Perovskite non-invasive blood glucose detection sensor based on photoplethysmography

Abstract

The application provides a perovskite non-invasive blood glucose detection sensor based on a photoelectric plethysmography, when the light absorption layer of the sensor is subjected to light radiation, absorbs photons to generate excitons, and diffuses to the metal electrode and the glass electrode on the two sides to form a potential difference, through the potential difference, a closed loop is formed in the sensor to further perform blood glucose detection and analysis. In the traditional lead-based perovskite, SnI2 is doped in the same proportion as PbI2 to reduce the band gap; at the same time, a protective layer is added between the light absorption layer and the transmission layer, which can not only passivate the perovskite and reduce the surface non-radiation recombination loss, but also can isolate the invasion of external water and oxygen due to the smooth film, large crystal grains and complete coverage of the perovskite light absorption layer, so as to serve as a self-sealing layer. The sensor has strong stability, can still maintain excellent photoelectric properties under atmospheric environment and high light intensity continuous irradiation, is more sensitive, accurate and scientific compared with the existing blood glucose detection.

Description

Technical Field

[0001] This application relates to the field of optoelectronic sensing technology, and in particular to a perovskite non-invasive blood glucose detection sensor based on photoplethysmography. Background Technology

[0002] Currently, the method for self-monitoring blood glucose at home is invasive, requiring patients to prick their finger to collect a subcutaneous blood sample for testing. This method is not only painful but also carries the risk of developing blood disorders due to repeated wounds. Therefore, the need for non-invasive blood glucose testing technology is urgent.

[0003] PPG (Positive Photoelectric Glucose) is a photoelectric technology used to detect changes in blood volume in tissue vessels. This technology is widely used in clinical measurements of heart rate (HR) and blood oxygen saturation (SaO2). Blood glucose testing, however, differs. It requires precise measurement of subtle changes in the infrared PPG signal under certain background signals. However, due to limitations in the performance of infrared photoelectric sensors and individual differences, blood glucose testing rarely achieves ideal results. Summary of the Invention

[0004] This application aims to address the technical problem of inaccurate blood glucose detection by providing a perovskite non-invasive blood glucose detection sensor based on photoplethysmography.

[0005] Specifically, this application provides a perovskite non-invasive blood glucose detection sensor based on photoplethysmography. The sensor is composed of a glass electrode, a first charge transport layer, a light-absorbing layer, a protective layer, a second charge transport layer, and a metal electrode stacked in sequence.

[0006] Specifically, when the light-absorbing layer is exposed to light radiation, it absorbs photons to generate excitons, which diffuse to the metal electrodes and glass electrodes on both sides to form a potential difference, thereby forming a closed loop in the sensor to generate current.

[0007] The protective layer is perovskite, and a layer of PEAI is spin-coated onto the perovskite to generate a two-dimensional perovskite PEA2Sn. 0.5 Pb 0.5 I4; wherein the perovskite is doped with SnI2 in the same proportion as PbI2;

[0008] The light radiation is characterized by near-infrared light above 800nm.

[0009] The exciton also includes:

[0010] The excitons are electron-hole pairs bound by Coulomb force;

[0011] When the exciton diffuses toward the metal electrode, it dissociates at the interface between the light-absorbing layer and the transport layer under the influence of the built-in electric field. Electrons transition to the excited state and enter the LUMO energy level, while holes remain in the HOMO energy level and become free carriers.

[0012] The perovskite also includes:

[0013] The perovskite is obtained using a vacuum evaporation vapor deposition method. The material is heated in an evaporator, causing it to sublimate. The evaporated particle stream is then directly directed onto a substrate, depositing a solid film on the substrate to form the perovskite. This method readily produces large-area, uniform, and dense films, significantly improving its cost-effectiveness.

[0014] Based on the same concept, this application also provides a method for fabricating a perovskite non-invasive blood glucose detection sensor based on photoplethysmography, as detailed below:

[0015] ITO or FTO conductive glass is used as the substrate for the glass electrode; the shape and number of electrodes are etched according to the required area size, and one of them is selected as the common electrode.

[0016] The etched substrate is ultrasonically cleaned sequentially with detergent, deionized water, acetone, and isopropanol; the cleaned substrate is dried in an oven and then placed in an oxygen plasma cleaner for O2 plasma treatment.

[0017] After plasma treatment, spin-coat PEDOT:PSS solution is performed. The solution is filtered using a filter head before spin-coating.

[0018] Annealing is performed to form a PEDOT:PSS thin film, which serves as a hole transport layer.

[0019] After annealing, the substrate is transferred to a coating machine and the quartz crucibles containing PbI2 and SnI2 powders are heated separately to evaporate the powders and obtain an orange-yellow PbI2 and SnI2 mixed film.

[0020] Heating the orange-yellow PbI2 and SnI2 mixed film yields a brownish-black MASn. 0.5 Pb 0.5 I3 perovskite;

[0021] Spin-coating a layer of PEAI onto the perovskite to generate a two-dimensional perovskite PEA2Sn 0.5 Pb 0.5 I4;

[0022] After the annealed perovskite is cooled, a PCBM thin film is spin-coated as an electron transport layer, and a BCP thin film is spin-coated as a hole blocking layer; finally, an Ag electrode is thermally evaporated as the top electrode.

[0023] The Ag electrode is deposited perpendicular to the glass electrode direction to form a complete sensor.

[0024] The plasma treatment is carried out in an ion cleaner using O2 plasma for 5-10 minutes at a pressure of 70 Pa and a power of 90 W.

[0025] The thickness of the PbI2 and SnI2 powder mixture is 220 nm.

[0026] The process of heating the orange-yellow PbI2 and SnI2 mixed film yields a brownish-black MASn. 0.5 Pb 0.5 I3 perovskite, also includes:

[0027] The deposited orange-yellow PbI2 and SnI2 mixed film was placed in a vacuum oven with a long crucible containing 50-300 mg of methylamine iodide powder. After the pressure in the vacuum oven was reduced to below 10 kPa, the crucible was heated to 150-200 °C and maintained for 10-30 min. During the heating process, the film gradually changed from orange-yellow to brownish-black MASn. 0.5 Pb 0.5 I3 perovskite.

[0028] Based on the same concept, the present invention also provides a blood glucose detection method using a perovskite non-invasive blood glucose sensor based on photoplethysmography, comprising the following steps:

[0029] S100: Near-infrared light is used to irradiate the skin and muscle tissue, and the near-infrared light penetrates the skin and muscle tissue to reach the blood glucose detection sensor.

[0030] S200: The metal electrode of the blood glucose detection sensor is attached to the skin and muscle tissue. The near-infrared light reaches the light-absorbing layer of the blood glucose detection sensor, so that when the light-absorbing layer is irradiated by light, it absorbs photons and generates excitons, which diffuse to the metal electrode and glass electrode on both sides to form a potential difference, thereby forming a closed loop of the sensor to generate current.

[0031] S300: Measure the PPG signal of blood glucose, and use the PPG signal and the real blood glucose value measured by the current blood glucose detection sensor as a set of data. Repeatedly test the PPG signal under different conditions to obtain a dataset; use feature engineering to extract the features of the PPG signal and replace them as input, and together with the real blood glucose value, form a feature dataset.

[0032] S400: Divide the feature dataset into a training set and a test set, and input them into a machine learning model for calculation to obtain the predicted blood glucose value.

[0033] The method of extracting PPG signal features using feature engineering also includes:

[0034] Feature engineering is used to extract features from PPG signals, including at least the features representing the PPG signal waveform such as double peaks, width, double peak width, diphthong notch, rise rate, and fall rate.

[0035] The machine learning model can be either a neural network or a deep learning model.

[0036] Compared with the prior art, the beneficial effects of this application are as follows:

[0037] The purpose of this invention is to provide a non-invasive blood glucose detection sensor based on photoplethysmography (PPG). This sensor incorporates SnI2 in a proportion equal to PbI2 into a traditional lead-based perovskite substrate to reduce its band gap. Simultaneously, a protective layer is added between the light-absorbing layer and the transmission layer; specifically, a two-dimensional perovskite PEA2Sn is generated by spin-coating a layer of PEAI onto the perovskite. 0.5 Pb 0.5 I4 can not only passivate perovskites and reduce surface non-radiative recombination losses, but also, due to its smooth film, large grains, and ability to completely cover the perovskite light-absorbing layer, it can serve as a self-encapsulating layer to isolate the intrusion of external water and oxygen.

[0038] This invention analyzes and calculates the measured PPG signal using a machine learning model, thereby achieving non-invasive blood glucose detection. Compared to existing blood glucose detection methods, it realizes a perovskite PPG sensor for non-invasive blood glucose detection. Furthermore, in the Clarke error network comparison between predicted and true values, the error falls within 100% of the range, further demonstrating that the detection sensitivity of the perovskite non-invasive blood glucose detection sensor based on photoplethysmography provided by this invention meets the standards and can be widely used in medical diagnosis. Moreover, this PPG sensor has extremely sensitive low-light response characteristics; even when using natural light as a light source, it can still sensitively measure the PPG waveform and provide accurate blood glucose prediction values. It exhibits strong stability, maintaining excellent photoelectric properties even in atmospheric environments and under continuous high-intensity light irradiation. This invention's sensor is compatible with current semiconductor processes and can be fabricated on a large scale. Attached Figure Description

[0039] Figure 1 This is a schematic diagram of a perovskite non-invasive blood glucose detection sensor based on photoplethysmography as described in this application.

[0040] Figure 2 This is a flowchart of the blood glucose detection method of the perovskite non-invasive blood glucose sensor based on photoplethysmography as described in this application.

[0041] Figure 3This is a comparison graph showing the current density-voltage characteristics of perovskite blood glucose sensors with and without PEAI passivation.

[0042] Figure 4 The image shows the results of a test conducted under atmospheric conditions to measure the stability of an optical switch under continuous operation for 5200 seconds.

[0043] Figure 5 This is a diagram showing how the photocurrent density of the sensor changes with variations in light intensity.

[0044] Figure 6 The weakest near-infrared light that the sensor described in this invention can measure under a certain light intensity.

[0045] Figure 7 This shows the extraction of PPG signals by this sensor under near-infrared light, a solar simulator, and sunlight as the light source.

[0046] Figure 8 The model MARD, trained on a dataset using near-infrared light as the light source, achieved a score of 2.48%.

[0047] Figure 9 This shows the predicted data under the sunlight simulator. Detailed Implementation

[0048] This application provides a perovskite non-invasive blood glucose detection sensor based on photoplethysmography to solve the technical problems of existing blood glucose meters being inaccurate, having high requirements for detection conditions, and being impractical.

[0049] The following describes in further detail a perovskite non-invasive blood glucose detection sensor based on photoplethysmography, with reference to specific embodiments and accompanying drawings.

[0050] Example 1:

[0051] Please see Figure 1 This application provides a perovskite non-invasive blood glucose detection sensor based on photoplethysmography. The sensor is composed of a glass electrode, a first charge transport layer, a light-absorbing layer, a protective layer, a second charge transport layer, and a metal electrode stacked in sequence.

[0052] Glass electrodes are used to provide structural support and protect other layers.

[0053] First charge transport layer: A layer located above the glass electrode, used to transport and collect the charge generated from the light-absorbing layer.

[0054] Light-absorbing layer: Located above the first charge transport layer, it has the property of absorbing photons and generating excitons (charge pairs).

[0055] Protective layer: Located above the light-absorbing layer, it is used to protect the light-absorbing layer from interference and damage from the external environment.

[0056] Second charge transport layer: Located above the protective layer, similar to the first charge transport layer, it is used to transport and collect the charge generated by excitons.

[0057] Metal electrodes: Located on top of the second charge transport layer, these electrodes collect and conduct charges, forming closed loops and generating current. The metal electrodes contact the skin and muscle tissue for blood glucose detection.

[0058] The purpose of the above structure is to optimize photoelectric conversion efficiency, provide stable charge transport and protective layer protection, and ensure accurate recording of relevant information required for blood glucose testing.

[0059] Specifically:

[0060] When exposed to light radiation, the light-absorbing layer absorbs photons and generates excitons, which diffuse to the metal and glass electrodes on both sides, creating a potential difference. This potential difference forms a closed loop inside the sensor, allowing current to flow. This current generation and flow can be used for further blood glucose detection and analysis.

[0061] The protective layer is perovskite, and a layer of PEAI is spin-coated onto the perovskite to generate a two-dimensional perovskite PEA2Sn. 0.5 Pb 0.5 I4; where SnI2 is doped into perovskite in the same proportion as PbI2; perovskite is a material with excellent light absorption and charge transport properties, and can be used in optoelectronic devices such as solar cells and photoelectric sensors. PEAI (Phenylethylammonium iodide) is a commonly used organic cationic material used for the modification and performance regulation of perovskite materials. By coating perovskite with PEAI, the stability and photoelectric properties of the perovskite material can be improved. Furthermore, doping perovskite with SnI2 in the same proportion as PbI2 can adjust the band structure and electronic conduction properties of the perovskite material, thereby further optimizing the sensor performance. The application of these material regulation strategies can improve the photoelectric conversion efficiency, stability, and sensitivity of the sensor, further enhancing the performance of non-invasive blood glucose monitoring sensors in measurement and analysis.

[0062] The light radiation used is characterized by near-infrared light above 800 nm. Near-infrared light has a longer wavelength, allowing it to penetrate tissues and skin effectively, and is harmless to the human body. The main reason for using near-infrared light in non-invasive blood glucose testing is that it interacts specifically with glucose molecules in the blood within this wavelength range. By measuring the spectral characteristics of blood absorption, scattering, and reflection, changes in blood glucose concentration can be observed.

[0063] The exciton also includes:

[0064] The excitons are electron-hole pairs bound by Coulomb force;

[0065] When the excitons diffuse toward the metal electrode, they dissociate at the interface between the light-absorbing layer and the transport layer under the influence of the built-in electric field. Electrons transition to an excited state, entering the LUMO (lowest unoccupied molecular orbital) energy level, while holes remain at the HOMO (highest occupied molecular orbital) energy level, becoming free charge carriers. This dissociation of electrons and holes converts the energy of photoexcitation into electrical energy, thereby generating a current in the sensor. This current can be connected to an external circuit via electrodes for measuring and analyzing blood glucose concentration, enabling non-invasive blood glucose detection. Through this mechanism, the formation and dissociation of excitons play a crucial role in converting light energy into electrical energy in non-invasive blood glucose sensors.

[0066] The perovskite also includes:

[0067] The perovskite is obtained using a vacuum evaporation vapor deposition method. The material is heated in an evaporator, causing it to sublimate. The evaporation particle stream is then directly directed onto a substrate, depositing a solid film of perovskite. This method readily produces large-area, uniform, and dense films. This method has several advantages. First, using a vacuum environment for evaporation deposition provides high purity and control, resulting in high-quality perovskite films. Second, this method facilitates the formation of large-area, uniform, and dense films, which is crucial for the performance and stability of perovskite materials. Furthermore, this method offers high cost-effectiveness, achieving relatively low production costs.

[0068] Example 2:

[0069] This application also provides a method for fabricating a perovskite non-invasive blood glucose detection sensor based on photoplethysmography, as detailed below:

[0070] ITO (indium oxide) or FTO (tin oxide) conductive glass is used as the substrate for the glass electrodes. The shape and number of electrodes are etched according to the required area size, and one electrode is selected as a common electrode. Specifically, the shape and number of electrodes can be etched based on the area size. Typically, depending on the specific device and application requirements, the required electrode shape can be defined on the conductive glass using methods such as photolithography, and multiple electrode areas can be embossed using etching methods. This allows one electrode to be selected as the common electrode for connection to other components or circuits. The etching and electrode shape definition process needs to be performed under specific process conditions to ensure good electrode quality and electrical performance. Furthermore, it is important to select appropriate etchants and etching times to control the shape and size of the electrodes and ensure that the electrodes maintain their smoothness during the etching process.

[0071] To better illustrate the fabrication method of the perovskite non-invasive blood glucose detection sensor based on photoplethysmography, taking the fabrication of a 5×1 perovskite photoelectric sensor as an example, an ITO conductive glass with a sheet resistance of 7Ω / square and an area of ​​15mm×15mm is selected. Six electrode patterns with a width of 2mm and an electrode spacing of 0.5mm can be etched, and one of them is selected as the common electrode.

[0072] Based on the etched substrate, ultrasonic cleaning is performed sequentially using detergent, deionized water, acetone, and isopropanol for 15-30 minutes; among which:

[0073] Using detergent: Immerse the etched substrate in the detergent solution. Ultrasonic cleaning can enhance the cleaning effect and remove impurities and organic matter from the substrate surface.

[0074] Use deionized water: Remove the substrate from the detergent and rinse it with deionized water to remove detergent residue and impurities.

[0075] Using acetone and isopropanol: Treating the substrate in acetone and isopropanol solutions respectively can remove residual organic matter and contaminants that may adhere to the substrate surface.

[0076] After cleaning, the substrate is dried in an oven and then placed in an oxygen plasma cleaner for 5-10 minutes at a pressure of 70 Pa and a power of 90 W. This step can remove organic residues and impurities from the substrate surface and provide good surface activity, thus providing good substrate conditions for the subsequent growth of nanomaterials.

[0077] After plasma treatment, PEDOT:PSS solution was spin-coated. The solution was filtered using a filter before spin-coating. The spin-coating parameters were 4000 rpm for 30 seconds.

[0078] The film is placed on a hot plate at 150°C for at least 10 minutes to undergo annealing to form a PEDOT:PSS film, which serves as a hole transport layer.

[0079] After annealing, the substrate is transferred to the coating machine, and the coating machine is evacuated to a pressure of 7×10⁻⁶. -3 Below Pa, quartz crucibles containing PbI2 and SnI2 powders are heated separately to allow both to evaporate. The evaporation rates are kept similar for both, resulting in a thickness of 150-300 nm, yielding an orange-yellow mixed PbI2 and SnI2 film. The thickness of the PbI2 and SnI2 mixed film is 220 nm.

[0080] Heating the orange-yellow PbI2 and SnI2 mixed film yields a brownish-black MASn. 0.5 Pb 0.5 I3 perovskite. Specifically: The deposited orange-yellow PbI2 and SnI2 mixed film is placed in a vacuum oven with a long crucible containing 50-300 mg of methylamine iodide powder. After the pressure in the vacuum oven is reduced to below 10 kPa, the crucible is heated to 150-200℃ and maintained for 10-30 min. During the heating process, the film gradually changes from orange-yellow to brownish-black MASn. 0.5 Pb 0.5 I3 perovskite. The heating temperature is 180°C; the heating duration is 15 minutes.

[0081] Spin-coating a layer of PEAI onto the perovskite to generate a two-dimensional perovskite PEA2Sn 0.5 Pb 0.5 I4: Transfer the sample to a glove box and, after cooling, spin-coat the perovskite surface with PEAI solution. The spin-coating parameters are 4000 rpm for 30 seconds and the concentration is 1-5 mg / mL. While removing the MAI adhering to the film surface, anneal at 100°C for 5-10 minutes on a hot plate to generate a two-dimensional perovskite protective layer.

[0082] After the annealed perovskite has cooled, a PCBM thin film is spin-coated in a glove box as an electron transport layer at a concentration of 20 mg / mL and spin-coating parameters of 1000 rpm for 30 s. Following this, a BCP thin film is spin-coated as a hole blocking layer at a concentration of 0.5-1 mg / mL and spin-coating parameters of 4000 rpm for 30 s. This completes the fabrication of the perovskite photodiode, and the device structure is as follows:

[0083] Glass / ITO / PEDOT:PSS / MASn 0.5 Pb 0.5 I3 / PEASn 0.5 Pb 0.5I3 / PCBM / BCP;

[0084] Finally, a 60-200 nm thick Ag electrode is thermally deposited as the top electrode. The strip-shaped Ag electrode is deposited perpendicular to the ITO electrode direction. This forms the complete device with the following structure:

[0085] Glass / ITO / PEDOT:PSS / MASn 0.5 Pb 0.5 I3 / PEASn 0.5 Pb 0.5 I3 / PCBM / BCP / Ag;

[0086] An 860nm laser is used to illuminate the finger (at any distance), and a perovskite sensor is used to detect the near-infrared light that passes through the finger. A source meter is used to collect the current change curve of the sensor. While testing the PPG waveform, a medical blood glucose meter is used to test the actual blood glucose value. Multiple tests are conducted to obtain PPG waveforms under different blood glucose conditions to form a dataset.

[0087] Example 3:

[0088] like Figure 2 As shown, the present invention also provides a blood glucose detection method based on a perovskite non-invasive blood glucose sensor using photoplethysmography, comprising the following steps:

[0089] S100: Near-infrared light is used to irradiate the skin and muscle tissue, and the near-infrared light penetrates the skin and muscle tissue to reach the blood glucose detection sensor.

[0090] S200: The metal electrode of the blood glucose detection sensor is attached to the skin and muscle tissue. The near-infrared light reaches the light-absorbing layer of the blood glucose detection sensor, so that when the light-absorbing layer is irradiated by light, it absorbs photons and generates excitons, which diffuse to the metal electrode and glass electrode on both sides to form a potential difference, thereby forming a closed loop of the sensor to generate current.

[0091] S300: Measure the PPG signal of blood glucose, and use the PPG signal and the real blood glucose value measured by the current blood glucose detection sensor as a set of data. Repeatedly test the PPG signal under different conditions to obtain a dataset; use feature engineering to extract the features of the PPG signal and replace them as input, and together with the real blood glucose value, form a feature dataset.

[0092] S400: Divide the feature dataset into a training set and a test set, and input them into a machine learning model for calculation to obtain the predicted blood glucose value.

[0093] The method of extracting PPG signal features using feature engineering also includes:

[0094] Feature engineering is used to extract features from PPG signals, including at least the features representing the PPG signal waveform such as double peaks, width, double peak width, diphthong notch, rise rate, and fall rate.

[0095] The machine learning model can be either a neural network or a deep learning model.

[0096] To better illustrate the blood glucose detection method of the perovskite non-invasive blood glucose sensor based on photoplethysmography, the sensor performance is tested in the following experiments:

[0097] Figure 3 The current density-voltage characteristics of perovskite blood glucose sensors with and without PEAI passivation were compared. Both devices, regardless of whether they were spin-coated with PEAI, exhibited good diode characteristics immediately after fabrication. However, after being exposed to the atmosphere for a period of time, the diode characteristics of the uncoated device were almost lost, and the current density-voltage curve became a straight line. The device with PEAI coating showed almost no change in performance and exhibited good atmospheric stability, thus opening up possibilities for its practical applications.

[0098] Optical switching stability reflects the device's ability to operate stably. In this invention, under continuous optical switching conditions for 5200 seconds, the photocurrent and dark current remained almost unchanged. Stability measurements were conducted under atmospheric conditions. Figure 4 .

[0099] Linear dynamic response (LDR) is an important parameter in the field of optoelectronic devices because the device needs to extract the detected light intensity from the corresponding photocurrent over a wide range of light intensity. For example... Figure 5 As the light intensity changes, the photocurrent density of the device also changes accordingly, with its coefficient of determination (R²) satisfying a linear relationship as high as 0.87, and the linear dynamic response reaching 204 dB. This device can respond to an extremely wide range of light, exhibiting a transient optical response to near-infrared lasers as low as 1.56 x 10⁻¹⁰ W cm⁻², with a responsivity as high as 0.4 A W⁻¹. This high optical response to weak light ensures that the device can detect subtle changes in light intensity caused by changes in blood glucose levels during PPG signal detection.

[0100] Blood glucose testing requires accurately obtaining a PPG waveform containing complete blood glucose information under a certain background light. To determine the weakest near-infrared light that this invention can measure under a specific light intensity, the final result is... Figure 6 This invention can achieve low-light recognition with a minimum power of 42nW in both dark and bright light conditions.

[0101] Figure 7The sensor's ability to extract PPG signals was tested under near-infrared light, a solar simulator, and sunlight as the light source. A distinct double-peak PPG waveform was extracted under all three conditions. Furthermore, the sensor also extracted a clear PPG waveform when the sun was obscured by clouds, demonstrating its sensitivity to changes in near-infrared light. Using natural light as the light source could reduce system power consumption and complexity, thus improving its commercial application prospects.

[0102] Figure 8 The model trained on a dataset using near-infrared light as the light source achieved a MARD of 2.48%. The true and predicted values ​​are depicted within a Clarke error grid. All estimated glucose values ​​are located in region A, the area on either side of the diagonal. The shape of this region indicates that the difference between the actual and reference blood glucose values ​​measured by the glucometer is less than 20%, and both the blood glucose value and the reference value are within the hypoglycemic range (<70 mg / dL). Data in this region can help doctors make accurate clinical judgments.

[0103] The experiment was repeated in the near-infrared laser to obtain a dataset from a solar simulator, and the model achieved a MARD of 7.78%. 91.7% of the data points were located in region A, and the remainder in region B. The prediction data from the solar simulator can still be used for medical diagnosis, such as... Figure 9 .

[0104] By setting the light source to two different wavelengths, 660nm and 910nm, and comparing the changes in light intensity transmitted through the finger, blood oxygen detection can be achieved.

[0105] If the substrate material of the sensor in this invention is changed from glass to a flexible material (such as PET), since the sensor can regard sunlight as a light source, the device does not need an external light source, the integrated circuit area is reduced, and it is expected to realize small blood glucose detectors in the shape of rings, nails, etc.

[0106] By extracting different PPG waveforms and analyzing other properties of the waveforms, they can be used for pregnancy detection, blink fatigue detection, etc.

[0107] Although exemplary embodiments have been described herein with reference to the accompanying drawings, it should be understood that the above exemplary embodiments are merely illustrative and are not intended to limit the scope of this application. Various changes and modifications can be made therein by those skilled in the art without departing from the scope and spirit of this application. All such changes and modifications are intended to be included within the scope of this application as claimed in the appended claims.

[0108] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0109] Although the description of this application has been made in conjunction with the specific embodiments described above, it will be apparent to those skilled in the art that many substitutions, modifications, and variations can be made based on the foregoing. Therefore, all such substitutions, modifications, and variations are included within the spirit and scope of the appended claims.

Claims

1. A perovskite non-invasive blood glucose detection sensor based on photoplethysmography, characterized in that, The sensor is composed of a glass electrode, a first charge transport layer, a light-absorbing layer, a protective layer, a second charge transport layer, and a metal electrode stacked in sequence. When the light-absorbing layer is exposed to light radiation, it absorbs photons to generate excitons, which diffuse to the metal electrode and the glass electrode on both sides to form a potential difference, thereby forming a closed loop in the sensor to generate current. The protective layer is perovskite, and a layer of PEAI is spin-coated onto the perovskite to generate a two-dimensional perovskite PEA2Sn. 0.5 Pb 0.5 I4; wherein the perovskite is doped with SnI2 in the same proportion as PbI2; The light radiation is characterized by near-infrared light above 800 nm; The perovskite also includes: The perovskite is obtained by vacuum evaporation deposition, in which the material is heated in an evaporator to sublimate, and the evaporation particle stream is directly directed onto the substrate to deposit a solid film on the substrate, thus obtaining the perovskite. ITO or FTO conductive glass is used as the substrate for the glass electrode; the shape and number of electrodes are etched according to the required area size, and one of them is selected as the common electrode. The etched substrate is ultrasonically cleaned sequentially with detergent, deionized water, acetone, and isopropanol; the cleaned substrate is dried in an oven and then placed in an oxygen plasma cleaner for O2 plasma treatment. After plasma treatment, spin-coat PEDOT:PSS solution is performed. The solution is filtered using a filter head before spin-coating. Annealing is performed to form a PEDOT:PSS thin film, which serves as a hole transport layer. After annealing, the substrate is transferred to a coating machine and the quartz crucibles containing PbI2 and SnI2 powders are heated separately to evaporate the powders and obtain an orange-yellow PbI2 and SnI2 mixed film. Heating the orange-yellow PbI2 and SnI2 mixed film yields a brownish-black MASn. 0.5 Pb 0.5 I3 perovskite; Spin-coating a layer of PEAI onto the perovskite to generate a two-dimensional perovskite PEA2Sn 0.5 Pb 0.5 I4; After the annealed perovskite is cooled, a PCBM thin film is spin-coated as an electron transport layer, and a BCP thin film is spin-coated as a hole blocking layer; finally, an Ag electrode is thermally evaporated as the top electrode. The Ag electrode is deposited perpendicular to the glass electrode direction to form a complete sensor.

2. The perovskite non-invasive blood glucose detection sensor based on photoplethysmography according to claim 1, characterized in that, The exciton also includes: The excitons are electron-hole pairs bound by Coulomb force; When the exciton diffuses toward the metal electrode, it dissociates at the interface between the light-absorbing layer and the transport layer under the influence of the built-in electric field. Electrons transition to the excited state and enter the LUMO energy level, while holes remain in the HOMO energy level and become free carriers.

3. The perovskite non-invasive blood glucose detection sensor based on photoplethysmography according to claim 1, characterized in that, The plasma treatment is carried out in an ion cleaner using O2 plasma for 5-10 minutes at a pressure of 70 Pa and a power of 90 W.

4. The perovskite non-invasive blood glucose detection sensor based on photoplethysmography as described in claim 1, characterized in that, The thickness of the PbI2 and SnI2 powder mixture is 220 nm.

5. A perovskite non-invasive blood glucose detection sensor based on photoplethysmography as described in claim 2, characterized in that, The process of heating the orange-yellow PbI2 and SnI2 mixed film yields a brownish-black MASn. 0.5 Pb 0.5 I3 perovskite, also includes: The deposited orange-yellow PbI2 and SnI2 mixed film was placed in a vacuum oven with a long crucible containing 50-300 mg of methylamine iodide powder. After the pressure in the vacuum oven was reduced to below 10 kPa, the crucible was heated to 150-200 °C and maintained for 10-30 min. During the heating process, the film gradually changed from orange-yellow to brownish-black MASn. 0.5 Pb 0.5 I3 perovskite.

6. A blood glucose detection method based on a perovskite non-invasive blood glucose sensor according to any one of claims 1-5, characterized in that, Includes the following steps: S100: Near-infrared light is used to irradiate the skin and muscle tissue, and the near-infrared light penetrates the skin and muscle tissue to reach the blood glucose detection sensor. S200: The metal electrode of the blood glucose detection sensor is attached to the skin and muscle tissue. The near-infrared light reaches the light-absorbing layer of the blood glucose detection sensor, so that when the light-absorbing layer is irradiated by light, it absorbs photons and generates excitons, which diffuse to the metal electrode and glass electrode on both sides to form a potential difference, thereby forming a closed loop of the sensor to generate current. S300: Measure the PPG signal of blood glucose, and use the PPG signal and the actual blood glucose value measured by the current blood glucose detection sensor as a set of data. Repeatedly test the PPG signal under different conditions to obtain a dataset. Feature engineering is used to extract PPG signal features to replace them as input, which together with the real blood glucose values ​​form a feature dataset. S400: Divide the feature dataset into a training set and a test set, and input them into a machine learning model for calculation to obtain the predicted blood glucose value.

7. The blood glucose detection method according to claim 6, characterized in that, The method of extracting PPG signal features using feature engineering also includes: Feature engineering is used to extract features from PPG signals, including at least the features representing the PPG signal waveform such as double peaks, width, double peak width, diphthong notch, rise rate, and fall rate.

8. The blood glucose detection method according to claim 7, characterized in that, The machine learning model can be either a neural network or a deep learning model.

Images