Alpha-Al2O3 nanoparticle high-precision particle size grading method based on four-layer non-equidistant density gradient liquid and collaborative sedimentation mechanism

Abstract

The invention discloses an alpha-Al2O3 nano-particle high-precision particle size grading method based on a four-layer non-equidistant density gradient solution and a collaborative sedimentation mechanism, which comprises the following steps: S1, carrying out surface dispersion treatment on alpha-Al2O3 nano-particles to form a uniform particle suspension with the mass fraction of 0.1-0.2 wt%; and S2, constructing four layers of non-equidistant density gradient liquid: sequentially and slowly injecting solvents with different densities into the centrifugal tube from bottom to top. According to the method, a four-layer non-equidistant density gradient liquid system is constructed, and a multi-stage differential sedimentation interface is formed in the centrifugation process, so that alpha-Al2O3 nanoparticles in different particle size sections obtain a relatively clear migration path, the mixing crossing probability of particles with approximate particle sizes at the interface is reduced, and the definition of a hierarchical structure is improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of nanoparticle grading, and in particular to a high-precision particle size grading method for α-Al2O3 nanoparticles based on a four-layer non-equidistant density gradient liquid and a synergistic sedimentation mechanism. BACKGROUND

[0002] α-Al2O3 has excellent thermal stability, electrical insulation and mechanical strength, and is widely used in high-end ceramics, polishing liquid, electronic packaging and other fields. In practical applications, the particle size distribution of particles has a significant impact on the terminal performance, especially in the sub-100 nm range, the smaller the particle size standard deviation, the more controllable the material performance, so it is necessary to classify particles in different particle size ranges with high precision.

[0003] Existing nanoparticle size grading techniques mainly include high-speed centrifugation, field flow fractionation, electrophoretic fractionation, sedimentation-anti-solvent method, and density gradient liquid differential centrifugation method, etc. Among them, the density gradient liquid differential centrifugation separation method has become a common choice because of its high separation efficiency for particles with large morphology. This method allows particles to be enriched at the interface according to their own sedimentation rate by building liquid layers with different densities, and in theory it can be used for particle size grading.

[0004] However, in the prior art, when using a single density liquid or a three-layer equidistant density liquid system to grade α-Al2O3 nanoparticles, the following typical technical bottlenecks are faced:

[0005] 1. The interface is not clear, and particles are easy to cross-layer sedimentation between interfaces, resulting in grading aliasing;

[0006] 2. The difference in interfacial tension is not enough to stably capture particles in a specific particle size range;

[0007] 3. The interface is disturbed strongly during operation, and the sample retention rate is low;

[0008] 4. The standard deviation of the particle size distribution of the obtained product is still large, the repeatability is poor, and it is difficult to meet the high-precision application requirements of high-density ceramic-based sintering, chemical mechanical polishing (CMP) slurry, etc.

[0009] Although there have been attempts to use a multi-layer density gradient liquid system to separate particles of different sizes, such as using a three-layer chloroform-ethanol-water system to grade TiO2 nanoparticles in some literature, the particle size standard deviation of the obtained sample is usually large, the interface is not clear, and the particle recovery rate is low, and the distribution control effect is limited. In addition, related research has not mentioned using slow-release sampling or medium-speed centrifugation interface disturbance control means to intervene in particle grading.

[0010] In summary, the existing multi-layer systems generally lack non-equidistant multi-section density design, and it is difficult to achieve stable high-resolution grading in the interval of 20-100 nm in terms of interface configuration, distribution control accuracy, and operation reproducibility. SUMMARY

[0011] The technical problems to be solved: The existing density gradient liquid grading system has multiple key technical problems in the process of fine separation of α-Al2O3 nanoparticles, including particle size overlap caused by unclear interface, low recovery rate caused by particle cross-layer settlement, and poor consistency and repeatability of particle size distribution in multi-batch operation. The existing multi-layer system generally lacks non-equidistant multi-section density design, and it is difficult to achieve stable high-resolution grading in the interval of 20-100 nm.

[0012] In view of the above technical problems, the present application provides a high-precision particle size grading method for α-Al2O3 nanoparticles based on four-layer non-equidistant density gradient liquid and synergistic sedimentation mechanism.

[0013] Technical scheme:

[0014] A high-precision particle size grading method for α-Al2O3 nanoparticles based on four-layer non-equidistant density gradient liquid and synergistic sedimentation mechanism, comprising the following steps:

[0015] S1, dispersing the α-Al2O3 nanoparticles on the surface to form a uniform particle suspension with a mass fraction of 0.1-0.2 wt%;

[0016] S2, constructing a four-layer non-equidistant density gradient liquid: slowly injecting different density solvents into the centrifuge tube from bottom to top, and the density of the solvents from bottom to top is 1.45±0.02 g / cm³, 1.30±0.02 g / cm³, 1.20±0.02 g / cm³ and 0.789±0.005 g / cm³; wherein the "non-equidistant density gradient" means that the density difference between the two adjacent layers is not equidistantly distributed, but is set to be a non-uniform gradient according to the sedimentation behavior of particles of different sizes, so as to form differentiated interception and enrichment conditions at the interface of each layer. Different density solvents can be achieved by various solvent combinations, preferably using chloroform-glycerol, chloroform-isopropanol, chloroform-ethanol and alcohol solvents to construct the four-layer non-equidistant density gradient liquid system; those skilled in the art can reasonably replace or adjust the types of solvents under the premise of meeting the density conditions and interlayer immiscibility.

[0017] S3, slowly adding the particle suspension prepared in S1 to the uppermost surface of the four-layer non-equidistant density gradient liquid without disturbing the interface;

[0018] S4, performing sedimentation fractionation under centrifugal conditions, wherein the temperature of the centrifugal operation is 18-28℃; the centrifugal rotation speed is 3000-5000 rpm, and the centrifugal time is 10-20 minutes; after centrifugation, the α-Al2O3 nanoparticles of different particle sizes form corresponding sedimentation bands at the interfaces between adjacent two layers of the four-layer non-equidistant density gradient liquid and at the interface between the lowermost layer of the solvent and the bottom of the centrifugal tube;

[0019] S5: extracting four groups of α-Al2O3 nanoparticle components from the sedimentation bands at the interfaces between adjacent two layers of the solvent and at the interface between the lowermost layer of the solvent and the bottom of the centrifugal tube, and obtaining four groups of α-Al2O3 nanoparticles with different particle size distribution concentration degrees after washing and drying.

[0020] Further, the particle size distribution interval of the α-Al2O3 nanoparticles in S1 is 20-100 nm, the particle size standard deviation is greater than 12 nm, and the tail of the particle size distribution is significantly expanded.

[0021] Further, the α-Al2O3 nanoparticles in S1 are dispersed in anhydrous ethanol for ultrasonic dispersion, the mass-volume ratio of the α-Al2O3 nanoparticles to anhydrous ethanol is 0.1-0.2 g:100 mL, and the ultrasonic dispersion conditions are 40 kHz, 120 W, and a dispersion time of 30-40 minutes.

[0022] Further, the anhydrous ethanol in S1 also contains a surfactant, and the surfactant includes PVP, PEG, and SDS.

[0023] Further, the injection time of each layer of solvent in S2 is not less than 5 minutes, and the injection angle is not greater than 45°.

[0024] Further, the solvents injected from bottom to top in S2 are, in order, chloroform-glycerol with a volume ratio of 4:1, chloroform-isopropyl alcohol with a volume ratio of 5:1, chloroform-ethanol with a volume ratio of 5:2, and anhydrous ethanol or a density-adjusted dilute alcohol mixture.

[0025] Further, the interface between adjacent two layers of solvent in S2 is covered with a trace of inert oil film.

[0026] Further, the sample loading rate of the particle suspension in S3 is between 0.5-1.5 mL / min, preferably 0.7±0.1 mL / min. A dropper, a slow-flow pump, or a syringe can be used to control the sample loading rate.

[0027] Further, the temperature of the centrifugal operation in S4 is 23±2℃, the centrifugal rotation speed is 4000 rpm±100 rpm, and the centrifugal time is 15±2 minutes.

[0028] Further, the extraction of the α-Al2O3 nanoparticles in the sedimentation zone in S5 is performed by extraction or pipetting, with an extraction rate of ≤1.0 mL / min, preferably 0.5±0.2 mL / min; and the sample obtained after centrifugation is washed with ethanol three times and dried at 60°C for 12 hours.

[0029] Advantages:

[0030] The hierarchical system constructed in the application has certain advantages in interface configuration, operation path and separation stability.

[0031] 1. The four-layer non-equidistant density gradient liquid system constructed in the application forms a plurality of differential sedimentation interfaces during centrifugation, so that the α-Al2O3 nanoparticles in different particle size sections obtain relatively clear migration paths, which is conducive to reducing the mixed crossing probability of particles with close particle sizes at the interface and improving the clarity of the hierarchical structure.

[0032] 2. The non-equidistant density gradient configuration combined with the synergistic sedimentation mechanism in the application enables the particles to migrate and separate step by step under the condition of multiple interfaces, which helps to improve the problem of unclear boundaries in the traditional equidistant or few-layer density system and improve the feasibility of particle size section control from the structure.

[0033] 3. In the hierarchical process described in the application, the four-layer density gradient system can maintain a relatively stable interface structure, which is conducive to reducing the invalid loss of particles during operation and improving the overall stability and repeatability of the hierarchical process.

[0034] 4. The process steps of the method described in the application are clear, the operation path is clear, and the process has certain tolerance to conditions such as centrifugal speed, temperature and particle dispersion state, which helps to reduce the influence of process fluctuations on the hierarchical results and enhance the engineering implementability of the method.

[0035] The system uses an industrially available solvent system and a conventional medium-speed centrifugal device, and can realize batch continuous classification by cooperating with an automatic liquid injection system, and has process feasibility for industrialization and expansion. BRIEF DESCRIPTION OF DRAWINGS

[0036] Figure 1 is a schematic diagram of the four-layer density gradient liquid construction and α-Al2O3 particle classification in the application;

[0037] Figure 2a is a transmission electron microscope (TEM) observation photo of the original α-Al2O3 nanoparticles to be separated;

[0038] Figure 2b is a particle size statistical column chart of the original α-Al2O3 nanoparticles to be separated;

[0039] Figures 3a–3d show TEM images and corresponding particle size histograms of α-Al2O3 nanoparticles in four particle size ranges: group A, group B, group C, and group D, respectively. Detailed Implementation

[0040] The present invention will now be described in detail with reference to the accompanying drawings.

[0041] To clearly illustrate the particle size distribution of the original sample to be separated in this invention, transmission electron microscopy (TEM) images of the original α-Al2O3 nanoparticles to be separated and the corresponding particle size distribution diagram are provided here.

[0042] Figure 2a Typical TEM images of the samples to be separated are shown. The images clearly show that there are obvious size differences and uneven dispersion of the original sample particles, indicating a wide overall particle size distribution.

[0043] Figure 2b The above is a bar chart showing the particle size distribution of the samples. According to the statistical results, the main distribution range of the particle size of the samples to be separated is within the range of 20–100 nm, but the standard deviation of the particle size is relatively large, usually exceeding 12 nm, and the tail of the particle size distribution is significantly extended, indicating the non-uniformity of particle size and poor distribution concentration.

[0044] Example 1:

[0045] A high-precision particle size classification method for α-Al₂O₃ nanoparticles based on a four-layer non-equidistant density gradient liquid and a synergistic sedimentation mechanism includes the following steps:

[0046] Step 1: Particle Dispersion

[0047] α-Al₂O₃ nanoparticles were selected. 0.1–0.2 g of the sample was added to 100 mL of anhydrous ethanol and ultrasonically dispersed at 40 kHz and 120 W for 30–40 minutes to form a homogeneous particle suspension with a mass fraction of 0.1–0.2 wt%. To improve the initial dispersibility of the particles and ensure stable sedimentation and interfacial separation in the density gradient system, a surfactant was added during the dispersion process. The surfactant could be selected from one or more of PVP, PEG, or SDS. Comparative observations showed that without the addition of a surfactant, the particles were prone to secondary agglomeration, and the stability of the hierarchical interface decreased.

[0048] Step 2: Construction of density gradient liquid

[0049] Four different solvents are slowly injected sequentially from bottom to top into a dry, clean centrifuge tube to construct a stable, layered structure with a non-uniformly decreasing density. To minimize interfacial disturbance, the injection time for each layer should be no less than 5 minutes, and the injection angle should not exceed 45°. The volume of solvent in each layer of the centrifuge tube should be the same (the volume difference between layers should not exceed ±10%), but this can be fine-tuned according to the desired fractionation window position. Preferably, the volume of solvent in each layer is approximately one-quarter of the total volume to ensure interfacial stability and gradient continuity.

[0050] Fourth layer (bottom): chloroform-glycerol (volume ratio 4:1), density 1.45±0.02 g / cm³;

[0051] The third layer: chloroform-isopropanol (5:1), with a density of 1.30±0.02 g / cm³;

[0052] The second layer is composed of chloroform and ethanol (5:2), with a density of 1.20 ± 0.02 g / cm³.

[0053] First layer (top): Anhydrous ethanol, density 0.789±0.005 g / cm³.

[0054] To further improve the stability of the layered interface, auxiliary measures for interface stability can be taken during the layering process, such as covering the layered interface with a small amount of inert oil film or moderately controlling the system operating temperature. These measures serve as auxiliary means of interface stability, aiming to reduce the impact of system viscosity fluctuations on the layered structure formation process. Within the preferred centrifugation operating temperature range of this invention, the density gradient layered structure can remain stable, and small fluctuations in operating temperature will not affect the implementation of the sedimentation and classification process.

[0055] Step 3: Sustained-release sample addition

[0056] The particle suspension obtained in step 1 is slowly added dropwise to the surface of the uppermost layer of the density gradient liquid. The addition rate can be controlled between 0.5 and 1.5 mL / min using a burette, a slow-flow pump, or a syringe to effectively avoid disturbing the layering interface.

[0057] Step 4: Centrifugation

[0058] Place the centrifuge tubes into a centrifuge and centrifuge at medium speed at 23±2℃. The preferred speed is 4000±100 rpm and the centrifugation time is 15±2 minutes, which allows particles of different sizes to form stable sedimentation zones at the density interfaces of the density gradient liquid, thereby enhancing the particle size differentiation effect.

[0059] Step 5: Graded extraction and washing and drying

[0060] After centrifugation, four groups of particle components were extracted from the sedimentation zones at each density interface. Interfacial extraction or slow extraction via pipetting could be used, with an extraction rate ≤1.0 mL / min, preferably 0.5±0.2 mL / min. Each sample was washed three times with ethanol and then dried at 60℃ for 12 hours to obtain four groups of samples (A, B, C, and D) with significantly different particle size distributions.

[0061] Step 6: Particle size statistical analysis

[0062] The four fractionated samples were observed and their particle size was measured by transmission electron microscopy (TEM). The number of particles measured in each group was no less than 200. Image processing software was used for calibration and particle size statistical analysis to obtain the average particle size μ and standard deviation σ of the corresponding samples.

[0063] Multiple parallel separation operations were performed during the experiment. Each experiment used a freshly prepared density gradient liquid system and was conducted under consistent operating conditions and processes. After each separation operation, sedimentation samples from groups A, B, C, and D were collected sequentially, washed with ethanol, dried, and then subjected to particle size statistical analysis. To minimize the impact of initial raw material differences on the results, α-Al₂O₃ nanoparticles prepared in the same batch were used as the fractionation material in all experiments, and the slow-release injection path and centrifugation parameters were kept consistent.

[0064] Repeated experiments showed that the particle size distribution of the obtained graded samples generally exhibited a Gaussian concentration trend, with most sample groups showing relatively low dispersion, demonstrating good grading concentration. Samples in other particle size ranges also showed relatively concentrated distribution characteristics, indicating that the grading system can distinguish particles in different size ranges. Under representative experimental conditions, the particle retention rate in each particle size range was generally at a high level.

[0065] Table 1:

[0066]

[0067] Note: The particle size statistics shown in Table 1 are from representative experimental batches and are used to reflect the particle size distribution trend of the classification system of this invention during actual operation, rather than to limit the absolute statistical consistency of each batch. The overall experimental results show good process robustness, classification consistency, and interface extraction stability, demonstrating feasibility for engineering scale-up applications. No significant interlayer disturbances were observed at the interfaces during the operation of each batch; the settling zones were sharp, the interface decomposition was clear, and the settling zone contours were stable. Samples in each particle size range maintained high retention levels under representative experimental conditions, with particle retention rates maintained between 85% and 95%, further corroborating the system's interface stability and particle size classification consistency.

[0068] To further reduce the impact of human intervention on the statistical results, each experiment was conducted independently by different operators, and each step was strictly performed in accordance with a unified operating procedure.

[0069] Figures 3a–3d show TEM images and corresponding particle size histograms of α-Al₂O₃ nanoparticles in four particle size ranges: Group A (20–35 nm), Group B (35–55 nm), Group C (55–75 nm), and Group D (75–100 nm). The yellow histograms represent the statistical frequency distribution for each particle size range, and the red curves are Gaussian fitted curves used to characterize the central tendency of the particle size distribution.

[0070] The particle size distributions shown in each figure are selected from representative batches from multiple experiments. Although there are slight differences in the tail of the distribution between different experiments, the main distribution range and the fitted curve maintain good consistency, further demonstrating that the classification system described in this invention has good reproducibility and process tolerance in nanoscale particle size control.

[0071] To systematically analyze the impact of different process parameters on particle settling behavior and classification results, Examples 2–5 were designed as trend derivation experiments. Based on the same classification system and operating path as Example 1, the changes in key parameters were compared and analyzed. The relevant results were mainly used to verify the direction of the influence of parameter changes on settling behavior and classification trends. The trend conclusions are all based on the repeatability observations of representative experiments and do not exceed the range that can be reasonably supported by existing experimental conditions.

[0072] Example 2: Comparison of the separation effect of a three-layer density liquid system

[0073] This embodiment is a comparative experiment used to illustrate the influence of different density gradient layers and configurations on the grading effect; the three-layer density system is only used as a control structure, and its purpose is to compare the separation effect with the four-layer non-equidistant density gradient system of the present invention, and does not constitute the technical solution of the present invention.

[0074] To verify the advantages of the four-layer non-equidistant density gradient liquid system in particle size distribution control, a comparative experiment was constructed, using a traditional three-layer density system to separate the same batch of α-Al₂O₃ nanoparticles. The experimental conditions, centrifugation speed, and particle concentration were kept consistent with those in Example 1, except that the density gradient structure was adjusted to the following three-layer system (from bottom to top):

[0075] Third layer (bottom): Chloroform-glycerol mixture (density 1.45 g / cm³);

[0076] Second layer: Chloroform-isopropanol mixture (density 1.30 g / cm³).

[0077] First layer (top): Anhydrous ethanol (density 0.789 g / cm³).

[0078] After slow-release injection, medium-speed centrifugation was performed. After grading, samples from each interface layer were collected, and their particle size distribution characteristics were compared and analyzed.

[0079] Comparative observation results show that in the traditional three-layer density gradient system, since the interlayer density distribution is not segmented to control the sedimentation behavior of particles of different sizes, the interfacial interception capacity is relatively limited. α-Al2O3 nanoparticles are prone to cross-interface migration during sedimentation, resulting in overlapping and enrichment of particles of different sizes at the interface, and the hierarchical boundaries are not clear enough.

[0080] In contrast, the four-layer non-equidistant density gradient system described in Example 1 sets different density interception conditions at different interfaces, making it easier for particles to accumulate at their corresponding interfaces during sedimentation, which is beneficial for forming sedimentation zones.

[0081] The above comparison results show that the traditional multi-layer density gradient structure alone is insufficient to provide a sufficient resolution control window for particles with similar sizes; while the four-layer non-equidistant density gradient configuration of the present invention, through targeted design of density differences, exhibits a better grading effect in fine segmentation control of particle size.

[0082] Example 3: Comparative Experiment under Different Centrifugation Speeds

[0083] To investigate the effect of centrifugation speed variation on particle sedimentation behavior and the stability of the classification interface, the same four-layer non-equidistant density gradient liquid system as in Example 1 was used. Under the condition of keeping the particle concentration, sample addition method and operating temperature consistent, only the centrifugation speed was compared.

[0084] This embodiment selected three representative rotation speed conditions for comparative experiments: low rotation speed (3000±100 rpm), medium rotation speed (4000±100 rpm), and relatively high rotation speed (5000±100 rpm); the centrifugation time was kept consistent for each group. After centrifugation, the formation state and particle size distribution characteristics of the sedimentation layer at each density interface were compared and observed.

[0085] Comparative observations show that at lower rotational speeds, the overall settling velocity of particles is slower, and particles in some size ranges are unable to form clear enrichment zones at the interface, resulting in less distinct interface zoning. At higher rotational speeds, the particle migration rate increases, and some particles with similar sizes easily cross the interface to settle, leading to a decrease in the clarity of the interface layers.

[0086] In contrast, under medium rotational speed conditions, the sedimentation zone boundaries at each density interface are clearer, and the distribution of the classification sections is relatively stable, which is beneficial for effectively distinguishing particles of different sizes. These results indicate that in the four-layer non-equidistant density gradient system, centrifugal rotational speed affects particle sedimentation paths and interface stability; therefore, reasonable control of centrifugal rotational speed helps improve the stability and repeatability of the classification process.

[0087] Example 4: Comparison of interfacial sedimentation behavior under different dispersion conditions

[0088] To evaluate the impact of changes in the initial dispersion state of particles on sedimentation behavior and the stability of the classification interface, the same four-layer non-equidistant density gradient liquid system and centrifugation conditions as in Example 1 were used. Under the premise of keeping the particle concentration, sample addition method and operating temperature consistent, comparative experiments were conducted on sedimentation behavior under different dispersion conditions.

[0089] The following comparison conditions are set in this embodiment:

[0090] Condition A: No dispersant (surfactant) was introduced;

[0091] Condition B: A dispersant (surfactant) is introduced into the particulate suspension to improve the initial dispersion state. The surfactant may be selected from one or more of PVP, PEG or SDS, and the amount added is 0.1-3 wt%, preferably 0.1-0.15 wt%.

[0092] After undergoing the same dispersion treatment, each system was subjected to slow-release injection and medium-speed centrifugation. After centrifugation, the formation state of the sedimentation layer at each density interface was compared and observed.

[0093] Comparative observation results show that, without the introduction of dispersant, secondary aggregation is prone to occur between particles, the boundaries of the sedimentation zone at some interfaces are not clear enough during the sedimentation process, and there is a certain degree of mixing of particles in different particle size ranges; while under the condition of improved dispersion, the sedimentation layer structure at each interface is relatively stable, and the distinguishability of the graded sections is enhanced.

[0094] The above results indicate that changes in particle dispersion state can affect the sedimentation path and interface behavior in non-equidistant density gradient systems. However, the classification effect described in this invention mainly comes from the density gradient structure and the synergistic sedimentation mechanism. The regulation of dispersion conditions helps to improve the system's operational stability and classification repeatability.

[0095] Example 5: The effect of temperature change on the stability of the settling process

[0096] To evaluate the effect of operating temperature on the sedimentation behavior of particles at the graded interface, a comparative experiment was conducted on the sedimentation process under different temperature conditions using the same four-layer non-equidistant density gradient liquid system as in Example 1, while maintaining the same particle concentration, sample addition method and centrifugation speed.

[0097] This embodiment selects representative temperature conditions below, close to, and above the preferred operating temperature range for comparison, with the specific settings as follows:

[0098] Condition A: 18±1℃ (lower temperature condition);

[0099] Condition B: 23±2℃ (medium temperature condition);

[0100] Condition C: 28±1℃ (higher temperature condition).

[0101] After each set of operations is completed, observe and compare the interface structure of the sedimentation layer with the distribution of the sample.

[0102] Comparative observations show that at lower temperatures, the system viscosity is relatively high, the particle migration rate slows down, and some small-diameter particles are retained near the interface, resulting in unclear classification interface boundaries. At higher temperatures, the system viscosity decreases, the particle migration rate accelerates, the sedimentation path changes, some particle size ranges migrate across the interface, and the interface partitions tend to widen.

[0103] In contrast, within the operating range near medium temperature conditions, the settlement layer structure at each density interface is relatively stable, and the boundaries of the grading sections are relatively clear, which is conducive to maintaining the stability and repeatability of the grading process.

[0104] Although the present invention has been disclosed above with reference to preferred embodiments, these are not intended to limit the invention. Any person skilled in the art can make various changes or modifications without departing from the spirit and scope of the invention. Therefore, the scope of protection of the present invention should be defined by the scope of the claims of this application.

Claims

1. A high-precision particle size classification method for α-Al₂O₃ nanoparticles based on a four-layer non-equidistant density gradient liquid and a synergistic sedimentation mechanism, characterized in that, Includes the following steps: S1. α-Al2O3 nanoparticles are surface dispersed to form a uniform particle suspension with a mass fraction of 0.1–0.2 wt%. S2. Construct four layers of non-equidistant density gradient liquid: Slowly inject solvents of different densities into centrifuge tubes from bottom to top. The solvent densities from bottom to top are 1.45±0.02 g / cm³, 1.30±0.02 g / cm³, 1.20±0.02 g / cm³ and 0.789±0.005 g / cm³, respectively. S3. Without disturbing the interface, slowly drop the particle suspension prepared in S1 onto the surface of the uppermost layer of the four non-equidistant density gradient liquids. S4. Sedimentation classification is carried out under centrifugation conditions, wherein the temperature of centrifugation is 18-28℃, the centrifugation speed is 3000-5000 rpm, and the centrifugation time is 10-20 minutes. After centrifugation, α-Al2O3 nanoparticles of different sizes form corresponding sedimentation zones at the interfaces of adjacent solvent layers in the four layers of non-equidistant density gradient liquid and at the interface between the bottom solvent layer and the bottom of the centrifuge tube. S5: Four groups of α-Al2O3 nanoparticles were extracted from the sedimentation zone at the interface between two adjacent solvent layers and at the interface between the bottom solvent layer and the bottom of the centrifuge tube. After washing and drying, four groups of α-Al2O3 nanoparticles with different particle size distribution concentrations were obtained.

2. The high-precision particle size classification method for α-Al₂O₃ nanoparticles based on a four-layer non-equidistant density gradient liquid and a synergistic sedimentation mechanism, as described in claim 1, is characterized in that... The α-Al2O3 nanoparticles in S1 have a particle size distribution range of 20–100 nm, a standard deviation of particle size greater than 12 nm, and a significant extension at the tail end of the particle size distribution.

3. The high-precision particle size classification method for α-Al₂O₃ nanoparticles based on a four-layer non-equidistant density gradient liquid and a synergistic sedimentation mechanism, as described in claim 1, is characterized in that... In S1, α-Al2O3 nanoparticles were dispersed in anhydrous ethanol and ultrasonically dispersed. The mass-to-volume ratio of α-Al2O3 nanoparticles to anhydrous ethanol was 0.1-0.2 g:100 mL. The ultrasonic dispersion conditions were 40 kHz, 120 W, and the dispersion time was 30-40 minutes.

4. The high-precision particle size classification method for α-Al₂O₃ nanoparticles based on a four-layer non-equidistant density gradient liquid and a synergistic sedimentation mechanism according to claim 3, characterized in that, The anhydrous ethanol in S1 also contains surfactants, including PVP, PEG and SDS.

5. The high-precision particle size classification method for α-Al₂O₃ nanoparticles based on a four-layer non-equidistant density gradient liquid and a synergistic sedimentation mechanism according to claim 1, characterized in that, The injection time for each layer of solvent in S2 shall not be less than 5 minutes, and the injection angle shall not exceed 45°.

6. A high-precision particle size classification method for α-Al₂O₃ nanoparticles based on a four-layer non-equidistant density gradient liquid and a synergistic sedimentation mechanism, as described in any one of claims 1-5, characterized in that, The solvents injected into S2 from bottom to top are: chloroform-glycerol in a volume ratio of 4:1, chloroform-isopropanol in a volume ratio of 5:1, chloroform-ethanol in a volume ratio of 5:2, and anhydrous ethanol or a diluted alcohol mixture adjusted for density.

7. The high-precision particle size classification method for α-Al₂O₃ nanoparticles based on a four-layer non-equidistant density gradient liquid and a synergistic sedimentation mechanism according to claim 1, characterized in that, A small amount of inert oil film covers the interface between two adjacent solvent layers in S2.

8. The high-precision particle size classification method for α-Al₂O₃ nanoparticles based on a four-layer non-equidistant density gradient liquid and a synergistic sedimentation mechanism according to claim 1, characterized in that, The addition rate of the particulate suspension in S3 is between 0.5 and 1.5 mL / min, and the addition rate is controlled by a burette, a slow-flow pump or a syringe.

9. A high-precision particle size classification method for α-Al₂O₃ nanoparticles based on a four-layer non-equidistant density gradient liquid and a synergistic sedimentation mechanism, as described in claim 1, is characterized in that... In S4, the centrifugation operation temperature is 23±2℃, the centrifugation speed is 4000 rpm±100 rpm, and the centrifugation time is 15±2 minutes.

10. A high-precision particle size classification method for α-Al₂O₃ nanoparticles based on a four-layer non-equidistant density gradient liquid and a synergistic sedimentation mechanism, as described in claim 1, is characterized in that... The α-Al2O3 nanoparticles in the sedimentation zone were extracted by extraction or pipetting at a rate ≤1.0 mL / min. After centrifugation, the sample was washed three times with ethanol and then dried at 60℃ for 12 hours.

Images