Raman spectrometer and data processing method thereof

Abstract

The invention provides a Raman spectrometer and a data processing method thereof. The Raman spectrometer comprises a laser module, a collection light path, a light splitting assembly and a detection assembly, the detection assembly adopts a universal CMOS area array image sensor; the collection light path transmits Raman scattering light generated by the sample to the light splitting assembly; the light splitting assembly comprises a blazed grating and an imaging lens. The focal length f of the imaging lens is smaller than or equal to 35 mm. The scribed line density D of the blazed grating is greater than or equal to 1000 lines per millimeter; the dispersion capability of the blazed grating is cooperatively matched with the focal length of the imaging lens: within a target Raman displacement range, the scribed line density D is increased along with the reduction of the focal length f, and the dispersion spectrum is ensured to be completely imaged on the target surface of the CMOS sensor. The invention aims to realize miniaturization and low cost of equipment, and simultaneously has high-resolution and high-sensitivity Raman spectrum detection performance.

Description

Technical Field

[0001] This invention relates to the field of spectral analysis technology, and in particular to a Raman spectrometer and its data processing method. Background Technology

[0002] Raman spectroscopy, based on the Raman scattering effect, is a non-destructive analytical technique that can provide the "fingerprint" information of a substance's molecules. It has indispensable and important applications in fields such as chemistry, materials science, biomedicine, pharmacy, and public safety.

[0003] However, the Raman scattering signal is extremely weak (typically only 10 times the intensity of the incident light). - ¹ 0 ~10 -6 This places extremely high demands on the sensitivity and signal-to-noise ratio of the detection system. Therefore, traditional high-performance Raman spectrometers (especially research-grade and mainstream desktop equipment) generally employ specially designed, high-performance, but expensive detectors, primarily charge-coupled device (CCD) linear array sensors or scientific-grade complementary metal-oxide-semiconductor (sCMOS) sensors. These sensors have advantages such as large pixel size (up to tens of micrometers), low noise, and wide dynamic range. To match these sensors and achieve high spectral resolution, traditional optical path designs almost invariably employ long focal length imaging lenses to "magnify" and image the dispersed spectrum on the detector. This "long focal length matching large pixel" design paradigm, while achieving excellent performance (spectral resolution up to 1-2 cm), has limitations. - ¹), but it also leads to a long optical path, complex structure, and difficult debugging of the system. In addition, the core detection module is highly customized and expensive, making the price of the whole machine generally range from hundreds of thousands to millions of RMB, which greatly limits the popularization and application of the technology.

[0004] To reduce the cost of Raman spectrometers, some industry attempts have been made to incorporate lower-cost, general-purpose CMOS area array sensors, which are already widely used in mass-market consumer electronics and machine vision, into Raman spectrometers. For example: Open-source projects (such as Open Raman) utilize a Sony IMX265 general-purpose CMOS sensor, a 50mm focal length lens, and a 1200 lines / mm grating in their performance version. However, this project still hasn't broken free from the traditional "telephoto" design philosophy, achieving a spectral resolution of only about 12cm. - ¹, significantly lower than 5-10cm of mainstream commercial equipment - ¹Level. At the same time, its optical path design is relatively simple, and stray light suppression is insufficient. Public information only shows the detection of strong signal organic solvents, and fails to verify its ability to detect weak standard signals such as high-order peaks of single-crystal silicon, that is, the sensitivity does not reach the mainstream level.

[0005] For example, while Chinese patent application CN207300889U discloses a Raman spectrometer using a CMOS detector and emphasizes its advantages such as low cost and low power consumption, its technical solution also uses a traditional long-focal-length or conventional focal-length imaging design. It fails to disclose how to address the resulting insufficient resolution or how to optimize the design to fully utilize the potential of the CMOS sensor to simultaneously achieve high sensitivity. Furthermore, its description of the technical effects fails to demonstrate that its solution can rival mainstream equipment in terms of resolution and sensitivity.

[0006] In summary, the existing technology has the following main drawbacks: (1) The contradiction between high performance and low cost: dedicated CCD / sCMOS sensors and matching long-focal complex optical paths ensure high performance, but result in extremely high cost; while the solution using general CMOS sensors reduces cost, but the performance (especially resolution and weak signal detection sensitivity) cannot meet the requirements of mainstream applications.

[0007] (2) Mismatch between design concept and device characteristics: The fundamental problem with existing low-cost solutions is that their design thinking is still confined to the traditional "long focal length matching large pixel" paradigm. When simply replacing a dedicated sensor with a general-purpose CMOS sensor with a smaller target surface, smaller pixel size, and higher pixel density, the traditional long focal length lens will over-amplify the spectrum, resulting in too many pixels covering a single spectral feature. This fails to convert the high pixel density advantage of the CMOS sensor into high spectral resolution, instead causing light energy dispersion and a decrease in signal-to-noise ratio. In other words, the existing technology forcibly fits the new device (general-purpose CMOS) into the old architecture (long focal length optical path), failing to conduct a systematic and adaptive redesign of the optical path based on the core characteristics of the new device (small target surface, high pixel density).

[0008] Therefore, there is an urgent need in this field for an innovative technical solution that can break free from the constraints of traditional design paradigms, fundamentally solve the performance bottleneck faced by general-purpose CMOS sensors in Raman spectroscopy applications, and achieve significant cost reduction without sacrificing or even reaching mainstream levels of spectral resolution and detection sensitivity. Summary of the Invention

[0009] The technical problem to be solved by this invention is: In view of the technical problems existing in the prior art, this invention provides a Raman spectrometer and its data processing method, which aims to achieve high resolution and high sensitivity Raman spectroscopy detection performance by co-designing a short focal length lens and a high line density blazed grating, and adapting it to a general-purpose CMOS area array sensor, with a low cost and miniaturized architecture.

[0010] To solve the above-mentioned technical problems, the technical solution proposed by this invention is as follows: A Raman spectrometer includes a laser module, a collection optical path, a dispersive spectrometer, and a detector. The detection component uses a general-purpose CMOS area array image sensor; The collecting optical path is used to collect the Raman scattered light generated by the sample and transmit the Raman scattered light to the dispersive spectrometer. The collecting optical path also includes a slit disposed in front of the dispersive spectrometer. The dispersive beam splitter includes a blazed grating and an imaging lens, wherein the blazed grating is used to disperse the Raman scattered light; The focal length of the imaging lens is f, and f≤35mm; The blazed grating has a line density of D, and satisfies D≥1000 lines / mm; The dispersion capability of the blazed grating is coordinated with the focal length of the imaging lens. For a given target Raman shift detection range, the grating line density D value increases accordingly as the focal length f value of the imaging lens decreases, so that the dispersed Raman spectrum can be completely imaged within the effective target surface of the general-purpose CMOS area array image sensor.

[0011] As a further improvement of the present invention, it also includes a data processing module, which is used to perform region of interest cropping and inter-row cumulative averaging algorithms on the spectral image acquired by the general CMOS area array image sensor: cropping the part of the row pixels with the smallest aberration in the sensor as the region of interest, and accumulating and averaging the spectral data of multiple rows to improve the spectral signal-to-noise ratio.

[0012] As a further improvement of the present invention: the optical path for collecting light includes a first focusing lens, a dichroic mirror, a Rayleigh light filtering module, and a second focusing lens arranged in sequence; the laser emitted by the laser module is reflected by the dichroic mirror and then focused onto the sample by the first focusing lens; the Raman scattered light generated by the sample is collected by the first focusing lens, passes through the dichroic mirror and enters the Rayleigh light filtering module, and after Rayleigh light is filtered out, it is focused onto the subsequent optical path by the second focusing lens.

[0013] As a further improvement of the present invention: the optical path of collection adopts a mirror replacement structure: some or all of the first focusing lens, the second focusing lens, the collimating lens, and the imaging lens are replaced with mirrors. The optical path design of the mirrors is adapted to the transmission and imaging requirements of Raman spectroscopy, so as to shorten the size of the device and reduce stray light interference.

[0014] As a further improvement of the present invention: the optical path for collecting light adopts a Raman probe replacement structure: the front-end optical path consisting of the first focusing lens, dichroic mirror, Rayleigh light filtering module and the second focusing lens is replaced as a Raman probe with fiber optic output. The fiber optic output end of the Raman probe is aligned with the slit to realize remote Raman scattering light detection.

[0015] As a further improvement of the present invention: the height of the slit is less than or equal to 3 mm and the width is less than or equal to 0.1 mm. The slit height is matched with the slit width while ensuring an effective spectral signal, forming a confocal pinhole-like spatial filter structure to suppress stray light interference.

[0016] As a further improvement of the present invention: the slit is disposed between the second focusing lens and the dispersive beam splitter, and the input end of the slit is aligned with the light output end of the second focusing lens.

[0017] As a further improvement of the present invention: the blazing wavelength of the blazing grating is optimized in coordination with the emission wavelength of the laser module, and the blazing wavelength parameter is determined according to the laser emission wavelength.

[0018] As a further improvement of the present invention: the laser module includes at least one laser, the laser including a single emission wavelength laser or two different emission wavelength lasers; when two different emission wavelength lasers are used, the system is configured with an optical path switching mechanism to realize the switching of the front-end optical path to adapt to the Raman scattering light detection requirements of different substances.

[0019] The present invention also provides a method for processing spectral data for a Raman spectrometer, applied to the Raman spectrometer, comprising the following steps: Step S1: Acquire a two-dimensional Raman spectral image acquired by the general CMOS area array image sensor in the Raman spectrometer. The two-dimensional Raman spectral image contains spatial dimension information and dispersion dimension information. Step S2: Based on the dispersion direction of the two-dimensional Raman spectrum image, select the region of interest that is perpendicular to the dispersion direction and has the smallest aberration, and extract multiple rows of effective spectral data within the region of interest; Step S3: Perform inter-row cumulative averaging on the multi-row effective spectral data. Utilize the characteristics of inter-row signal coherence and noise incoherence to suppress spectral noise and generate one-dimensional Raman spectral data with enhanced signal-to-noise ratio to meet the requirements of Raman shift detection.

[0020] Compared with the prior art, the beneficial effects of the present invention are as follows: 1. This invention replaces the dedicated CCD or sCMOS sensor in traditional Raman spectrometers with a widely available and low-cost general-purpose CMOS area array image sensor. It also creatively adopts a collaborative design scheme of short focal length imaging lens (f≤35mm) and high line density blazed grating (D≥1000 lines / mm), breaking the traditional idea of ​​"trading long focal length for precision". This invention achieves a significant simplification of the optical path structure and miniaturization of the overall equipment, enabling the Raman spectrometer to maintain high performance while significantly reducing hardware cost and size.

[0021] 2. This invention achieves full-spectrum coverage and high resolution compatibility in a compact optical path by constructing a precise collaborative mechanism of "blazed grating dispersion capability - imaging lens focal length". Furthermore, by adopting a general-purpose CMOS area array sensor and integrating ROI clipping and inter-row accumulation algorithms, it significantly reduces hardware costs while greatly improving the signal-to-noise ratio and weak signal detection capability. Finally, it forms a highly modular and scalable optical architecture, which significantly enhances the adaptability and functional flexibility of the device to different detection scenarios, thereby systematically solving the technical contradiction of traditional Raman spectrometers in balancing miniaturization, low cost and high performance. Attached Figure Description

[0022] Figure 1 This is a schematic diagram of the overall structure of the Raman spectrometer in an embodiment of the present invention.

[0023] Figure 2 These are the spectra of a series of comparative experiments used in embodiments of the present invention to verify the collaborative design principle of "short focal length lens + high line density grating"; wherein Figure 2 (a) is the Raman spectrum of isopropanol acquired using a general-purpose CMOS area array sensor, a 50mm telephoto lens, and a 1200 lines / mm, 500nm blazed wavelength grating. Arrow 1 points to a Raman characteristic peak, and arrow 2 points to two adjacent Raman characteristic peaks. (b) is the Raman spectrum of isopropanol acquired by replacing the 50mm telephoto lens with a 16mm short focal length lens, while keeping the other configurations unchanged. Arrow 3 points to the Raman characteristic peak corresponding to arrow 1, and arrow 4 points to two adjacent Raman characteristic peaks corresponding to arrow 2. (c) is the Raman spectrum of isopropanol acquired using a 35mm short focal length lens, an 1800 lines / mm high line density grating, and a high-sensitivity CMOS sensor. Arrow 5 points to the Raman characteristic peak corresponding to arrow 1, and arrow 6 points to two adjacent Raman characteristic peaks corresponding to arrow 2.

[0024] Figure 3 This is the measured Raman spectrum of a neon lamp standard light source in an embodiment of the present invention.

[0025] Figure 4 This is the Raman spectrum of sodium citrate dihydrate measured in an embodiment of the present invention.

[0026] Figure 5 The image shows the Raman spectrum of sodium citrate dihydrate downloaded from the Chemicalbook website. Detailed Implementation

[0027] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0028] like Figure 1 As shown, this embodiment provides a Raman spectrometer, including a laser module, a collection optical path, a dispersive spectrometer, and a detector. The detection component uses a general-purpose CMOS area array image sensor; The collecting optical path is used to collect the Raman scattered light generated by the sample and transmit the Raman scattered light to the dispersive beam splitter. The collecting optical path also includes a slit disposed in front of the dispersive beam splitter. The dispersive beam splitter includes a blazed grating and an imaging lens, wherein the blazed grating is used to disperse the Raman scattered light; The focal length of the imaging lens is f, and f≤35mm; The blazed grating has a line density of D, and satisfies D≥1000 lines / mm; The blazed grating's dispersion capability works in conjunction with the focal length of the imaging lens. For a given target Raman shift detection range, the blazed grating's scribe line density D increases as the focal length f of the imaging lens decreases, ensuring that the dispersed Raman spectrum can be completely imaged within the effective target surface of a general-purpose CMOS area array image sensor.

[0029] This embodiment achieves a dispersion coordination mechanism of "short focal length imaging lens + high line density blazed grating" through the coordinated adaptation of optical components, and realizes the core performance target by combining it with a general-purpose CMOS area array image sensor. The Raman spectrometer includes the relative installation positions and connection relationships of the laser module, the light collection path, the dispersive beam splitter, the detector, and the data processing module. All components are fixed and integrated by an optical bracket. The overall device size is reduced by more than 40% compared with traditional devices, effectively solving the problems of large size and poor portability of traditional devices. The detection component utilizes a general-purpose CMOS area array image sensor, which offers advantages in area array imaging. It can acquire two-dimensional spectral images for subsequent noise reduction processing. Compared to dedicated CCD / sCMOS sensors, this significantly reduces hardware costs. Furthermore, the area array imaging characteristics provide a foundation for multi-dimensional data acquisition and noise suppression. The imaging lens is a short-focal-length lens with a focal length ≤35mm. This not only effectively controls the device size but also increases the light energy density per pixel, adapting to the small pixel size and high pixel density characteristics of general-purpose CMOS sensors, avoiding light dispersion and ensuring signal strength. The blazed grating uses a high-dispersion grating with a line density ≥1000 lines / mm. Preferably, in this embodiment, the line density ranges from 1200 lines / mm to 2400 lines / mm, precisely matching the dispersion requirements of the short-focal-length lens, ensuring spectral dispersion accuracy, and supporting high-resolution detection. In addition, the mounting brackets for each component are made of lightweight aluminum alloy, further reducing the overall weight of the device while ensuring structural stability and improving portability. The bracket surface is anodized, effectively preventing corrosion and wear, and extending the device's service life.

[0030] To achieve systematic compatibility between short-focal-length imaging lenses and high-line-density blazed gratings, this embodiment is based on the target Raman shift detection range (typically set to 500cm). - ¹Up to 3500cm - ¹ Based on the matching rules between the characteristic peak detection of most organic and inorganic substances and the effective target surface of a general CMOS area array image sensor, a core collaborative design logic was constructed. The core of this collaborative logic lies in ensuring that the complete Raman spectrum after dispersion can be accurately projected and contained within the limited target surface of the sensor by precisely matching the dispersive capability of the blazed grating with the imaging field of view of the short-focal-length imaging lens. To verify this design logic, a series of comparative experiments were conducted in this embodiment, with the specific configuration as follows: like Figure 2As shown in (a), the Raman spectrum of isopropanol was acquired using a general-purpose CMOS area array sensor (model IMX265), a 50mm telephoto lens, and a 1200 lines / mm, 500nm blazed wavelength grating. The experimental results show that the spectral feature line indicated by arrow 1 has a width of 21 pixels. At arrow 2, two adjacent Raman feature peaks merge into one due to insufficient spectral line resolution and cannot be clearly distinguished. This indicates that although a telephoto lens combined with a conventional grating can magnify the spectral image on a general-purpose CMOS area array sensor, it suffers from low resolution and overlapping spectral lines.

[0031] like Figure 2 As shown in (b), in Figure 2 Based on (a), only the 50mm telephoto lens was replaced with a 16mm short telephoto lens, while all other configurations remained unchanged, to obtain the isopropanol Raman spectrum. Experimental results show that the spectral line width at arrow 3 decreased to 7 pixels, demonstrating a significant line refinement effect, indicating that shortening the lens focal length can improve spectral resolution. Arrow 4 points to the overall imaging spot of the Raman spectrum after dispersion; this spot area is significantly smaller than... Figure 2 (a) indicates that the spectral image formed by a short focal length lens becomes smaller, resulting in a wider spectral range acquired by the same sensor, while simultaneously reducing resolution. Therefore, a higher line density grating is needed to broaden the spectral image, so that the spectral range acquired by the same sensor remains unchanged, only improving spectral resolution.

[0032] like Figure 2 As shown in (c), the configuration was further optimized, employing a 35mm short focal length lens, a 1800 lines / mm high-density grating, and a high-sensitivity CMOS sensor to acquire the same isopropanol sample. Arrow 5 points to the same Raman characteristic peak corresponding to arrows 1 and 3, and the spectral linewidth was further optimized to approximately 5 pixels, resulting in a higher resolution. Figure 2 (b) Further enhancement. Arrow 6 points to the two sets of adjacent Raman characteristic peaks corresponding to arrow 2. After compensation by a high-density grating, the two originally fused spectral lines are clearly separated, and the spectral coverage is complete. This proves that increasing the grating density can compensate for the insufficient field of view of a short focal length lens, and further verifies that the synergistic design of "short focal length lens + high-density grating" can simultaneously achieve both "high resolution" and "full spectral coverage".

[0033] Based on the above experimental results, this embodiment further summarizes the systematic design principle of "coordinated matching between the focal length f of the short focal length lens and the blazed grating line density D". The relevant dispersion coordination mechanism, optical path transmission process, and the relative positions of the imaging lens, blazed grating, and general-purpose CMOS area array image sensor, as well as the dispersion and transmission path of Raman scattered light, are all described. The effective target surface of the sensor is marked with a dashed box. When the focal length f of the imaging lens decreases, its imaging field of view shrinks. If the grating line density D remains unchanged, the dispersed Raman spectrum will exceed the effective target surface of the general-purpose CMOS sensor, resulting in a loss of spectral information. Therefore, it is necessary to increase the D value to improve the grating dispersion capability and ensure that the complete spectrum is imaged within the target surface. This coordinated design can achieve a balance between "short focal length miniaturization" and "high resolution, full spectral coverage", breaking the design limitations of the traditional "long focal length for precision" approach. For example, when f=35mm, a grating with D=1000 lines / mm to 1200 lines / mm is sufficient to cover the target area; when f=16mm (lower limit of short focal length), a grating with D=2000 lines / mm to 2400 lines / mm is required to ensure complete spectral imaging. This synergistic relationship has been verified through the above experiments and optical simulations, achieving a spectral integrity of over 99% and a spectral resolution of 4cm. - ¹, meeting the detection accuracy requirements of mainstream application scenarios.

[0034] Simulation verification uses Zemax optical design software to build an optical model consistent with the actual equipment. Parameters of different focal length lenses and corresponding gratings are input to simulate spectral dispersion and target projection effects. Experimental verification uses standard samples (such as single-crystal silicon and polystyrene) for actual measurement. The integrity and resolution of spectral acquisition under different parameter combinations are compared. Verification data are recorded in the accompanying experimental report and can serve as a basis for performance assurance.

[0035] To optimize the signal output quality of a general-purpose CMOS area array image sensor and improve the detection accuracy of Raman spectroscopy, this embodiment also includes a data processing module. The data processing module is used to perform region of interest cropping and inter-row cumulative averaging algorithms on the spectral images acquired by the general-purpose CMOS area array image sensor: the part of the row pixels with the smallest aberration in the sensor is cropped as the region of interest, and the spectral data of multiple rows are cumulatively averaged to improve the spectral signal-to-noise ratio.

[0036] The data processing module, as the core optimization component of the Raman spectrometer, is designed based on the imaging characteristics of a general-purpose CMOS area array sensor. It aims to compensate for the slightly higher noise level of general-purpose CMOS sensors compared to dedicated sensors, significantly improving the spectral signal-to-noise ratio and ensuring weak signal detection capabilities. In the two-dimensional spectral images acquired by the sensor, the aberrations of pixels in different rows vary, with larger aberrations in edge rows, introducing additional noise. Therefore, region of interest (ROI) cropping is necessary to remove edge noise pixels and reduce invalid data interference. The core criterion for ROI cropping is "selecting the rows with the smallest aberrations." For example, when using the IMX664 sensor (with 1520 rows of pixels), the spectral imaging typically occupies about 500 rows of pixels. Experiments have determined that the middle 100 rows of pixels represent the region with the smallest aberrations. The data processing module automatically extracts this region as the effective data source, which can improve the signal purity of subsequent processing by more than 30%. Aberration detection uses professional image quality analysis equipment to quantify row pixel aberrations in standard resolution images acquired by the sensor, determine the aberration value distribution curves of different rows of pixels, and delineate the range of the minimum aberration area to ensure the scientific nature and accuracy of the cropping standard.

[0037] The principle of the inter-row cumulative averaging algorithm is that "inter-row signals are coherent, while inter-row noise is incoherent." When N rows of effective spectral data are cumulatively averaged, the signal-to-noise ratio can theoretically be improved. This represents a 10-fold increase in signal-to-noise ratio (e.g., accumulating 100 rows of data). After processing with the above ROI cropping and inter-row averaging algorithm, the system achieves a higher signal-to-noise ratio at 1440cm. - A weak third-order Raman peak of monocrystalline silicon can be clearly detected at point ¹, with a measured signal-to-noise ratio greater than 10:1, fully meeting the requirements for high-precision detection. To achieve real-time processing of the algorithm, the data processing module can be implemented using an FPGA chip or a microcontroller: FPGA chips (such as the Xilinx Artix-7 series) are preferred, as they have abundant logic resources, support parallel processing, and can control the spectral processing delay to within 10ms; if a microcontroller solution is used, the algorithm code needs to be deeply optimized to ensure the core noise reduction function under limited resources.

[0038] To suppress stray light interference and improve the purity of Raman scattered light signals, this embodiment also includes a slit disposed in the light collection path and located before the dispersive beam splitter; the height of the slit is less than or equal to 3 mm and the width is less than or equal to 0.1 mm. The slit height is matched with the slit width while ensuring an effective spectral signal, forming a confocal pinhole-like spatial filter structure to suppress stray light interference.

[0039] This slit is a confocal pinhole-like spatial filter structure that filters stray light deviating from the main optical path by limiting the beam aperture, allowing only the Raman-scattered main signal light to enter the dispersive beam splitter. Compared to a slit-less solution, stray light suppression efficiency can be improved by more than 6 times. The width and height parameters of the slit are matched to meet the balance requirements of "stray light suppression" and "signal throughput guarantee," as detailed below: Width primarily affects spectral resolution, luminous flux, and signal-to-noise ratio (SNR). A narrower width results in stronger spatial filtering, better stray light suppression, and a corresponding increase in spectral resolution, but at the expense of lower luminous flux, potentially leading to a decrease in SNR. Conversely, an excessively wide width results in insufficient stray light suppression, affecting signal purity. The preferred range is between 0.01 mm and 0.1 mm, with 0.05 mm being the optimal value, balancing stray light suppression and signal strength.

[0040] Height primarily affects luminous flux and signal integration efficiency, thus impacting the signal-to-noise ratio (SNR). Too low a height limits the effective number of pixel rows covered by the beam, reducing the signal integration space and lowering the SNR; too high a height introduces more stray light, also leading to a decrease in SNR. The optimal height range is between 0.1mm and 3mm, with 1mm being the preferred value, ensuring sufficient luminous flux while controlling stray light entry and optimizing the SNR.

[0041] The slit, through the coordinated design of its width and height, forms a confocal pinhole-like filter structure, which effectively suppresses stray light while ensuring that the Raman signal has sufficient light flux and signal-to-noise ratio to meet the requirements of high-sensitivity Raman spectroscopy detection.

[0042] The slit is processed using precision photolithography to ensure that the dimensional accuracy of the slit width and height is controlled within ±0.005mm, and the slit edge is smooth and burr-free, avoiding beam deviation or additional stray light caused by processing errors.

[0043] The slit is made of stainless steel or other metals with a blackened surface treatment to reduce secondary stray light from beam reflection. Compared to slits made of ordinary metal without this treatment, secondary stray light is reduced by more than 40%. The slit must be precisely aligned with the output end of the light-collecting path to ensure the main signal light is incident perpendicularly. Precise installation is achieved using an optical positioning fixture with a positioning error not exceeding 0.1mm, effectively ensuring spectral imaging accuracy and preventing resolution degradation due to installation deviations. The optical positioning fixture is equipped with a fine-tuning knob and a laser alignment tool. During installation, the laser alignment tool is used to calibrate the coaxiality between the slit center and the optical path center, and then the fine-tuning knob is used to adjust the horizontal and vertical positions of the slit to ensure the alignment accuracy meets requirements.

[0044] In this embodiment, the optical path includes a first focusing lens, a dichroic mirror, a Rayleigh light filtering module, and a second focusing lens arranged in sequence. The laser emitted by the laser module is reflected by the dichroic mirror and then focused onto the sample by the first focusing lens. The Raman scattered light generated by the sample is collected by the first focusing lens, passes through the dichroic mirror and enters the Rayleigh light filtering module. After Rayleigh light is filtered out, it is focused onto the subsequent optical path by the second focusing lens.

[0045] The optical path for collection employs a coaxial backscattering design, which simplifies the optical path structure, reduces the size of the device, and improves the collection efficiency of Raman light. Compared to a non-coaxial design, the collection efficiency is improved by more than 30%, and the optical path integration is higher, facilitating device miniaturization. The laser incident light and Raman scattered light collection share the same first focusing lens, which further simplifies the optical path structure, reduces the size of the device, and lowers hardware costs. The first focusing lens uses a convex lens with a numerical aperture (NA) ≥ 0.25, preferably a lens with NA = 0.3, which can focus the laser into a spot with a diameter ≤ 10 μm, improving laser excitation efficiency, enhancing the Raman scattering signal intensity of the sample, and efficiently collecting Raman scattered light. The first focusing lens adopts an achromatic lens group design, which can effectively correct chromatic aberration and spherical aberration, ensuring the roundness and sharpness of the laser focused spot, and avoiding the decrease in excitation efficiency or sample damage caused by spot distortion.

[0046] A dichroic mirror in the optical path is a wavelength-selective optical element based on an interference thin film. Its characteristics include high reflectivity for laser wavelengths and high transmittance for Raman scattered light wavelengths. By precisely matching the dichroic mirror parameters to the selected laser wavelength—for example, when the laser wavelength is 532nm, selecting a dichroic mirror with a reflection center wavelength of 532nm and a transmission band of 550nm~800nm—can achieve highly efficient separation of the excitation light and the Raman signal light, with a separation efficiency exceeding 99%. This effectively prevents strong laser light from entering the detection optical path and interfering with Raman signal detection.

[0047] The Rayleigh scattering module is a core component for suppressing background noise. It filters out Rayleigh scattered light, which has the same wavelength as the laser but an intensity several orders of magnitude higher than the Raman signal. This module uses a combination of long-pass filters, with the cutoff wavelength of the filters slightly larger than the laser wavelength. For example, when the laser wavelength is 532nm, a long-pass filter with a cutoff wavelength of 535nm can be used to completely filter out Rayleigh light while ensuring that the Raman signal (above 550nm) passes through completely.

[0048] The Raman signal light, after being filtered and purified as described above, is first focused onto the slit position by the second focusing lens. A collimating lens is precisely positioned behind the slit, its function being to collimate the diverging light emanating from the slit into a parallel beam. This parallel beam is a necessary prerequisite for the subsequent efficient dispersion of the blazed grating, maximizing the utilization of input light energy. The collimating lens is preferably an achromatic lens with a focal length of 50mm, and its focal length is designed in conjunction with the parameters of the second focusing lens and the short-focal-length imaging lens to ensure the generation of a high-quality spectral image within a compact space.

[0049] The Rayleigh light filtering module's long-pass filter assembly adopts a dual-layer filter structure. The first layer is a preliminary filter, and the second layer is a high-precision cutoff filter, which can effectively improve the thoroughness of Rayleigh light filtering, while reducing the absorption loss of Raman signals by the filter and ensuring signal transmission efficiency. The dichroic mirror adopts a dielectric film coating process to improve the selectivity of reflection and transmission and reduce stray light reflection.

[0050] To adapt to miniaturized detection scenarios, this embodiment provides a first alternative solution for the optical path collection. This solution optimizes device size and stray light performance by replacing lens assemblies with reflectors. Specifically, the optical path collection adopts a reflector-replacement structure: some or all of the first focusing lens, second focusing lens, collimating lens, and imaging lens are replaced with reflectors. The optical path design of the reflectors is adapted to the transmission and imaging requirements of Raman spectroscopy, thereby reducing device size and stray light interference. When replacing the collimating lens, the reflector used must have collimation functionality; for example, an off-axis parabolic mirror can be used.

[0051] This alternative is based on the following design considerations: lens assemblies have inherent chromatic aberration issues, and the light path is linear, occupying a large space; reflectors do not have lens chromatic aberration, and the length of the device can be shortened through a folded light path design, while reducing stray light interference from beam reflection, making them suitable for miniaturized, low-interference detection scenarios. Specifically, the replacement method for the reflector can be chosen according to actual needs: only the first and second focusing lenses can be replaced, retaining the imaging lens; alternatively, all three lenses can be replaced with reflectors, with the imaging lens replaced by a reflective short-focal-length lens (the focal length still needs to be ≤35mm).

[0052] The reflector is selected from either aluminum or silver film reflectors, with silver film reflectors being preferred due to their reflectivity ≥95%. Compared to aluminum film reflectors, this represents a 10% or more improvement in reflection efficiency, ensuring no loss of Raman signal strength. The reflector is fixed by an adjustable bracket with an angle adjustment range of 0° to 90°. The angle can be adjusted according to the optical path transmission requirements, achieving precise beam transmission, adapting to different equipment structural designs, improving equipment assembly and debugging efficiency, and reducing production difficulty. Experimental data shows that using this alternative structure can further reduce the equipment size by 20% to 40%, and further reduce stray light interference by 15%, fully meeting the requirements for portable testing. Performance comparison data can be verified in conjunction with structural parameters. The reflector surface is treated with a multi-layer protective film to prevent oxidation and discoloration, extending the reflectivity stabilization period. The adjustable bracket uses a precision gear transmission structure, with an adjustment accuracy of up to 0.1°, ensuring the accuracy and stability of angle adjustment and facilitating optical path calibration.

[0053] This embodiment provides a second alternative to the optical path collection, replacing the traditional front-end optical path with a Raman probe. Specifically, the optical path collection adopts a Raman probe replacement structure: the entire front-end optical path consisting of the first focusing lens, dichroic mirror, Rayleigh filter module, and second focusing lens is replaced with a Raman probe with fiber optic output. The fiber optic output end of the Raman probe is aligned with the slit to achieve remote Raman scattering light detection.

[0054] In this embodiment, the Raman probe is an integrated front-end component, internally integrating functional components such as a focusing lens, a dichroic mirror, and a Rayleigh light filtering module. Its function is completely consistent with traditional front-end optical paths. Its core advantage lies in transmitting Raman signals via optical fiber, achieving separation between the front-end optical path and the host device, expanding the detection scenarios of the equipment, and enabling remote detection of samples that are difficult to access, while maintaining the same detection accuracy as traditional optical paths. The overall structure of the Raman probe's alternative collection optical path is shown, indicating the connection relationships between the Raman probe, transmission fiber, slit, and dispersive beam splitter, as well as the sample placement position and signal transmission path during remote detection. The Raman probe uses silica fiber with a core diameter of 50μm~200μm and a numerical aperture of 0.22~0.25. Compared to ordinary optical fiber, signal transmission loss is reduced by more than 25%, and the transmission distance can reach 5~10 meters, meeting the needs of most remote detection scenarios. The quartz optical fiber adopts a low-loss coating design to reduce signal attenuation during transmission, while also having good bending resistance, making it easy to install and operate on site. The Raman probe's housing adopts a waterproof and dustproof design with an IP65 protection rating, which can be adapted to complex field testing environments and prevent dust and moisture from entering the interior and affecting component performance.

[0055] The Raman probe features a protective window made of sapphire, offering high light transmittance, wear resistance, and high-temperature resistance. This prevents sample contamination or damage to internal components, extending the probe's lifespan and environmental adaptability while reducing maintenance costs. The key to Raman probe installation is ensuring precise alignment between the fiber optic output and the slit. Alignment is achieved using a fiber optic mount and positioning fixture, with a positioning error not exceeding 0.1mm. This ensures the signal transmitted remotely accurately enters the subsequent optical path, avoiding spectral distortion caused by signal offset and guaranteeing consistent accuracy between remote and near-field detection. This alternative solution requires no modification to the backend dispersive beam splitter, detector, or data processing module, offering strong compatibility. It allows for flexible switching between the traditional and Raman probe optical paths depending on the detection scenario. A quick-connect connector is preferred for rapid switching between the two optical paths, with a switching time of ≤1 minute. No professional personnel are required for setup, lowering the barrier to entry. To ensure the slit effectively suppresses stray light, its installation position must be precisely defined based on the different structures of the collection optical path.

[0056] To ensure that the slit fully exerts its stray light suppression effect, its installation position needs to be precisely defined according to the different structures of the light collection path. In this embodiment, the slit is set between the second focusing lens and the dispersive beam splitter, and the input end of the slit is aligned with the light output end of the second focusing lens.

[0057] Specifically, the core installation principle of the slit is that it should be located "between the light output end of the collection optical path and the light input end of the dispersive beam splitter," ensuring that all Raman signal light filtered by the collection optical path must pass through the slit into the dispersive beam splitter to maximize stray light suppression. The installation details of the slit are as follows, depending on the different collection optical path structures: When using a basic optical path structure (composed of a first focusing lens, a dichroic mirror, a Rayleigh light filtering module, and a second focusing lens in sequence), the light output end of the second focusing lens is the final output end of the collecting optical path. The slit is positioned at the focal point of the second focusing lens, and the input end is precisely aligned with the light output end to ensure that the beam enters the slit perpendicularly. The slit is fixed by an optical bracket, and its position and angle can be finely adjusted with an alignment accuracy of ≤0.1mm. When using a mirror-replacement optical path structure (replacing part or all of the lens components with a mirror), if the second focusing lens has been replaced, the slit is positioned between the light output end of the mirror optical path and the dispersive beam splitter, and the input end is aligned with the light output end of the mirror. When using a Raman probe-replacement optical path structure (integrating the front-end optical path function with a Raman probe with fiber optic output), the fiber optic output end of the Raman probe is the light output end of the collecting optical path. The input end of the slit is aligned with the fiber optic output end, with a distance of 2mm~5mm from the fiber optic output end to avoid light throughput loss caused by signal light divergence. All three slit mounting configurations ensure effective stray light suppression and signal transmission efficiency, adapting to the detection needs of various scenarios. For slit mounting with different optical path structures, dedicated positioning templates are provided. These templates are marked with the standard installation position and alignment baseline of the slit, enabling rapid initial slit positioning and improving assembly efficiency. Fine-tuning and calibration further ensure alignment accuracy.

[0058] To improve the diffraction efficiency of the blazed grating and ensure that the Raman signal intensity meets the detection requirements, the matching relationship between its blazed wavelength and the laser emission wavelength needs to be optimized. In this embodiment, the blazed wavelength of the blazed grating and the emission wavelength of the laser module are optimized in tandem, and the blazed wavelength parameter is determined based on the laser emission wavelength.

[0059] Specifically, the blaze wavelength of the blazed grating is used as a core parameter, defined as the wavelength at which the grating's diffraction efficiency is highest. This is to ensure the Raman scattered light (target range 500cm) is... - ¹~3500cm - ¹) To achieve the highest diffraction efficiency during the detection process, the blaze wavelength and the laser emission wavelength must be matched. The core design principle is: "The difference between the blaze wavelength and the laser emission wavelength should be controlled within a preset range" to ensure that the main wavelength of the target Raman signal falls within the high diffraction efficiency range of the grating.

[0060] Based on the above matching principle, the specific parameter correspondence can be further clarified: when the emission wavelength of the laser module is 532nm (suitable for inorganic matter detection), the blaze wavelength of the blaze grating is selected as 500nm±50nm, preferably 500nm. This matching relationship allows for Raman signals in the range of 550nm~800nm ​​(corresponding to 500cm²) to be within the range of 550nm~800nm. - ¹~3500cm -¹Raman shift) diffraction efficiency ≥80%; when the laser emission wavelength is 785nm (adapted for organic matter detection, reducing fluorescence interference), the blaze wavelength is selected as 750nm±50nm, preferably 750nm, ensuring Raman signal diffraction efficiency ≥80% in the 780nm~1000nm range. Furthermore, the blazed grating size is preferably 25mm×25mm×6mm, adapting to the overall structural design of the equipment, and the incident angle is close to the Littorhal angle, further improving diffraction efficiency. The incident angle of the blazed grating is precisely positioned using an optical support, with the Littorhal angle deviation controlled within ±0.5° to ensure optimal diffraction efficiency; the grating surface is treated with an anti-reflective coating to reduce stray light from surface reflection and improve signal contrast.

[0061] To expand the application range of Raman spectrometers and adapt to the detection needs of different substances, this embodiment further provides a multi-wavelength configuration scheme for the laser module and a corresponding optical path switching mechanism. Specifically, the laser module includes at least one laser, which may be a single-emission wavelength laser or two different emission wavelength lasers. When two different emission wavelength lasers are used, the system is configured with an optical path switching mechanism to switch the front-end optical path to adapt to the Raman scattering light detection needs of different substances.

[0062] Based on the above configuration, the key parameters of the laser module are set as follows: the emission wavelength range of the laser module is 500nm~830nm (the mainstream laser wavelength range for Raman detection), and the linewidth of the laser needs to be less than 0.2nm to ensure the stability of laser excitation, avoid spectral peak broadening caused by excessive linewidth, and ensure spectral resolution. Compared with lasers with larger linewidths, the spectral peak clarity is improved by more than 40%. The configuration of the laser module can be selected according to application requirements: if only a single type of substance is to be detected, a single emission wavelength laser is selected, preferably a 532nm laser, which can reduce equipment costs and adapt to specific scenario requirements; if both organic and inorganic substances need to be detected, a dual-wavelength combination laser of 532nm and 785nm is preferred, which can cover the detection needs of most substances. The laser uses a semiconductor laser, which features small size, low power consumption, and high stability. The laser output power can be continuously adjusted within the range of 0~100mW to adapt to the detection needs of samples with different fluorescence intensities. The laser module is equipped with a temperature control unit to keep the laser operating temperature at 25±0.5℃, avoiding wavelength drift caused by temperature fluctuations and ensuring detection stability.

[0063] The optical path switching mechanism is used to switch the front-end optical path (mainly the dichroic mirror and Rayleigh filter module, which must be matched with the laser wavelength) to different laser wavelengths. It employs either an electromagnetic switching valve or a mechanical switching structure, with an electromagnetic switching valve being preferred due to its fast switching response (switching time ≤ 0.5 seconds) and ability to achieve automated control. Compared to mechanical switching structures, it is more convenient to operate, offers higher switching accuracy, and avoids optical path interference caused by switching deviations. The core of the switching mechanism is to ensure the accuracy of the optical path after switching, preventing a decrease in excitation efficiency caused by laser incident angle deviations. Precise positioning after optical path switching is achieved through positioning pins, with a positioning error ≤ 0.05mm, ensuring consistent detection accuracy after switching and avoiding performance fluctuations caused by switching. In addition, the laser module is equipped with a laser power adjustment component, with a power adjustment range of 1mW~100mW. The laser power can be adjusted according to sample characteristics to avoid damaging samples (such as biological samples) with excessive power, while ensuring the excitation intensity of weak signal samples (such as trace substances), improving the sample adaptability of the equipment. The electromagnetic switching valve adopts a contactless design, has a long service life, and provides smooth and vibration-free switching, avoiding optical path deviation during switching. The positioning pin is made of hard alloy, which has good wear resistance and ensures stable positioning accuracy after long-term use. The laser power adjustment component achieves continuous power adjustment through a high-precision potentiometer with an adjustment accuracy of 0.1mW, which can accurately match the excitation requirements of different samples.

[0064] To fully verify the core performance of the Raman spectrometer in this embodiment, standard tests were performed on its spectral resolution and system sensitivity. For example... Figure 3 As shown, a neon lamp was used as the standard light source to test this system. The test equipment parameters were: IMX664 sensor, 25mm focal length, F0.8 imaging lens, 1800 lines / mm, 500nm blazed grating, and 10μm slit width. After ROI cropping and inter-row averaging in the data processing module, a one-dimensional neon lamp spectrum was obtained. Analysis of the characteristic peaks in the figure showed that the full width at half maximum (FWHM) of the strongest peak (585.25nm) was 3.7cm. - ¹ indicates that the system's spectral resolution reaches 3~4 cm⁻¹. - ¹ It meets the performance requirements of mainstream desktop Raman spectrometers. The corresponding test data are shown in Table 1: Table 1 Test Data Table

[0065] Table 1 presents the quantitative test results of the system in this embodiment on the neon lamp spectrum. Data analysis shows that the system performs well in the range of 500–3500 cm⁻¹. - ¹Characteristic peaks can be effectively resolved across the entire spectrum, with the full width at half maximum (FWHM) of each peak ranging from 1.83 to 3.70 cm⁻¹. -¹ Between these values, the strongest peak (585.25 nm) has a free wave height (FWHM) of 3.70 cm⁻¹. - ¹ This demonstrates that the system's spectral resolution reaches 3–4 cm⁻¹. - ¹ The design goal is to reach the high level of mainstream desktop Raman spectrometers. At the same time, the signal-to-noise ratio (characterized by relative intensity) of each characteristic peak is stable, indicating that the system has good signal consistency and detection sensitivity across the entire spectrum, meeting the requirements of high-precision spectral analysis.

[0066] In this embodiment, a high-purity single-crystal silicon wafer (crystal orientation 100) was used for sensitivity testing. The test configuration was the same as that for spectral resolution verification, using a 532nm narrow-linewidth solid-state laser (50mW power), a slit width of 100μm, a slit height of 1mm, an integration time of 267 seconds, and multiple averaging. After data processing, the system achieved a sensitivity of 1440cm. - The third-order weak Raman peak of single-crystal silicon is clearly visible at position ¹, with a measured signal-to-noise ratio greater than 10:1, demonstrating that this system possesses excellent weak signal detection capabilities and high sensitivity. To further verify the reliability of the spectral data from this system, it is compared with spectra from publicly available databases. Figure 4 , Figure 5 As shown, the Raman spectrum of sodium citrate dihydrate measured by this system (excited by a 532nm laser, with a power of 50mW and an integration time of 5s) is highly consistent with the corresponding standard spectrum downloaded from the Chemicalbook website in terms of characteristic peak position, relative intensity, and spectral line shape. This further verifies that the Raman spectrometer in this embodiment has reached a professional level in terms of resolution and signal-to-noise ratio.

[0067] This embodiment also provides a spectral data processing method for a Raman spectrometer. This method is a proprietary data processing method designed for the Raman spectrometer and cannot be implemented independently of the hardware structure of the Raman spectrometer. The implementation details and hardware adaptation of each step are as follows: Step S1: Acquire a two-dimensional Raman spectral image from a general-purpose CMOS area array image sensor in a Raman spectrometer. The two-dimensional Raman spectral image contains spatial dimension information and dispersion dimension information.

[0068] Two-dimensional Raman spectral images acquired by a general-purpose CMOS area array image sensor have a spatial dimension corresponding to the spatial information of the sample, and a dispersion dimension corresponding to the wavelength (wavenumber) information of the Raman spectrum. Taking the IMX415 sensor as an example, its acquired two-dimensional image resolution is 3840×2160 pixels, where the horizontal direction is the dispersion direction (corresponding to a Raman shift of 500 cm⁻¹). - ¹~3500cm -¹), with the vertical direction being the spatial direction, image data is transmitted to the data processing module via the sensor's MIPI-CSI interface at a transmission rate ≥1.5Gbps, ensuring real-time transmission and processing of spectral images. The sensor is equipped with a gain adjustment function, dynamically adjusting the analog gain according to signal strength. When detecting weak Raman signals, the gain is increased to enhance the electrical signal; when detecting strong signals or in the presence of fluorescence background, the gain is decreased to avoid signal saturation, ensuring the linearity and integrity of the spectral data.

[0069] In the static spectral acquisition application scenario of the Raman spectrometer in this embodiment, there is no relative movement between the sensor, the sample, and the spectrum during exposure. Therefore, the rolling shutter mode has significant advantages. Compared to global shutter sensors, rolling shutter CMOS sensors have advantages such as simple pixel structure, high fill factor, generally lower readout noise, and higher cost. These characteristics are particularly suitable for detection requirements where Raman signals are weak and system sensitivity and cost control are critical. In this static imaging scenario, the inherent line-by-line exposure characteristic of the rolling shutter does not cause image distortion, and its advantages of high sensitivity and low cost are fully utilized, making it the preferred solution of this invention.

[0070] Step S2: Based on the dispersion direction of the two-dimensional Raman spectrum image, select the region of interest that is perpendicular to the dispersion direction and has the smallest aberration, and extract multiple rows of effective spectral data within the region of interest.

[0071] The determination of dispersion direction is based on the structural design of the dispersive beam splitter (the dispersion direction of the blazed grating is horizontal). Therefore, rows of pixels perpendicular to the horizontal direction (i.e., the vertical direction) are selected as the processing objects. The selection of the region of interest is achieved through a preset threshold, with a preset aberration threshold of 0.1 pixels. The data processing module automatically identifies row pixel regions with aberrations ≤ 0.1 pixels in the vertical direction and extracts multiple rows of valid data within this region. For example, rows 500 to 600 (a total of 100 rows) in the vertical direction are extracted as valid data. The data processing module has a built-in aberration judgment algorithm that can perform row pixel aberration analysis on the acquired 2D image in real time, automatically delineating the region of interest without manual intervention, thus improving the automation level of detection. For different types of CMOS sensors, the aberration threshold and the range of valid row extraction can be adjusted through the parameter configuration interface to improve the adaptability of the method.

[0072] Step S3: Perform inter-row cumulative averaging on the multiple rows of valid spectral data. Utilize the characteristics of inter-row signal coherence and noise incoherence to suppress spectral noise and generate one-dimensional Raman spectral data with enhanced signal-to-noise ratio to meet the requirements of Raman shift detection.

[0073] The specific process of inter-row cumulative averaging is as follows: for the 100 rows of valid data extracted, the values ​​of pixels at the same wavelength / wavenumber position are accumulated column by column, and then divided by the number of rows (100) to obtain the average pixel value, finally generating 3840×1 one-dimensional Raman spectral data (only retaining the dispersion dimension information); document experimental data verification shows that the spectral data processed by this method has a noise intensity reduction of more than 90% and a spectral resolution of up to 4cm. - ¹, can clearly distinguish adjacent Raman peaks (e.g., 5cm apart) - ¹Two Raman peaks), perfectly suited to the high-precision requirements of Raman displacement detection.

[0074] The data processing steps are as follows: First, a two-dimensional spectral image (labeled with spatial and dispersion dimensions) is acquired from a general-purpose CMOS area array image sensor. This is followed by the extraction of the region of interest (selecting the row of pixels with the smallest aberration). Then, a one-dimensional Raman spectral data output is generated after inter-row cumulative averaging. The image includes comparative annotations of the spectra before and after processing (including noise suppression and baseline correction effects), visually demonstrating the improvement in spectral signal-to-noise ratio and accuracy achieved through data processing. The processed one-dimensional Raman spectral data can be transmitted to a host computer via a data interface (such as USB or Ethernet). The host computer software can perform functions such as spectral peak identification, wavenumber calibration, data storage and export, and supports real-time spectral display for easy observation of the detection results. A baseline correction algorithm is also incorporated into the data processing to remove the influence of baseline drift on the detection results, further improving the accuracy of the spectral data.

[0075] The above description is merely a preferred embodiment of the present invention and is not intended to limit the invention in any way. Although the present invention has been disclosed above with reference to preferred embodiments, it is not intended to limit the invention. Therefore, any simple modifications, equivalent changes, and alterations made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention should fall within the protection scope of the present invention.

Claims

1. A Raman spectrometer, characterized in that, Includes a laser module, a light-collecting path, a dispersive beam splitter, and a detector. The detection component uses a general-purpose CMOS area array image sensor; The collecting optical path is used to collect the Raman scattered light generated by the sample and transmit the Raman scattered light to the dispersive spectrometer. The collecting optical path also includes a slit disposed in front of the dispersive spectrometer. The dispersive beam splitter includes a blazed grating and an imaging lens, wherein the blazed grating is used to disperse the Raman scattered light; The focal length of the imaging lens is f, and f≤35mm; The blazed grating has a line density of D, and satisfies D≥1000 lines / mm; The dispersion capability of the blazed grating is coordinated with the focal length of the imaging lens. For a given target Raman shift detection range, the grating's scribe line density D increases accordingly as the focal length f of the imaging lens decreases, so that the dispersed Raman spectrum can be completely imaged within the effective target surface of the general-purpose CMOS area array image sensor.

2. The Raman spectrometer according to claim 1, characterized in that, It also includes a data processing module, which is used to perform region of interest cropping and inter-row cumulative averaging algorithms on the spectral images acquired by the general CMOS area array image sensor: cropping the row pixels with the smallest aberration in the sensor as the region of interest, and accumulating and averaging multiple rows of spectral data to improve the spectral signal-to-noise ratio.

3. The Raman spectrometer according to claim 1, characterized in that, The optical path includes a first focusing lens, a dichroic mirror, a Rayleigh light filtering module, and a second focusing lens arranged in sequence. The laser emitted by the laser module is reflected by the dichroic mirror and then focused onto the sample by the first focusing lens. The Raman scattered light generated by the sample is collected by the first focusing lens, passes through the dichroic mirror, enters the Rayleigh light filtering module, filters out the Rayleigh light, and is then focused onto the subsequent optical path by the second focusing lens.

4. The Raman spectrometer according to claim 3, characterized in that, The optical path for collecting light adopts a mirror replacement structure: some or all of the first focusing lens, the second focusing lens, the collimating lens, and the imaging lens are replaced with mirrors. The optical path design of the mirrors is adapted to the transmission and imaging requirements of Raman spectroscopy, so as to shorten the size of the device and reduce stray light interference.

5. The Raman spectrometer according to claim 3, characterized in that, The optical path for collecting light adopts a Raman probe replacement structure: the front-end optical path consisting of the first focusing lens, dichroic mirror, Rayleigh light filtering module, and second focusing lens is replaced with a Raman probe with fiber optic output. The fiber optic output end of the Raman probe is aligned with the slit to achieve remote Raman scattering light detection.

6. The Raman spectrometer according to claim 1, characterized in that, The slit has a height of less than or equal to 3 mm and a width of less than or equal to 0.1 mm. The slit height is matched with the slit width to ensure an effective spectral signal, forming a confocal pinhole-like spatial filter structure to suppress stray light interference.

7. The Raman spectrometer according to claim 6, characterized in that, The slit is positioned between the second focusing lens and the dispersive beam splitter, with the input end of the slit aligned with the light output end of the second focusing lens.

8. The Raman spectrometer according to claim 1, characterized in that, The blazing wavelength of the blazed grating is optimized in conjunction with the emission wavelength of the laser module, and the blazing wavelength parameter is determined based on the laser emission wavelength.

9. The Raman spectrometer according to claim 1, characterized in that, The laser module includes at least one laser, which may be a single-wavelength laser or two different-wavelength lasers. When two different-wavelength lasers are used, the system is equipped with an optical path switching mechanism to switch the front-end optical path to adapt to the Raman scattering detection requirements of different substances.

10. A method for processing spectral data for a Raman spectrometer, applied to the Raman spectrometer according to any one of claims 1-9, characterized in that, Includes the following steps: Step S1: Acquire a two-dimensional Raman spectral image acquired by the general CMOS area array image sensor in the Raman spectrometer. The two-dimensional Raman spectral image contains spatial dimension information and dispersion dimension information. Step S2: Based on the dispersion direction of the two-dimensional Raman spectrum image, select the region of interest that is perpendicular to the dispersion direction and has the smallest aberration, and extract multiple rows of effective spectral data within the region of interest; Step S3: Perform inter-row cumulative averaging on the multiple rows of valid spectral data. Utilize the characteristics of inter-row signal coherence and noise incoherence to suppress spectral noise and generate one-dimensional Raman spectral data with enhanced signal-to-noise ratio to meet the requirements of Raman shift detection.

Images