A raman spectrum detection method, device and dynamic monitoring method for microalgae oil

Abstract

The application discloses a kind of microalgae oil Raman spectrum detection method, equipment and dynamic monitoring method, belong to spectral detection technical field.The method first in first microalgae sample with different laser power acquisition time domain signal, according to characteristic peak intensity determines damage threshold, and constructs power gradient sequence, and then from time domain signal extraction each laser power corresponding gate delay and gate width, current gate delay is calculated and fitted to obtain the gate gradient function between laser power and gate width.Subsequently to second microalgae sample sequentially using the laser power in this power gradient sequence carries out fluorescence quenching processing, if residual fluorescence background intensity is higher than preset threshold, then return to adjust laser power, otherwise under corresponding gate delay, enable gate switch, call gate gradient function to determine gate width, collect second Raman spectrum.Finally based on second Raman spectrum obtains the lipid distribution data of second microalgae sample.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of spectral detection, in particular to a Raman spectrum detection method, device and dynamic monitoring method for microalgae oil. BACKGROUND

[0002] Microalgae is an important potential source of producing bio-oil and biodiesel, and the synthesis of its oil is a dynamic biological process. Although the traditional chemical extraction method is accurate, it is extremely time-consuming and completely destroys the cell structure, resulting in the complete loss of all key spatial distribution information about oil in the cell population (heterogeneity) and single cell (such as lipid droplet distribution). Raman spectrum technology is sensitive to lipid chemical bonds and does not require complex staining, and is suitable for lipid analysis in principle. Chinese patent application No. CN118688179A discloses a method for real-time in-situ analysis of microalgae biochemical components combining Raman spectrum detection and neural network model. The method collects the Raman original spectrum of the microalgae sample, constructs and trains a neural network model combining the dry weight of the microalgae cells and the fucoxanthin yield, and finally realizes quantitative prediction of the target variable. However, microalgae cells are rich in chlorophyll, carotenoids and other fluorescent substances, which will produce fluorescence background much stronger than Raman signal under laser excitation, seriously submerging the weak characteristic Raman peak, leading to difficulty in extracting characteristic information and limiting the accuracy of the quantitative model. Chinese patent application No. CN113252637A discloses a fluorescence background suppression system and method in Raman spectrum detection. The method enhances signal collection efficiency and suppresses fluorescence background by combining a specially designed fiber probe (central excitation, surrounding collection) with a coded aperture technique. However, this method mainly focuses on static optimization at the hardware level, and its suppression effect largely depends on the fixed optical configuration. Fluorescence suppression lacks prior algorithms, and the optical configuration cannot adaptively and actively regulate the dynamic changes of the fluorescence characteristics of the sample itself. Therefore, this method needs to be further improved. SUMMARY

[0003] To solve the above problems, the present application provides a Raman spectrum detection method and device for microalgae oil. The present application selects and optimizes the laser power and the gating width through power gradient and gating gradient, and then performs laser surface scanning and Raman spectrum detection. This method effectively reduces fluorescence interference, avoids damaging the target monitoring components, and improves the signal-to-noise ratio and detection accuracy of Raman imaging.

[0004] Further, the present application also provides a dynamic monitoring method for microalgae stress treatment. The present application performs Raman spectrum detection on microalgae samples treated by ultrasonic stress to generate the total amount of lipids in different stress time periods. The present application uses a set of first microalgae samples, avoiding repeated power and gating parameter adjustment.

[0005] The application object of the application can be achieved by the following technical means:

[0006] A Raman spectrum detection method of microalgae oil, based on the following steps:

[0007] Step 1: Select multiple test areas in the first microalgae sample, use a first laser with different laser powers to irradiate the test areas respectively, disable the gate switch, and collect corresponding multiple time domain signals;

[0008] Step 2: Generate the first Raman spectrum of the time domain signal, and calculate the characteristic peak intensity, determine the damage threshold of the laser power according to the characteristic peak intensity, and determine the power gradient sequence according to the damage threshold;

[0009] Step 3: Find the gate delay and gate width of each laser power according to the time domain signal, fit the gate gradient function according to different laser powers and corresponding gate widths, and calculate the current gate delay according to different gate delays;

[0010] Step 4: Select the laser power from the power gradient sequence in turn, and use a second laser with the laser power to perform fluorescence quenching treatment on the observation area of the second microalgae sample;

[0011] Step 5: After quenching, use the second laser to face scan the observation area of the second microalgae sample, enable the gate switch under the current gate delay, select the gate width of the second laser from the gate gradient function, and collect the second Raman spectrum;

[0012] Step 6: Calculate the fluorescence background intensity in the second Raman spectrum, if the fluorescence background intensity is greater than the quenching threshold, return to step 4, otherwise enter step 7;

[0013] Step 7: Extract multiple lipid characteristic peaks based on the second Raman spectrum, and generate the lipid distribution data of the second microalgae sample according to the signal intensity of the lipid characteristic peaks.

[0014] In the application, in step 2, the Raman shift range of ±15 cm -1 around the Raman shift of any Raman characteristic peak is selected as the characteristic peak window of the Raman characteristic peak, multiple baseline points are selected on both sides of the characteristic peak window, a polynomial is used to fit the baseline points to obtain a baseline spectrum, the signal intensity of the first Raman spectrum in the characteristic peak window is subtracted from the signal intensity of the corresponding baseline spectrum to obtain the net signal intensity of the characteristic peak window, and the integral area of the net signal intensity corresponding to each Raman characteristic peak is the characteristic peak intensity.

[0015] In the present application, in step 2, the power interval is determined according to the sampling sensitivity and the damage threshold of the device, the power center is determined according to the preset proportion of the damage threshold, the laser power is extracted from the power center to the direction of the sampling sensitivity and the damage threshold at equal intervals, and a power gradient sequence is formed.

[0016] In the present application, in step 3, the peak time t1 of the first maximum value and the valley time t2 of the first minimum value of the time domain signal are found, the delay T1=t1-t0 is gated, t0 is the emission time of the first laser, and the width T2=2(t2-t1) is gated.

[0017] In the present application, in step 3, the mean value of the delay of different laser powers is the current delay, the different laser powers and the corresponding gate widths are extracted, and based on the nonlinear relationship between the laser power and the corresponding gate width, a function of the gate width monotonically decreasing with the increase of the laser power and asymptotically tending to the laser pulse width is fitted as the gate gradient function.

[0018] In the present application, in step 5, the observation area of the second microalgae sample is scanned, the scanning step length in the horizontal axis direction is Δx, the scanning step length in the vertical axis direction is Δy, Δx and Δy are determined according to the spot diameter d of the second laser, the time domain signal containing the scanning position information is obtained, and the second Raman spectrum of the time domain signal is extracted based on the corresponding gate width.

[0019] In the present application, in step 6, a background band of the second Raman spectrum is selected, the average value of the signal intensity of all spectral data points in the background band is calculated, and the average value is taken as the fluorescence background intensity under the corresponding laser power. The background band is a continuous interval that does not contain a characteristic peak window.

[0020] In the present application, in step 7, based on the signal intensity of one or more lipid characteristic peaks in the second Raman spectrum corresponding to each scanning position, the total amount of lipids at the scanning position is calculated, and the first content matrix is constructed according to the total amount of lipids and the scanning position of all scanning positions. Spatial difference method is performed on the first content matrix to generate the second content matrix, the total amount of lipids in the second content matrix is converted into pixel value, and the lipid distribution data is generated.

[0021] A Raman spectrum detection device for implementing the Raman spectrum detection method of the microalgae oil, comprising:

[0022] A tunable laser configured to emit a first laser or a second laser;

[0023] A slide configured to carry the first microalgae sample or the second microalgae sample;

[0024] A confocal optical path assembly configured to focus the first laser or the second laser on the slide.

[0025] a gated detection module configured to collect Raman scattering signals and generate time-domain signals, the gated detection module having a gating switch;

[0026] a signal analysis module configured to reconstruct first and second Raman spectra in a frequency domain from the time-domain signals;

[0027] a data analysis module configured to determine a power gradient sequence from the first Raman spectrum and fit a gating gradient function;

[0028] an image generation module configured to generate lipid distribution data from the second Raman spectrum;

[0029] a system control module configured to control the tunable laser, the slide, and the gated detection module,

[0030] wherein the system control module controls laser power of the tunable laser according to the power gradient sequence, the system control module controls gating width of the gated detection module according to the gating gradient function, and the system control module moves the slide according to a focus position of the first or second laser.

[0031] A dynamic monitoring method for microalgae stress treatment, comprising the following steps:

[0032] Step 100: divide the microalgae sample to be tested into a first group of microalgae samples and multiple groups of second microalgae samples, and treat the second microalgae samples in different time periods by using an ultrasonic stress method, wherein the parameters of the ultrasonic stress method are as follows: the ultrasonic frequency range is 20-80 kHz, the ultrasonic power density range is 5-275 W / L, the stress treatment mode is pulse mode, the single pulse duration is 1-10 minutes, and the pulse interval is 0.5-2 hours;

[0033] Step 200: detect the first and second microalgae samples according to the Raman spectrum detection method of the microalgae oil, and generate multiple groups of lipid distribution data;

[0034] Step 300: combine the multiple groups of lipid distribution data to construct a lipid distribution data spectrum sequence, and generate the total amount of lipids in different time periods.

[0035] The Raman spectrum detection method, device and dynamic monitoring method of the microalgae oil fat according to the present application have the beneficial effects that: the Raman spectrum detection method first performs self-adaptive regulation in the safe range of the laser power, determines the damage threshold by testing the characteristic peak intensity under different laser powers, and constructs the power gradient sequence of the safe excitation based on the damage threshold, so that the cell damage or lipid oxidation problem caused by the high-power laser is effectively avoided. And the best gating width is automatically matched according to the fitted gating gradient function, so that the time gating parameter is no longer dependent on the artificial experience setting. The best quenching effect is obtained by the method of real-time evaluation of the fluorescence background intensity, the second Raman spectrum based on the gating switch is collected after quenching, and the effective control of the fluorescence background is realized. Finally, the second Raman spectrum of the second microalgae sample is collected by accurately adjusting the current gating delay and gating width, the delay fluorescence interference is maximally reduced, the signal-to-noise ratio of the lipid characteristic peak is improved, and the high-resolution imaging of the lipid distribution data of the second microalgae sample is realized. BRIEF DESCRIPTION OF DRAWINGS

[0036] Figure 1 A flowchart of the Raman spectrum detection method of the microalgae oil fat according to the present application;

[0037] Figure 2 A comparison chart of the main detection substances with different Raman shifts according to the present application;

[0038] Figure 3 A schematic diagram of the first laser and the time domain signal according to the present application;

[0039] Figure 4 A schematic diagram of the background waveband according to the present application;

[0040] Figure 5 A schematic diagram of the gating gradient function according to the present application;

[0041] Figure 6 A schematic diagram of the original second Raman spectrum of different observation regions according to the present application;

[0042] Figure 7 A schematic diagram of the corrected second Raman spectrum according to the present application;

[0043] Figure 8 A block diagram of the Raman spectrum detection device for implementing the Raman spectrum detection method of the microalgae oil fat according to the present application;

[0044] Figure 9 A light path diagram of the Raman spectrum detection method according to the present application, the solid arrows are optical signals, and the dashed arrows in the diagram are electrical signals;

[0045] Figure 10 A flowchart of the dynamic monitoring method of the microalgae stress treatment according to the present application;

[0046] Figure 11 A schematic diagram of the lipid distribution data spectrum of different stress time periods of the present application;

[0047] Figure 12 A schematic diagram of the total amount of lipids of different stress time periods of the present application.

[0048] In the drawing, the reference signs are: tunable laser 1, band-pass filter 2, mirror 3, first lens 4, first beam splitter 5, broadband light source 6, filter 7, second beam splitter 8, second lens 9, glass slide 10, optical path adjustment module 11, gate switch 12, photodetector 13. DETAILED DESCRIPTION

[0049] The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application.

[0050] Compared with traditional methods, Raman spectroscopy can directly collect spectra of intact cells, and can qualitatively identify the class of lipid compounds and relatively quantitatively evaluate the content change by analyzing the signal intensity of the characteristic peaks in a specific Raman shift interval. However, microalgae cells are rich in chlorophyll and pigment proteins and other components, which will produce strong autofluorescence background. When the laser is excited, the signal intensity of the fluorescence background is often several orders of magnitude higher than the Raman signal, resulting in the complete annihilation of the weak lipid characteristic peaks.

[0051] In addition, after the laser photons are absorbed by the fluorescent molecules, the fluorescent molecules are transitioned to a high-energy state. Under high-intensity irradiation, these fluorescent molecules cannot release energy back to the ground state through the normal way, and the chemical bonds of the fluorescent molecules will irreversibly break or change structure, thereby permanently losing the ability to emit light. The energy threshold required for this process is relatively low. When the laser energy is absorbed and converted into local heat, it causes the vibration of lipid molecules to intensify. When the temperature rises to a certain extent, it may cause the breakage of chemical bonds or the oxidation and isomerization of lipid molecules, but the energy threshold required for this process is relatively high. Therefore, the present application accurately determines the safe power gradient sequence that does not significantly damage the lipid molecules through pre-experiments, thereby selecting the second laser, and using the second laser for fluorescence quenching pretreatment of the second microalgae sample, and then using the gate switch to obtain the second Raman spectrum of the second microalgae sample, to obtain the lipid distribution data of the second microalgae sample. This method effectively removes the fluorescence interference, while ensuring the integrity of the microalgae sample structure, and significantly improves the Raman signal quality.

[0052] A preferred implementation process of the present application is as follows. Example 1

[0053] Reference Figure 1 The Raman spectroscopy detection method for microalgae oil of the present application described in detail in this embodiment includes the following steps:

[0054] Step 1: Select multiple test regions in the first microalgae sample, irradiate the test regions with the first laser at different laser powers, disable the gate switch, and collect the corresponding multiple time-domain signals. The first microalgae sample is divided into multiple test regions with the same area. In this embodiment, the linear dimension of each test region is not less than 10 times the spot diameter of the first laser. In another embodiment, a single microalgae cell in the first microalgae sample can also be set as a test region. At this time, the spot size of the first laser is configured to be not larger than the size of the single cell to achieve single-cell resolution irradiation and detection.

[0055] There is significant subcellular structure heterogeneity inside the microalgae cell, for example, chloroplasts are rich in chlorophyll, lipid droplets are rich in neutral lipids, nuclei contain nucleic acids, and the composition of the cytoplasmic matrix is also uneven. If laser irradiation is performed only at a single site, the collected Raman signal may only reflect a local component (such as pure chlorophyll or pure lipid droplets). Therefore, in this embodiment, the first laser is used to irradiate and collect time-domain signals at multiple spatial sites corresponding to the test region, and the time-domain signal of the test region is generated based on the mean value of the time-domain signals of all spatial sites in the test region.

[0056] Step 2: Generate the first Raman spectrum of the time-domain signal and calculate the characteristic peak intensity, determine the damage threshold of the laser power according to the characteristic peak intensity, and determine the power gradient sequence according to the damage threshold. Select a Raman shift interval of ±15 cm -1 around the Raman shift of any Raman characteristic peak as the characteristic peak window of the Raman characteristic peak, select multiple baseline points on both sides of the characteristic peak window, fit the baseline points with a polynomial to obtain a baseline spectrum, subtract the signal intensity of the corresponding baseline spectrum from the signal intensity of the first Raman spectrum in the characteristic peak window to obtain the net signal intensity of the characteristic peak window, and the integral area of the net signal intensity corresponding to each Raman characteristic peak is the characteristic peak intensity. In this embodiment, the selected Raman characteristic peaks are selected from the peak positions related to lipids in microalgae. Specifically, referring to Figure 2 , the following peak positions are selected as Raman characteristic peaks: 1007 cm -1 -1009 cm -1 , 1066 cm -1 , 1266 cm -1 , 1445 cm -1 , and 1660 cm -1 .

[0057] Within the range where photo-damage does not occur, the intensity of Raman characteristic peaks usually increases with the increase of laser power (approximately linear), but once the damage threshold is exceeded, the intensity of characteristic peaks will suddenly decrease, distort or disappear, when the intensity of characteristic peaks first decreases, it indicates that the photo-thermal or photo-chemical effect of laser has begun to produce detectable and irreversible effects on the chemical structure or spatial conformation of the target lipid molecules, and the laser power at this time is taken as the damage threshold. According to the sampling sensitivity and damage threshold of the device, the power interval is determined, the power center is determined according to the preset proportion of the damage threshold, and the laser power is extracted from the power center to the direction of the sampling sensitivity and the damage threshold at equal intervals to form a power gradient sequence.

[0058] Step 3: Find the gate delay and gate width of each laser power according to the time domain signal, fit the gate gradient function according to the different laser powers and the corresponding gate width, and calculate the current gate delay according to the different gate delays. As Figure 3 , find the peak time t1 of the first maximum value and the valley time t2 of the first minimum value of the time domain signal, the gate delay T1 = t1-t0, t0 is the emission time of the first laser, and the gate width T2 = 2(t2-t1). Raman scattering is a transient process, which occurs immediately after laser pulse, and fluorescence emission involves the lifetime of electronic excited state, usually with nanosecond delay, so in the time domain signal, the early stage is Raman signal + elastic scattering, and then the fluorescence signal gradually rises, mainly Raman signal before the minimum value, and mainly fluorescence signal after the minimum value. The mean value of the gate delay of different laser powers is the current gate delay, and the different laser powers and the corresponding gate width are extracted, and based on the nonlinear relationship between the laser power and the corresponding gate width, a function of the gate width monotonically decreasing with the increase of the laser power and asymptotically tending to the laser pulse width is fitted as the gate gradient function, for details, refer to Example Three.

[0059] Step 4: Select laser power from the power gradient sequence in turn, and use the second laser with the laser power to perform fluorescence quenching treatment on the observation area of the second microalgae sample. From the power gradient sequence, select laser power in turn, and use the second laser corresponding to the laser power to irradiate the observation area for a fixed irradiation time to perform fluorescence quenching treatment. The fixed irradiation time is 1-5 seconds, and in this embodiment, the fixed irradiation time is preferably 3 seconds, at which time only the fluorescence background pretreatment step is performed, and the time domain signal is not collected, so the gate switch does not need to be enabled. The wavelength of the second laser is the same as that of the first laser.

[0060] Step 5: After quenching, the observation area of the second microalgae sample is scanned by the second laser, the gate switch is enabled under the current gate delay, the gate width of the second laser is selected from the gate gradient function, and the second Raman spectrum is collected. The observation area of the second microalgae sample is scanned, the scanning step in the horizontal axis direction is Δx, the scanning step in the vertical axis direction is Δy, Δx and Δy are determined according to the spot diameter d of the second laser, and the value ranges of Δx and Δy are both [2d, 3d]. The time domain signal containing the scanning position information is obtained, and the second Raman spectrum of the time domain signal is extracted based on the corresponding gate width.

[0061] Step 6: The fluorescence background intensity in the second Raman spectrum is calculated, if the fluorescence background intensity is greater than the quenching threshold, return to step 4, otherwise enter step 7. Referring to the method for obtaining the characteristic peak window of the first Raman spectrum in step 2, the characteristic peak window of the second Raman spectrum is obtained, the continuous interval not containing the characteristic peak window is taken as the background waveband, and the background waveband of the second Raman spectrum is selected according to Figure 4 , the average value of the signal intensity of all spectral data points in the background waveband is calculated, and the average value is taken as the fluorescence background intensity under the corresponding laser power. The quenching threshold is set based on the analysis requirement of the lipid characteristic peak, in this embodiment, the average net signal intensity of the lipid characteristic peak is measured by pre-experiment, and the maximum allowed fluorescence background intensity is calculated inversely according to the criterion of ensuring that the signal-to-noise ratio of the lipid characteristic peak is not less than 10, and the maximum allowed fluorescence background intensity is taken as the quenching threshold. It is verified that the reasonable value range of the quenching threshold is 100 to 500 a.u.

[0062] Step 7: A plurality of lipid characteristic peaks are extracted based on the second Raman spectrum, and the lipid distribution data of the second microalgae sample is generated according to the signal intensity of the lipid characteristic peak. Based on the signal intensity of one or more lipid characteristic peaks in the second Raman spectrum corresponding to each scanning position, the total amount of lipids at the scanning position is calculated, a first content matrix is constructed according to the total amount of lipids at all scanning positions and the scanning positions, a spatial difference method is performed on the first content matrix, a second content matrix is generated, the total amount of lipids in the second content matrix is converted into pixel value, and the lipid distribution data is generated. The lipid characteristic peak is a Raman shift located in the interval of 600-2000 cm -1 , which is related to the carbon skeleton vibration or carbon-carbon unsaturated bond vibration in the lipid molecule. In this embodiment, the lipid characteristic peak is 1066 cm -1 C-C stretching bond, 1266 cm -1 H-C= bending plane, 1445 cm -1 C-H2 bending bond and 1660 cm -1 C=H stretching bond. Example Two

[0063] Referring toFigure 5 The embodiment further discloses a fitting method of the gated gradient function in step 3, and a method of determining the gating width according to the laser power when the gated gradient function is actually used in step 5.

[0064] Fitting the gated gradient function. Under the condition of disabling the gating switch, a series of discrete laser powers P are selected, the time-domain signal of the observation area of the first microalgae sample corresponding to each laser power is collected, and the gating width T2 is extracted based on the time-domain signal. The laser power P is taken as the independent variable, and the gating width T2 is taken as the dependent variable. The exponential decay function is fitted by using the nonlinear least squares method, and the gated gradient function T2 = τ + (T0-τ)e -kP wherein τ is the laser pulse width, representing the minimum time resolution of the equipment, k is the decay coefficient, e is the natural constant according to the experimental data fitting, and T0 is the gating width under the minimum power of the equipment, which is usually determined according to the sensitivity of the equipment.

[0065] Determination of the gating width. The laser power in step 5 has been determined, and the theoretical optimal gating width T3 is first calculated by calling the gated gradient function. However, due to the limitation of the actual hardware performance, the gating width is usually discretely set with a fixed time step, and cannot be accurately configured with any continuous value. Therefore, the theoretical optimal gating width T3 is mapped to the nearest available hardware gear: T4 = round(T3 / Δt)Δt, wherein Δt is the minimum time adjustment step of the gating detection module (for example, Δt = 100 ps), round() represents rounding, and T4 is the gating width of the laser power. Embodiment three

[0066] Reference Figure 6 and Figure 7 The embodiment further discloses a preferred method of generating lipid distribution data according to the second Raman spectrum.

[0067] Constructing the lipid quantitative model. In this embodiment, partial least squares regression is used as the basic modeling framework to eliminate the multicollinearity between the Raman spectrum variables and extract the most relevant potential variables of the lipid. A series of microalgae samples with gradient total lipid content are prepared, and the total lipid content of the microalgae samples is accurately measured by using a chemical reference method (such as Soxhlet extraction method) as a reference value Y. The Raman spectrum of these microalgae samples is collected under the same conditions, and the net signal intensity (the peak area or peak height of the lipid characteristic peak after baseline correction of the Raman spectrum) of the lipid characteristic peak in the Raman spectrum is extracted as the independent variable X. The linear or nonlinear calibration curve between X and Y is established by least squares fitting, and the calibration curve is the lipid quantitative model for converting the signal intensity into the total lipid content.

[0068] Baseline correction of the second Raman spectrum. The least square fitting method can automatically adapt to different fluorescence background patterns and achieve accurate subtraction of baseline drift. To eliminate the interference of abnormal peaks caused by cosmic rays on the characteristic peaks of lipids in the second Raman spectrum, the least square fitting method is used in this embodiment: taking each spectral data point in the second Raman spectrum as the center, a low-order polynomial fitting is performed in the window composed of the previous and next n spectral data points. When the measured value of any spectral data point deviates from the fitted value by more than K times the noise standard deviation threshold, the spectral data point is determined to be an abnormal point and is replaced by the fitted value. For the fluorescence background caused by microalgae itself, an asymmetric weighted least square smoothing algorithm is used for adaptive baseline estimation, that is, a baseline spectrum vector b with the same length as the original second Raman spectrum y is solved. The objective function for solving the baseline spectrum vector b is where I is the total number of spectral data points in the second Raman spectrum, y i is the signal intensity of the second Raman spectrum at the i th spectral data point, b i is the estimated intensity of the to-be-solved baseline spectrum vector at the i th spectral data point, and the weight w i is asymmetrically distributed according to the difference between y i and b i , that is, a smaller weight w i is given when y i >b i , and a larger weight w i is given when y i ≤b i , and γ is a smoothing penalty coefficient for balancing the fitting error and the baseline smoothness, the value of γ is determined by leave-one-out cross-validation optimization, (Δ 2 b i ) is the second-order difference of the baseline spectrum vector, that is, Δ 2 b i =b i-1 -b i +b i+1 , Δ 2 b i is used to constrain the curvature smoothness of the baseline. The above objective function is minimized by iterative optimization, and finally the baseline spectrum vector b adapted to the second Raman spectrum is obtained. After subtracting b from the original second Raman spectrum y, the corrected second Raman spectrum is obtained.

[0069] Construction of lipid distribution data. The corrected second Raman spectral data matrix is imported into the established lipid quantitative model, and the total amount of lipids q corresponding to each scanning position is obtained by projection operation. The total amount of lipids is mapped to the corresponding scanning position to generate a first content matrix. After the first content matrix is subjected to bicubic spline interpolation to improve the spatial resolution, the system deviation between samples is eliminated by quantile normalization to obtain a second content matrix. Finally, the total amount of lipids in the second content matrix is converted into hue and saturation parameters of the HSV color space through a linear color mapping function c = aq + b, where a controls the color dynamic range, b sets the reference hue, and c is the pixel value. The spatial encoding false color image generated is the lipid distribution data. Example Four

[0070] With reference to Figure 8 A Raman spectrum detection device for implementing the Raman spectrum detection method of microalgae oil, comprising: a tunable laser, a glass slide, a confocal optical path assembly, a gated detection module, a signal analysis module, a data analysis module, an image generation module, and a system control module. The tunable laser is configured to emit a first laser or a second laser. The glass slide is configured to carry the first microalgae sample or the second microalgae sample. The confocal optical path assembly is configured to focus the first laser or the second laser on the glass slide. The gated detection module is configured to collect Raman scattering signals and generate time-domain signals, and the gated detection module has a gating switch. The signal analysis module is configured to reconstruct first and second Raman spectra in the frequency domain according to the time-domain signals. The data analysis module is configured to determine a power gradient sequence according to the first Raman spectrum and fit a gating gradient function. The image generation module is configured to generate lipid distribution data according to the second Raman spectrum. The system control module is configured to control the tunable laser, the glass slide, and the gated detection module. Further, the system control module controls the laser power of the tunable laser according to the power gradient sequence. The system control module controls the gating width of the gated detection module according to the gating gradient function. The system control module moves the glass slide according to the focusing position of the first laser or the second laser.

[0071] With reference to Figure 9The confocal optical path assembly includes a band-pass filter 2, a filter 7, a first beam splitter 5, a second beam splitter 8, a mirror 3, a first lens 4, and a second lens 9. The gated detection module includes a photodetector 13 and a gated switch 12. In this embodiment, a broadband light source 6 is further included, which is used to output broadband light covering the visible light band to assist in illumination. When step 1 is performed, the tunable laser 1 emits first laser light, which first passes through a band-pass filter 2 to screen specific wavelengths (for example, 532 nm), is then turned by a mirror 3, is collimated by a first lens 4, and is incident on a first beam splitter 5. At the same time, the broadband light source 6 emits broadband light, which is filtered by a filter 7 and then coupled to the first beam splitter 5. The first laser light and the broadband light are combined at the first beam splitter 5, are transmitted through the second beam splitter 8, and are then focused by the second lens 9 to the surface of a first microalgae sample on a slide 10, thereby achieving excitation of the first laser light and auxiliary illumination of the broadband light. Raman scattered photons generated by the first microalgae sample after being excited by the laser light return along the original optical path, are collected and reflected by the second beam splitter 8, then enter an optical path adjustment module 11, and are transmitted to the photodetector 13 for collection of time-domain signals. The time-domain signals are finally analyzed by a signal analysis module. The optical path adjustment module 11 is used to control the relative time of the Raman scattered photons reaching the photodetector 13, thereby achieving precise synchronization with the gated switch 12. Embodiment five

[0072] With reference to Figure 10 A method for dynamically monitoring stress treatment of microalgae, comprising the following steps:

[0073] Step 100: Divide the microalgae sample to be tested into a group of first microalgae samples and multiple groups of second microalgae samples, and treat the second microalgae samples in different time periods by using an ultrasonic stress method. The parameters of the ultrasonic stress method are as follows: the ultrasonic frequency range is 20 kHz-80 kHz, and the ultrasonic power density range is 5 W / L-275 W / L. The stress treatment is in a pulse mode, the single pulse duration is 1-10 minutes, and the pulse interval is 0.5-2 hours. In this embodiment, the second microalgae samples are divided into an experimental group and a blank control group. The blank control group is not treated, and the experimental group is treated by using ultrasonic waves with frequencies of 20 kHz, 28 kHz, 40 kHz, and 68 kHz respectively.

[0074] Step 200: Detect the first microalgae samples and the second microalgae samples according to the Raman spectrum detection method of microalgae oil, and generate multiple groups of lipid distribution data. Extract the lipid characteristic peaks (1066 cm - 1 C-C stretch, 1266 cm -1 H-C= in-plane bending, 1445 cm -1C-H2 bending and 1660 cm -1 C=H stretch) to generate pseudo-color lipid distribution data, and to realize the visualization characterization of the lipid density gradient at the subcellular scale.

[0075] Step 300: Constructing a lipid distribution data spectrum sequence by combining multiple sets of lipid distribution data, and generating the total amount of lipids at different time periods. By analyzing the time series of the lipid distribution data spectrum sequence, the dynamic process and spatial heterogeneity of microalgae lipid accumulation at different time periods under stress treatment are quantitatively revealed. Specifically, as shown in Figure 11 , typical lipid distribution data on the 1st, 2nd, 3rd, 4th, 5th, and 6th days of stress treatment are respectively displayed. As shown in Figure 12 , the total amount of lipids at different time periods is respectively displayed. Within two days of stress, the total amount of lipids in each experimental group is always stable in the range of 28.1%-28.5%, and with the increase of stress time, the total amount of lipids in the experimental group increases, and gradually stabilizes on the 4th day. Specifically, the total amount of lipids in the experimental groups corresponding to 20kHz, 28kHz, 40kHz, and 68kHz increases to 33.27%, 33.45%, 33.31%, and 33.07% respectively on the 5th day. From the lipid distribution data spectrum sequence and the total amount of lipids, it can be observed that the lipid distribution and total amount of the blank control group do not change significantly within the 6-day cultivation period, while the total amount of intracellular lipids in all experimental groups (20kHz, 28kHz, 40kHz, and 68kHz) shows a gradient rising trend with the extension of cultivation time. Among them, under the condition of applying 28kHz ultrasonic wave stress, the lipid metabolism response of microalgae cells is the most significant and efficient. Example Six

[0076] In this embodiment, Scenedesmus obliquus (No. FACHB-12) is selected as the microalgae sample to be tested, and the dynamic monitoring method for microalgae stress treatment is implemented. The Scenedesmus obliquus has important application value in the field of bioenergy due to its high proliferation rate and significant oil-protein double-component accumulation characteristics.

[0077] Pre-treatment before stress treatment. After the Scenedesmus obliquus was activated by the BG11 standard medium (1.7g per liter of culture solution), it was subjected to photoautotrophic amplification culture in a glass reactor to obtain a microalgae suspension. The cultivation system was strictly controlled at the following optimized parameters: light intensity 3500-4000 lux (incandescent light source, light-dark cycle 12h:12h), oscillation frequency once every 12 hours, and temperature gradient maintained at 26±1°C. To accurately control the best induction method of lipid synthesis, the biomass was dynamically monitored using a multi-dimensional method: every 48 hours, 5ml of uniform microalgae suspension was taken, centrifuged at 3000xg for 2min, and the supernatant was used to measure OD 680Values, with deionized water as blank control. And 8 μL microalgae suspension (pre-treated by hemocytometer) was taken synchronously to complete under inverted fluorescence microscope (IX73, Olympus, Japan), and microalgae cell identification and counting statistics were performed with ViewProfessional Suite image analysis system (version 2.4). When the culture system entered the late exponential growth phase (OD 680 ≥1.5 and microalgae cell concentration >5 x 10 6 cells / mL), it was determined as the optimal induction window period of lipid synthesis metabolism, and the microalgae suspension was used as the microalgae sample to be tested. The microalgae sample to be tested was divided into a first microalgae sample and a second microalgae sample. The first microalgae sample was used to determine the power gradient sequence, and the second microalgae sample was used to start the subsequent ultrasonic stress treatment.

[0078] Selection of ultrasonic stress treatment device. The ultrasonic stress treatment device used was HGD3000 program-controlled ultrasonic generating device (frequency range 20-80 kHz, power control 0-100% numerical control, driving power 200-2200 W), equipped with four groups of modular titanium alloy transducers (nominal frequency 20, 28, 40, 60 kHz). The main body of the ultrasonic stress treatment device was seamlessly coupled with the bottom of the customized sonochemical reactor (cylindrical glass cavity, inner diameter 12 cm, effective volume 8 L) through silicon-based heat-conducting glue, realizing efficient conduction of acoustic energy. In this embodiment, the working parameters of the ultrasonic stress treatment device were set as follows: constant power 100 W, pulse time 3 minutes, and basic working voltage 220 V (±10% fluctuation compensation).

[0079] Ultrasonic stress treatment. After centrifugation (4000 rpm, 5 min) and 3 times of phosphate buffer washing, 5 L of the second microalgae sample was resuspended in fresh BG11 medium and distributed into 5 groups of parallel photobioreactors (6 L / group, initial OD 680 =0.8±0.05), light intensity 180 μmol / m² / s, continuous aeration (2% CO2, 0.2 vvm), temperature maintained at 25±0.5°C, and the same light-dark cycle as the culture stage. Four groups of the second microalgae sample were used as experimental groups, and one group of the second microalgae sample was used as a blank control group. Numerical control program was used to apply 20 kHz, 28 kHz, 40 kHz and 60 kHz ultrasonic treatment (fixed parameters: constant power 100 W, 3 minutes pulse per hour) to the four groups of experimental groups, and the blank control group only maintained the basic culture.

[0080] Collecting lipid distribution data. From the start of ultrasonic stress treatment for 6 consecutive days, every 24 h, accurately weigh 3 g of agar powder into 100 mL of distilled water, magnetically stir and boil for 3 min until completely dissolved, cool naturally to 45°C to obtain a warm agar solution. Mix 1 mL of the second microalgae sample in the experimental group with 3 mL of the warm agar solution evenly, quickly pour it into a sterile petri dish, and after it is fully solidified, use a sterile blade to cut the agar gel into a thickness of about 1 mm. A 532 nm laser excitation confocal Raman spectrometer (spectral range covering 491-2185 cm -1 , the laser power determined by the Raman spectrum detection method of the microalgae oil is 15 mW, and the 50x objective lens is focused) to collect the second Raman spectrum of the second microalgae sample. Among them, the observation area of the second microalgae sample is a single microalgae cell, and a face scan is performed in the observation area with a step size of 0.7 μm (20×20 μm² observation area), and the lipid distribution data of the second microalgae sample is obtained. The face scan passes through the silicon wafer 520.5 cm -1 The characteristic peak corrects the wave number drift and eliminates the matrix interference by background (pure agar) subtraction to ensure the reliability of the spectral data. A total of 30 sets of lipid distribution data are collected within 6 days, and then a sequence of lipid distribution data spectra is obtained.

[0081] The above only describes the preferred embodiments of the present application and is not intended to limit the present application. Any modifications, equivalent replacements and improvements made within the spirit and principle of the present application shall be included in the protection scope of the present application.

Claims

1. A Raman spectroscopy method for detecting microalgal lipids, characterized in that, Includes the following steps: Step 1: Select multiple test areas in the first microalgae sample, irradiate the test areas with first lasers of different laser powers, disable the gating switch, and collect the corresponding multiple time-domain signals; Step 2: Generate the first Raman spectrum of the time-domain signal and calculate the intensity of the characteristic peak. Determine the damage threshold of the laser power based on the intensity of the characteristic peak, and determine the power gradient sequence based on the damage threshold. Step 3: Find the gate delay and gate width for each laser power based on the time domain signal, fit the gate gradient function with different laser powers and corresponding gate widths, and calculate the current gate delay based on different gate delays; Step 4: Select laser power sequentially from the power gradient sequence, and use the second laser of that power to perform fluorescence quenching treatment on the observation area of ​​the second microalgae sample; Step 5: After quenching, the second laser surface is used to scan the observation area of ​​the second microalgae sample. Under the current gate delay, the gate switch is activated, the gate width of the second laser is selected from the gate gradient function, and the second Raman spectrum is acquired. Step 6: Calculate the fluorescence background intensity in the second Raman spectrum. If the fluorescence background intensity is greater than the quenching threshold, return to step 4; otherwise, proceed to step 7. Step 7: Extract multiple lipid characteristic peaks based on the second Raman spectrum, and generate lipid distribution data for the second microalgae sample based on the signal intensity of the lipid characteristic peaks. In step 3, the peak time t1 when the time-domain signal first reaches a maximum and the valley time t2 when the signal first reaches a minimum are found. The gating delay T1 = t1 - t0, where t0 is the emission time of the first laser, and the gating width T2 = 2(t2 - t1). In step 3, the mean gating delay for different laser powers is the current gating delay. Different laser powers and their corresponding gating widths are extracted. Based on the nonlinear relationship between the laser power and the corresponding gating width, a function is fitted where the gating width monotonically decreases with increasing laser power and asymptotically approaches the laser pulse width as the gating gradient function. In step 5, the observation area of ​​the second microalgae sample is scanned, with a scanning step size of Δx in the horizontal direction and Δy in the vertical direction. Δx and Δy are determined according to the spot diameter d of the second laser. The values ​​of Δx and Δy are both in the range of [2d, 3d]. A time-domain signal containing scanning position information is obtained, and the second Raman spectrum of the time-domain signal is extracted based on the corresponding gate width.

2. The Raman spectroscopy detection method for microalgal lipids according to claim 1, characterized in that, In step 2, a range of ±15cm is selected, centered on the Raman shift where any Raman characteristic peak is located. -1 The Raman shift interval is taken as the characteristic peak window of the Raman characteristic peak. Multiple baseline points are selected on both sides of the characteristic peak window. The baseline points are fitted by a polynomial to obtain the baseline spectrum. The signal intensity of the first Raman spectrum in the characteristic peak window is subtracted from the signal intensity of the corresponding baseline spectrum to obtain the net signal intensity of the characteristic peak window. The integral area of ​​the net signal intensity corresponding to each Raman characteristic peak is the characteristic peak intensity.

3. The Raman spectroscopy detection method for microalgal lipids according to claim 1, characterized in that, In step 2, the power range is determined based on the sampling sensitivity and damage threshold of the device, the power center is determined according to the preset ratio of the damage threshold, and the laser power is extracted sequentially at equal intervals from the power center in the direction of sampling sensitivity and damage threshold to form a power gradient sequence.

4. The Raman spectroscopy detection method for microalgal lipids according to claim 2, characterized in that, In step 6, the background band of the second Raman spectrum is selected, and the average signal intensity of all spectral data points in the background band is calculated. This average value is used as the fluorescence background intensity under the corresponding laser power. The background band is a continuous interval that does not contain the characteristic peak window.

5. The Raman spectroscopy detection method for microalgal lipids according to claim 4, characterized in that, In step 7, the total lipid content at each scanning position is calculated based on the signal intensity of one or more lipid characteristic peaks in the second Raman spectrum corresponding to each scanning position. A first content matrix is ​​constructed based on the total lipid content at all scanning positions and the scanning position. The spatial difference method is performed on the first content matrix to generate a second content matrix. The total lipid content in the second content matrix is ​​converted into pixel values ​​to generate lipid distribution data.

6. A Raman spectroscopy detection device for implementing the Raman spectroscopy detection method for microalgal lipids as described in claim 1, characterized in that, include: A tunable laser, configured to emit a first laser or a second laser; A glass slide is configured to hold either the first microalgae sample or the second microalgae sample; The confocal optical path assembly is configured to focus either a first laser or a second laser onto a glass slide; The gated detection module is configured to collect Raman scattering signals and generate time-domain signals. The gated detection module has a gated switch. The signal analysis module is configured to reconstruct the first and second Raman spectra in the frequency domain based on the time-domain signal; The data analysis module is configured to determine the power gradient sequence based on the first Raman spectrum and fit a gated gradient function; The image generation module is configured to generate lipid distribution data based on a second Raman spectrum. The system control module is configured to control the tunable laser, the glass slide, and the gated detection module. The system control module controls the laser power of the tunable laser according to the power gradient sequence; the system control module controls the gate width of the gated detection module according to the gated gradient function; and the system control module moves the glass slide according to the focusing position of the first laser or the second laser.

7. A method for dynamic monitoring of microalgae stress treatment, characterized in that, Includes the following steps: Step 100: Divide the microalgae samples to be tested into a first group of microalgae samples and multiple groups of second group of microalgae samples. Use the ultrasonic stress method to treat the second group of microalgae samples at different time periods. The parameters of the ultrasonic stress method are: ultrasonic frequency range of 20kHz-80kHz, ultrasonic power density range of 5W / L-275W / L, and the stress treatment method is pulse mode, with a single pulse duration of 1 to 10 minutes and a pulse interval of 0.5 to 2 hours. Step 200: Detect the first microalgae sample and the second microalgae sample using the Raman spectroscopy detection method for microalgae lipids according to claim 1, and generate multiple sets of lipid distribution data; Step 300: Combine multiple sets of lipid distribution data to construct a lipid distribution data spectrum sequence and generate the total lipid amount for different time periods.

Images