementing this technical solution; the above description is merely a preferred embodiment. It should be noted that those skilled in the art can make various improvements and modifications without departing from the principles of this invention, and these improvements and modifications should also be considered within the scope of protection of this invention. All components not explicitly stated in this embodiment can be implemented using existing technologies.

Claims

1. A non-destructive detection system for trace anthocyanins in fermented fruit wine based on dual-wavelength photothermal deflection, characterized in that, It includes a dual-wavelength pump source, a dichroic mirror beam combiner, an orthogonal optical fluid cell, a probe source, and a position-sensitive detector; in, The output end of the dual-wavelength pump light source is coupled to the input end of the dichroic mirror combiner in the spatial optical path; a square vertical laminar flow channel is formed inside the orthogonal optical fluid control cell, with the direction of the fermentation fruit wine fluid flow in the square vertical laminar flow channel as the Z-axis, and the output end of the dichroic mirror combiner is orthogonally aligned with the X-axis side of the square vertical laminar flow channel in the spatial optical path; the probe beam output by the probe light source is orthogonally aligned with the Y-axis side of the square vertical laminar flow channel in the orthogonal optical fluid control cell in the spatial optical path, and the probe beam transmitted from the orthogonal optical fluid control cell illuminates the photosensitive target surface of the position-sensitive detector in the far-field optical path.

2. The non-destructive detection system for trace anthocyanins in fermented fruit wine based on dual-wavelength photothermal deflection according to claim 1, characterized in that, The system also includes a main control unit, a modulation drive circuit, a constant current injection pump, a signal conditioning module, and an analog-to-digital converter; in, The control signal output terminal of the main control unit is electrically connected to the input terminal of the modulation drive circuit; the power output terminal of the modulation drive circuit is electrically connected to the input terminal of the dual-wavelength pump light source; the communication control terminal of the main control unit is electrically connected to the constant current injection pump, the probe light source and the analog-to-digital converter respectively. The constant flow injection pump is connected to the bottom liquid inlet of the orthogonal optical fluidic cell via a fluid microtube; the current signal output of the position-sensitive detector is electrically connected to the input of the signal conditioning module; the voltage output of the signal conditioning module is electrically connected to the analog input of the analog-to-digital converter; and the digital signal output of the analog-to-digital converter is closed-loop connected back to the digital input bus of the main control unit.

3. The non-destructive detection system for trace anthocyanins in fermented fruit wine based on dual-wavelength photothermal deflection according to claim 1, characterized in that, The orthogonal optical fluid control cell is constructed with an insulating and heat-resistant material as its main body. Inside this main body is a square vertical laminar flow channel, with a liquid inlet at the bottom and a liquid outlet at the top. A thermally conductive metal temperature control jacket is tightly fitted to the outer side of the square vertical laminar flow channel. The orthogonal optical fluid control cell is equipped with a closed-loop temperature control stabilization module, which integrates a TEC semiconductor temperature control chip, a liquid cooling circulation chamber, and a PID precision temperature controller. The TEC semiconductor temperature control chip is tightly fitted to the outer flat wall of the temperature control jacket. The temperature sensor of the PID precision temperature controller is attached to the outer wall of the temperature control jacket adjacent to the square vertical laminar flow channel, specifically downstream of the flow direction of the fermented fruit wine fluid. The liquid cooling circulation chamber is installed on the back of the TEC semiconductor temperature control chip. The side wall of the liquid cooling circulation chamber has a coolant inlet and a coolant outlet.

4. The non-destructive detection system for trace anthocyanins in fermented fruit wine based on dual-wavelength photothermal deflection according to claim 3, characterized in that, With the direction of fluid flow within the square vertical laminar flow channel as the Z-axis, a Y-axis probe window and a Y-axis receiver probe window are provided along the Y-axis of the square vertical laminar flow channel. These two windows are arranged opposite each other on both sides of the square vertical laminar flow channel, and the windows are parallel to each other. With the direction of fluid flow within the square vertical laminar flow channel as the Z-axis, a pump light incident window is provided along the X-axis of the square vertical laminar flow channel, allowing the combined pump beam to be incident on one side into the square vertical laminar flow channel. The probe window, receiver probe window, and pump light incident window are all made of quartz. On the transmission paths of the X-axis pump light and Y-axis probe light corresponding to the square vertical laminar flow channel, the temperature control jacket and the main body are respectively provided with through mechanical light-transmitting holes, and the quartz light windows are respectively embedded and sealed inside the mechanical light-transmitting holes to form the light-transmitting sidewall of the square vertical laminar flow channel. The pump light output from the dual-wavelength pump source is spatially collinearly combined by a dichroic mirror combiner to form a pump beam with completely overlapping spatial paths. This combined pump beam enters the square vertical laminar flow channel through the pump light incident window. The probe beam output from the probe source is orthogonally aligned with the Y-axis probe light window of the orthogonal optical fluid control cell in the spatial optical path, passes through the square vertical laminar flow channel, and is transmitted out from the relatively parallel Y-axis receiving probe light window. The probe beam transmitted from the orthogonal optical fluid control cell illuminates the photosensitive target surface of the position-sensitive detector in the far-field optical path. In the square vertical laminar flow channel, the pump beam output from the dual-wavelength pump source and the probe beam output from the probe source intersect and are orthogonal in space, and the intersection point is the interaction center.

5. The non-destructive detection system for trace anthocyanins in fermented fruit wine based on dual-wavelength photothermal deflection according to claim 2, characterized in that, The main control unit outputs a square wave pulse width modulation signal through a modulation drive circuit to drive the dual-wavelength pump light source in a time-division multiplexing manner, so that the dual-wavelength pump light source alternately outputs anthocyanin characteristic excitation light or reference light; the wavelength of the anthocyanin characteristic excitation light is 500 nm to 550 nm; the wavelength of the reference light is 780 nm to 1100 nm; and the wavelength of the probe light output by the probe light source is 630 nm to 1600 nm.

6. The non-destructive detection system for trace anthocyanins in fermented fruit wine based on dual-wavelength photothermal deflection according to claim 2, characterized in that, The dual-wavelength pump source can be replaced with a tunable laser; the position-sensitive detector can be replaced with a four-quadrant photodiode or a linear CCD / CMOS array camera; and the constant-flow injection pump can be replaced with an injection pump or a pneumatically driven microfluidic system.

7. The non-destructive detection system for trace anthocyanins in fermented fruit wine based on dual-wavelength photothermal deflection according to claim 2, characterized in that, The system also includes a control device for generating and sending control commands; the main control unit is communicatively connected to the control device and is used to control the execution device according to the received control commands; a transimpedance amplifier is built into the signal conditioning module; a digital lock-in amplifier is built into the digital signal processing module of the main control unit. The current signal output terminal of the position-sensitive detector is connected to the input terminal of the transimpedance amplifier; the output terminal of the transimpedance amplifier is connected to the analog input terminal of the analog-to-digital converter; the digital signal output terminal of the analog-to-digital converter is connected to the input terminal of the digital lock-in amplifier; and the output terminal of the digital lock-in amplifier is communicatively connected to the control device.

8. A method for non-destructive detection of trace anthocyanins in fermented fruit wine using a dual-wavelength photothermal deflection system according to any one of claims 1 to 7, characterized in that, Includes the following steps: System initialization and benchmark calibration: Establish the thermodynamic steady state and hydrodynamic steady state inside the square vertical laminar flow channel, and establish the initial physical coordinate reference zero point of the light spot on the photosensitive target surface of the position-sensitive detector; After system initialization and reference calibration, the system is started, and the dual-wavelength pump light source operates based on a preset modulation frequency. Alternating outputs of anthocyanin-specific excitation light or reference light are directed into the interaction center of a square vertical laminar flow channel. Simultaneously, a probe light source is activated, and as the probe beam passes over the interaction center, a position-sensitive detector captures the real-time displacement of the light spot. The displacement of the light spot is then preprocessed to obtain the physical spatial differential displacement. The concentration of anthocyanins in the current fermented fruit wine was calculated using the sensitivity mapping concentration equation.

9. The method according to claim 8, characterized in that, The modulation frequency 10 Hz - 1 kHz; ;in, Total deflection shift resulting from the combined effects of anthocyanin-specific absorption and basal absorption, unit: μm; Background deflection displacement caused solely by the base of the fermented fruit wine, in μm; The sensitivity-mapping concentration equation is as follows: ;in, The overall photothermal response sensitivity coefficient of the system, in mg / (L·μm); Zero-point calibration intercept of the system, unit: mg / L.

10. The method according to claim 8, characterized in that, When the absolute value of the real-time temperature deviation inside the square vertical laminar flow channel is ≤0.1℃ and is maintained for 60 s, the system determines that the thermodynamic steady state has been established; when the Reynolds number of the fermented fruit wine fluid inside the square vertical laminar flow channel is <50, the system determines that the hydrodynamic steady state has been established.

Images