Method for detecting content of magnetic substances in lithium carbonate

Abstract

The invention provides a method for detecting the content of magnetic substances in lithium carbonate. The method comprises the following steps: S1, carrying out sample collection and pretreatment; s2, accurately weighing the sample and fixing the volume of the sample; s3, carrying out gradient magnetic field magnetic separation and enrichment; s4, carrying out ultrasonic-assisted washing and purification on the enrichment liquid; s5, magnetic substance precipitation dissolution and constant volume are carried out; s6, a step of carrying out ICP-MS instrument calibration and parameter optimization; s7, carrying out detection mother liquor determination and multivariate statistical error correction; s8, carrying out blank experiment correction; s9, carrying out parallel experimental verification; s10, a result calculation and judgment step is carried out; and S11, carrying out method verification and filing. According to the method, the problems in the prior art are creatively solved by organically combining gradient magnetic field separation and enrichment, ultrasonic-assisted purification, ICP-MS (Inductively Coupled Plasma Mass Spectrometry) accurate detection and multivariate statistical error correction aiming at the core pain point of the existing technology for detecting the magnetic substances in the lithium carbonate.

Description

Technical Field

[0001] This invention specifically relates to a method for detecting the content of magnetic substances in lithium carbonate. Background Technology

[0002] Lithium carbonate, as a key strategic resource, is widely used in lithium-ion batteries, pharmaceuticals, ceramics, glass, and other fields. The lithium-ion battery industry is its primary consumption scenario, accounting for over 70%. With the rapid development of industries such as new energy vehicles and energy storage equipment, the market's requirements for the purity of lithium carbonate are becoming increasingly stringent, especially for battery-grade lithium carbonate, which needs to achieve a purity of over 99.9% and has extremely strict control over impurity content. Magnetic substances, a typical harmful impurity in lithium carbonate, mainly originate from elemental metals such as iron, cobalt, and nickel, as well as oxides, introduced during the mining, crushing, and flotation processes of raw lithium ore (such as spodumene and lepidolite), and metal debris generated from the wear and tear of production equipment (such as crushers, grinders, and pipelines). The presence of these magnetic substances can seriously affect product quality: in lithium-ion batteries, magnetic impurities can cause internal short circuits, shorten cycle life, reduce rate performance, and even pose safety hazards; in the pharmaceutical field, magnetic impurities may cause adverse reactions such as organ irritation and inflammation; in the production of high-end ceramics and glass, magnetic impurities can lead to uneven product color and reduced transparency, lowering the product's added value.

[0003] Currently, there is an increasingly urgent need in the industry to detect the content of magnetic materials in lithium carbonate. Various countries have established relevant standards. For example, my country's GB / T11075-2023 standard for lithium carbonate clearly stipulates that the iron content in battery-grade lithium carbonate must be ≤0.001%, and the cobalt and nickel content must be ≤0.0005%. However, existing detection methods still have many limitations and cannot meet the actual needs for high-precision, high-sensitivity, and rapid detection, especially for the detection of trace (ppb-level) magnetic materials. The accuracy and repeatability of existing technologies need to be improved.

[0004] Prior to this method, the detection of magnetic material content in lithium carbonate mainly employed gravimetric methods, magnetic separation-atomic absorption spectrometry, and direct detection methods such as inductively coupled plasma mass spectrometry (ICP-MS). These methods suffer from the following core challenges: First, the detection limits are too high, with existing methods typically having detection limits of 10-50 ppb, which is insufficient to meet the detection requirements of battery-grade lithium carbonate for magnetic materials below the ppb level, and cannot accurately identify the potential risks posed by trace impurities. Second, there are many interfering factors; the lithium carbonate matrix itself has certain ion interference, and the detection process is easily affected by environmental, reagent, and equipment factors, resulting in large detection errors, poor repeatability, and relative standard deviations (RSDs) often exceeding 5%. First, the operation is complex and time-consuming. For example, the gravimetric method requires multiple cumbersome steps such as magnetic separation, filtration, drying, and weighing, and a single test takes more than 8 hours, which cannot meet the needs of industrial batch testing. Second, it cannot achieve precise separation and enrichment of magnetic materials. Existing magnetic separation methods mostly use a single magnetic field strength, which is difficult to completely separate impurities with different magnetic strengths (such as strongly magnetic iron and weakly magnetic iron oxide), resulting in some magnetic impurities not being detected and causing the test results to be low. Third, it lacks a scientific error correction mechanism. Existing methods mostly use simple blank correction and do not consider the systematic and random errors in each link of the detection process, which cannot achieve accurate calibration of the test results and cannot meet the requirements for method stability and accuracy.

[0005] Furthermore, existing publicly available detection methods do not organically combine magnetic separation enrichment, matrix interference elimination, and statistical error correction. They either emphasize separation while neglecting detection accuracy or emphasize detection while neglecting the thoroughness of impurity separation. These methods fail to effectively address the core pain points in the industry, such as "inaccurate detection of trace magnetic substances, low detection efficiency, and difficulty in controlling interference," thus hindering the improvement of lithium carbonate product quality and the high-quality development of the industry.

[0006] In summary, this application proposes a method for detecting the content of magnetic materials in lithium carbonate to solve the above-mentioned problems. Summary of the Invention

[0007] The purpose of this invention is to address the shortcomings of existing technologies by providing a method for detecting the content of magnetic substances in lithium carbonate, which can effectively solve the aforementioned problems.

[0008] To achieve the above requirements, the technical solution adopted by the present invention is: to provide a method for detecting the content of magnetic substances in lithium carbonate, the method comprising the following steps: S1: Steps for sample collection and pretreatment; S2: The steps for accurately weighing and adjusting the volume of the sample; S3: Steps for gradient magnetic field separation and enrichment; S4: The step of performing ultrasonic-assisted washing and purification of the enrichment solution; S5: Steps for precipitating, dissolving, and adjusting the volume of magnetic materials; S6: Steps for ICP-MS instrument calibration and parameter optimization; S7: Steps for determining the mother liquor and correcting for multivariate statistical errors; S8: Steps for performing blank experiment correction; S9: Steps for conducting parallel experimental verification; S10: Steps for calculating and judging the results; S11: Steps for method verification and archiving.

[0009] The advantages of this method for detecting the magnetic material content in lithium carbonate are as follows: 1. High detection precision and accuracy: This method ensures a magnetic material separation and recovery rate of ≥99% through multi-cycle enrichment with a gradient magnetic field. Combined with ultrasonic-assisted purification, it thoroughly removes matrix interference. ICP-MS is used for precise detection, and a multivariate statistical error correction model is employed to effectively eliminate systematic and random errors in each step. Furthermore, a blank experiment is used for secondary correction, lowering the detection limit to below 0.1 ppb. The relative standard deviation (RSD) of the detection results is ≤2.5%, and the spiked recovery rate is between 95% and 105%, significantly superior to existing methods (detection limit 10-50 ppb, RSD ≥5%). It can accurately detect trace magnetic materials below the ppb level, meeting the stringent testing requirements for battery-grade lithium carbonate and providing a scientific basis for improving the quality of lithium carbonate products.

[0010] 2. High detection efficiency and low cost: This method optimizes the operation process of each step, controlling the time for a single detection to within 4 hours. Compared with the existing gravimetric method (more than 8 hours) and magnetic separation-atomic absorption spectrometry method (more than 6 hours), the detection efficiency is improved by more than 50%. Moreover, each step can be partially parallelized, making it suitable for industrial batch detection. At the same time, the reagents (superior purity hydrochloric acid and nitric acid) and equipment (ICP-MS and gradient magnetic field separation device) used in the method are all conventional laboratory reagents and equipment, eliminating the need for special and expensive equipment. The optimized and reasonable reagent dosage can effectively reduce detection costs and improve the economic benefits of enterprises.

[0011] 3. Outstanding creativity, breaking through the limitations of existing technology, for the first time organically combining gradient magnetic field multi-cycle separation and enrichment, ultrasonic-assisted purification, ICP-MS precise detection and multivariate statistical error correction, and designed a unique error correction model and gradient magnetic field parameters, which solves the core pain points of existing methods such as incomplete separation, difficulty in controlling interference and unscientific error correction. It is significantly different from existing publicly available detection methods and fills the technical gap in the precise detection of lithium carbonate magnetic materials below ppb level.

[0012] 4. High versatility and wide applicability: This method is applicable to the detection of lithium carbonate samples of different grades, such as battery grade and industrial grade. It can simultaneously detect multiple magnetic substances such as iron, cobalt, and nickel. There is no need to adjust the detection process for different grades and different magnetic substances; only the judgment threshold needs to be adjusted according to industry standards. At the same time, the method has been validated in multiple dimensions, both indoors and outdoors, and can achieve accurate detection in different laboratories and by different operators. It is highly versatile and can be widely used in lithium carbonate production enterprises, third-party testing institutions, research institutes, and other scenarios, and has good promotion and application value.

[0013] 5. The method is standardized and highly traceable. It includes 11 logically coherent steps, each with clearly defined operating methods, working principles, and control parameters, forming a standardized operating procedure that is easy for operators to learn and master, reducing operational difficulty and human error. At the same time, it requires the standardized archiving of all relevant data during the testing process for a period of no less than 5 years, achieving full traceability of the testing process. In the event of quality disputes or testing errors, the cause of the problem can be quickly traced, ensuring the authority and credibility of the test results.

[0014] 6. Highly practical and driving industrial development, this method can not only accurately detect the content of magnetic substances in lithium carbonate, but also provide targeted rectification suggestions for unqualified samples, helping manufacturers optimize production processes, reduce the introduction of magnetic impurities, improve the quality of lithium carbonate products, and promote the high-quality development of downstream industries such as lithium-ion batteries, pharmaceuticals, and ceramics. At the same time, the promotion and application of this method can standardize the quality control system of the lithium carbonate industry, fill the technical gap in the field of trace magnetic substance detection in the industry, and enhance the core competitiveness of my country's lithium carbonate industry.

[0015] 7. The method exhibits good stability and high repeatability. Parallel experiments have verified that the method can obtain stable and accurate detection results under different experimental conditions and different batches of operations, avoiding the problem of poor repeatability of existing methods. It can meet the needs of routine laboratory testing and long-term stable industrial testing, providing strong support for the promotion of the method. Attached Figure Description

[0016] The accompanying drawings, which are included to provide a further understanding of this application and form part of this application, use the same reference numerals to denote the same or similar parts. The illustrative embodiments of this application and their descriptions are used to explain this application and do not constitute an undue limitation of this application. In the drawings: Figure 1 A schematic flowchart of a method for detecting the content of magnetic materials in lithium carbonate according to an embodiment of this application is shown. Detailed Implementation

[0017] To make the objectives, technical solutions and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and specific embodiments.

[0018] In the following description, references to "an embodiment," "an embodiment," "an example," "example," etc., indicate that the described embodiment or example may include a particular feature, structure, characteristic, property, element, or limitation, but not every embodiment or example necessarily includes that particular feature, structure, characteristic, property, element, or limitation. Furthermore, the repeated use of the phrase "an embodiment according to this application," while possibly referring to the same embodiment, does not necessarily refer to the same embodiment.

[0019] For simplicity, certain technical features known to those skilled in the art are omitted in the following description.

[0020] Example 1: According to one embodiment of this application, this method addresses the shortcomings of existing technologies by combining gradient magnetic field separation and enrichment, ultrasound-assisted purification, precise ICP-MS detection, and multivariate statistical error correction. The method comprises 11 steps, each logically connected and interconnected. Each step provides a necessary prerequisite for the next, and the next step extends the previous one, ensuring the scientific rigor, accuracy, and efficiency of the detection process. The specific steps are as follows: Step 1: Sample Collection and Preprocessing The core purpose of this step is to collect representative lithium carbonate samples and remove physical impurities (such as dust and large particles) from the samples through pretreatment to avoid interference with subsequent magnetic separation and detection. This also ensures sample homogeneity, laying the foundation for accurate weighing and detection. The operating method is as follows: Using a sterile sampler, samples are collected from the upper, middle, and lower parts of the lithium carbonate finished product silo according to the requirements of GB / T6678-2003 "General Rules for Sampling of Chemical Products". Three samples are collected from each part, with each sample weighing no less than 50g. During sampling, direct contact between the sampler and metal equipment should be avoided to prevent the introduction of external magnetic impurities. The nine collected samples are combined and placed in a clean agate mortar (agate is non-magnetic, avoiding interference). Manual grinding is performed, controlling the grinding force at 50-80N and the grinding time at 15-20 minutes, until all samples pass through a 200-mesh nylon sieve (nylon sieves are non-magnetic). To prevent the introduction of magnetic impurities through the metal sieve, the sieved sample is placed in a vacuum drying oven at a temperature of 105±2℃ and a vacuum of 0.08-0.10MPa for 2 hours. This is to remove adsorbed and free water from the sample and prevent moisture from affecting the subsequent magnetic separation efficiency and the accuracy of the detection results. After drying, the sample is removed and placed in a desiccator to cool to room temperature (cooling time is about 30 minutes to prevent moisture absorption). After cooling, the sample is reduced to 10g using the quartering method to obtain a uniform sample for testing. This sample is then placed in a sterile sealed bag for later use. The sample is numbered and recorded (including sampling time, sampling site, pretreatment time, etc.).

[0021] Working Principle: Sample collection employs a multi-point mixed sampling method, effectively avoiding the problem of insufficient sample representativeness caused by sampling from a single location. This ensures that the collected sample accurately reflects the magnetic material content of the entire batch of lithium carbonate. The selection of an agate mortar and nylon sieve completely avoids interference from magnetic impurities introduced by metal materials, ensuring the cleanliness of the pretreatment process. Vacuum drying quickly removes moisture from the sample and avoids high-temperature decomposition of lithium carbonate (the decomposition temperature of lithium carbonate is 723℃; drying at 105℃ will not cause sample decomposition), while also reducing the impact of moisture on the magnetic field strength during subsequent magnetic separation. Quartering reduces the sample volume while maintaining sample representativeness, improving subsequent operational efficiency and avoiding sample waste. This step provides a uniform, clean, and interference-free sample for accurate weighing in step 2. Incomplete sample pretreatment can lead to increased weighing errors, and physical impurities can interfere with the magnetic separation process, affecting the separation effect of magnetic materials.

[0022] Step 2: Accurately weigh and dilute the sample The core purpose of this step is to accurately weigh a certain mass of the sample to be tested and dissolve it in a specific solvent to prepare a homogeneous sample mother liquor. This provides a quantitative sample system for subsequent magnetic separation and enrichment, ensuring the repeatability and comparability of the test results. The operation method is as follows: Take out the sample to be tested after pretreatment in step 1 and weigh it using a 0.0001g electronic balance (accuracy 0.0001g, metrologically calibrated). Before weighing, preheat the balance for 30 minutes and calibrate the zero point of the balance to avoid the influence of ambient temperature (controlled at 25±2℃) and humidity (controlled at 50±5%RH) on the weighing accuracy of the balance. Use the difference method to weigh the sample. First, weigh the clean 50mL volumetric flask (recorded as m0). Then, use an agate spatula to take an appropriate amount of sample and put it into the volumetric flask. Weigh the total mass of the volumetric flask and the sample again (recorded as m1). The sample mass m = m1 - m0. Control the sample weighing mass to be 2.0000±0.0002g. Perform three parallel weighings to ensure that the weighing error is ≤0.0g. 1%; Slowly add 20 mL of hydrochloric acid solution (hydrochloric acid concentration 1+1, prepared by mixing analytical grade hydrochloric acid and ultrapure water at a volume ratio of 1:1, ultrapure water resistivity ≥18.2 MΩ·cm, to avoid interference from impurities in the water) to the volumetric flask containing the weighed sample. Gently shake the volumetric flask while adding the solution to ensure complete dissolution of the sample and avoid incomplete dissolution due to excessively high local concentrations. After the sample is completely dissolved (observe that there is no white precipitate or turbidity in the volumetric flask), rinse the inner wall of the volumetric flask 3-4 times with ultrapure water, ensuring that all the rinsing solution flows into the volumetric flask. Then, dilute the solution to the mark with ultrapure water, shake well, and let it stand for 10 minutes to obtain the sample mother solution, which is labeled as mother solution A. Prepare three parallel portions of mother solution A, labeled A1, A2, and A3, for later use.

[0023] Working Principle: The core of the precise weighing of the 0.00001 g electronic balance is the use of the electromagnetic force balance principle. By calibrating the zero point and controlling environmental conditions, systematic and random errors in the weighing process are reduced, ensuring the accuracy of sample mass. The sample mass is controlled at around 2.0000 g to balance the efficiency of subsequent magnetic separation and detection sensitivity. If the sample amount is too small, the magnetic material content will be too low, resulting in a weak detection signal. If the sample amount is too large, it will increase matrix interference and the difficulty of magnetic separation. Hydrochloric acid solution is chosen because lithium carbonate is easily soluble in hydrochloric acid. The reaction equation is Li2CO3 + 2HCl = 2LiCl + H2O + CO2↑. The generated LiCl is easily soluble in water and will not produce precipitation, ensuring complete sample dissolution. The use of analytical grade hydrochloric acid and ultrapure water avoids interference from magnetic impurities and other ions in the reagents. The volume adjustment operation ensures that the concentration of the sample stock solution is uniform. The preparation of 3 parallel stock solutions provides a basis for subsequent parallel experiments and error analysis, ensuring the repeatability of the detection results. The completion of this step provides a quantitative, uniform, and completely dissolved sample system for the magnetic separation and enrichment in step 3. If the weighing is inaccurate or the sample is not completely dissolved, it will directly lead to incomplete separation of magnetic materials and excessive deviation in the detection results.

[0024] Step 3: Gradient magnetic field magnetic separation and enrichment The core objective is to utilize the differences in magnetic strength among different magnetic materials to completely separate and enrich the magnetic materials (strong and weak magnetic) in the sample mother liquor using a gradient magnetic field. This solves the problem that existing single magnetic field strengths cannot completely separate different magnetic impurities, improves the enrichment efficiency and separation purity of magnetic materials, and lays the foundation for lowering the detection limit and improving the detection accuracy in subsequent detection processes. The operating method is as follows: A self-designed gradient magnetic field separation device is selected (this device includes a magnetic field generator, separation column, collection tank, and temperature control system; the magnetic field strength can be continuously adjusted within a range of 0.1-1.5T). First, the separation device is cleaned and calibrated. The inner wall of the separation column is rinsed 3-4 times with ultrapure water to remove residual impurities. Then, the magnetic field strength is calibrated to ensure that the error is ≤ ±0.01T. The mother liquor A (A1, A2, A3) prepared in step 2 is poured into three clean separation columns, with 20mL of mother liquor A added to each column. The flow rate of the separation column is controlled at 1.0-1.5mL / min (the flow rate is adjusted by a peristaltic pump with a speed of 50-80r / min). The magnetic field generator is turned on, and the gradient magnetic field parameters are set: the initial magnetic field strength is 0.1T, and then it is uniformly increased to 1T at a rate of 0.05T / min. A 1.5T magnetic field strength was maintained for 5 minutes, then decreased to 0.1T at a rate of 0.1T / min to complete one gradient magnetic field cycle. A total of three gradient magnetic field cycles were performed to ensure that impurities with different magnetic intensities (strongly magnetic elements such as iron and cobalt, and weakly magnetic elements such as Fe2O3 and NiO) could be adsorbed by the magnetic field. During the magnetic field cycle, the temperature of the separation column was controlled at 25±1℃ using a temperature control system to avoid excessively high or low temperatures affecting the magnetic field strength and the adsorption effect of magnetic substances. After magnetic separation was completed, the magnetic field generator was turned off, and the column was allowed to stand for 5 minutes to allow the magnetic substances adsorbed on the inner wall of the separation column to completely detach and flow into the collection tank, resulting in a magnetic substance enrichment solution, labeled as enrichment solution B. Three parallel enrichment solutions B (B1, B2, and B3) were obtained. The remaining mother liquor in the separation column was collected and labeled as waste liquid for later use (for subsequent matrix interference analysis).

[0025] Working Principle: The core working principle of the gradient magnetic field is to selectively adsorb and separate substances with different magnetic susceptibility by utilizing the continuous change in magnetic field strength. The greater the magnetic susceptibility of a magnetic substance, the lower the magnetic field strength required for adsorption, and vice versa. Magnetic substances in lithium carbonate mainly include strongly magnetic substances (magnetic susceptibility χ ≥ 10). - ³) and weakly magnetic materials (magnetic susceptibility χ=10) -6 -10 -³), a single magnetic field strength can only adsorb substances with a certain magnetic strength, and cannot achieve comprehensive separation. However, the gradient magnetic field, increasing from 0.1T to 1.5T, can sequentially adsorb weakly magnetic, moderately magnetic, and strongly magnetic substances. Three magnetic field cycles can further improve the adsorption efficiency, ensuring that the separation and recovery rate of magnetic substances is ≥99%. The temperature control system controls the separation column temperature at 25℃ because temperature affects the magnetic permeability of the magnetic field. Too high a temperature will cause the magnetic field strength to decrease, resulting in poor adsorption of magnetic substances. Too low a temperature will cause the solution viscosity to increase, affecting the shedding and collection of magnetic substances. The peristaltic pump controls the flow rate at 1.0-1.5mL / min, which can ensure that the magnetic substances in the mother liquor have enough time to contact the magnetic field and avoid the magnetic substances flowing out of the separation column before being adsorbed due to the flow rate being too fast. The separation column is made of quartz material (non-magnetic and with good chemical stability), which can avoid the introduction of magnetic impurities and chemical interference by the separation column itself. The completion of this step provides a high-concentration, high-purity magnetic material enrichment solution for the washing of the enrichment solution in step 4. If the magnetic separation is not thorough, some magnetic material will remain in the waste liquid, resulting in low detection results that cannot meet the needs of trace detection.

[0026] Step 4: Ultrasonic-assisted washing and purification of the enrichment solution The core objective of this step is to remove lithium carbonate matrix impurities (such as Li) from enrichment solution B. + Cl -To remove other non-magnetic impurities, this process avoids matrix interference with ICP-MS detection in subsequent steps, improves the purity of the enrichment solution, ensures the accuracy of the detection signal, and lays the foundation for the dissolution of the enriched substance in step 5. The operating method is as follows: Pour the three enrichment solutions B (B1, B2, B3) obtained in step 3 into three 50mL centrifuge tubes respectively. Add 10mL of ultrapure water (resistivity ≥18.2MΩ·cm) to each centrifuge tube and gently shake the centrifuge tubes for 1 minute to fully mix the enrichment solution with the ultrapure water. Place the centrifuge tubes in an ultrasonic cleaner, set the ultrasonic power to 200-250W, the ultrasonic frequency to 40kHz, and the ultrasonic time to 15 minutes. During the ultrasonic process, control the water temperature of the ultrasonic cleaner to 25±2℃ to avoid excessively high water temperature causing oxidation of magnetic materials (such as Fe elemental oxidation to Fe2O3), which would affect subsequent detection results. After ultrasonication, place the centrifuge tubes in a high-speed centrifuge, set the centrifugation speed to 8000r / min, and the centrifugation time to 10 minutes. At 25℃, magnetic substances precipitate at the bottom of the centrifuge tube, while non-magnetic impurities and matrix impurities dissolve in the supernatant. After centrifugation, the supernatant is slowly aspirated with a pipette (avoiding contact with the bottom precipitate) and poured into a waste liquid collection bottle, retaining the magnetic substance precipitate at the bottom of the centrifuge tube. 10 mL of ultrapure water is added to each centrifuge tube again, and the above ultrasonic, centrifugation, and supernatant aspiration operations are repeated three times until the supernatant is colorless and transparent (the absorbance of the supernatant in the 200-800 nm wavelength range is ≤0.001 as measured by a UV-Vis spectrophotometer, indicating that matrix impurities and non-magnetic impurities have been completely removed). After washing, the magnetic substance precipitate at the bottom of the centrifuge tube is retained and labeled as precipitate C. Three parallel precipitates C (C1, C2, and C3) are obtained for later use.

[0027] Working Principle: The core principle of ultrasonic-assisted cleaning is the cavitation effect of ultrasound. When ultrasound propagates in a liquid, it generates a large number of tiny bubbles. These bubbles expand and burst rapidly, producing a strong impact force that can remove matrix impurities (Li) adsorbed on the surface of magnetic materials. + Cl -Ultrasonic washing thoroughly removes both magnetic and non-magnetic impurities. Compared to traditional manual washing, ultrasonic washing is more efficient and thorough, reducing the number of washes and preventing the loss of magnetic materials during the washing process. Controlling the ultrasonic power and frequency is crucial. Too low a power (<200W) will not generate sufficient impact force, resulting in incomplete impurity removal; too high a power (>250W) will cause magnetic particles to break, affecting subsequent dissolution. A 40kHz ultrasonic frequency optimizes the cavitation effect, balancing washing effectiveness and the stability of the magnetic materials. High-speed centrifugation utilizes centrifugal force to precipitate denser magnetic materials at the bottom of the centrifuge tube, while less dense impurities dissolve in the supernatant. A centrifugation speed of 8000 rpm and a centrifugation time of 10 minutes ensure complete precipitation of magnetic materials, preventing residual magnetic materials in the supernatant and thus avoiding detection errors. Multiple washes, verified using a UV-Vis spectrophotometer, ensure complete removal of matrix and non-magnetic impurities, because Li... + Cl - Matrix impurities exhibit weak absorption in the UV-Vis region; an absorbance ≤0.001 indicates that the impurity content is below the detection limit and will not interfere with subsequent ICP-MS detection. This step provides a high-purity magnetic material precipitate for the enrichment dissolution in step 5. Incomplete washing can cause matrix impurities to suppress or enhance signals during ICP-MS detection, leading to significant deviations in the results.

[0028] Step 5: Precipitation, dissolution, and volume adjustment of magnetic substances The core purpose of this step is to completely dissolve the high-purity magnetic material precipitate C in a specific solvent to prepare a homogeneous detection mother solution, thereby converting the magnetic material into an ionic state, which facilitates subsequent ICP-MS detection, ensures the stability and accuracy of the detection signal, and lays the foundation for the ICP-MS detection in step 6. The procedure is as follows: Add 5 mL of nitric acid solution (5% nitric acid concentration, prepared by mixing analytical grade nitric acid and ultrapure water at a volume ratio of 1:19; analytical grade nitric acid avoids introducing impurities, and low concentration ensures complete dissolution of the magnetic material while avoiding matrix interference in subsequent ICP-MS detection due to excessive nitric acid) to the centrifuge tubes corresponding to the three precipitates C (C1, C2, C3) obtained in step 4. Place the centrifuge tubes in a constant temperature water bath, set the water bath temperature to 60±2℃, and the water bath time to 20 minutes. During the water bath, gently shake the centrifuge tubes once every 5 minutes to promote the dissolution of the magnetic material precipitate. Observe the dissolution of the precipitate until it is completely dissolved (no black or brown precipitate residue). If the precipitate is not completely dissolved, add analytical grade nitric acid dropwise (0.1 mL each time) until the precipitate is completely dissolved and record the dissolution rate. Total amount of acid added (for subsequent error correction); After the precipitate is completely dissolved, remove the centrifuge tube from the constant temperature water bath and place it in a fume hood to cool to room temperature (cooling time is about 20 minutes to avoid excessive temperature causing solution evaporation and affecting concentration accuracy); Slowly transfer the cooled solution to a 25mL volumetric flask, rinse the inner wall of the centrifuge tube with ultrapure water 3-4 times, ensuring that all rinsing solution flows into the volumetric flask to ensure no magnetic material residue; Dilute the solution in the volumetric flask to the mark with ultrapure water, shake well and let stand for 5 minutes to obtain the magnetic material detection mother solution, labeled as detection mother solution D. Prepare 3 parallel detection mother solutions D (D1, D2, D3) for later use. At the same time, prepare a blank solution: Take a 25mL volumetric flask, add 5mL of nitric acid solution of the same concentration, dilute to the mark with ultrapure water, and label it as blank solution D0 for subsequent blank correction.

[0029] Working principle: Magnetic material precipitation mainly includes elemental metals such as iron, cobalt, and nickel, as well as their oxides. These substances can react with nitric acid to form water-soluble nitrates, such as Fe + 4HNO3 (dilute) = Fe(NO3)3 + NO↑ + 2H2O, Fe2O3 + 6HNO3 = 2Fe(NO3)3 + 3H2O, and CoO + 2HNO3 = Co(NO3)2 + H2O. Low-concentration (5%) analytical grade nitric acid ensures complete dissolution of the magnetic material while avoiding the matrix effect caused by excessive nitric acid (excessive nitric acid will cause ion inhibition in ICP-MS detection, affecting the detection signal intensity). Heating to 60℃ in a constant temperature water bath can accelerate the dissolution reaction rate. This process shortens the dissolution time and avoids the volatilization of nitric acid due to high temperatures (>70℃), which would increase the solution concentration and affect the detection results. Shaking the centrifuge tube ensures that the precipitate and nitric acid are in full contact, preventing incomplete dissolution due to excessively high local concentrations. Rinsing the inner wall of the centrifuge tube with ultrapure water ensures that there are no magnetic material residues, preventing low detection results due to residues. Adjusting the volume to 25 mL controls the concentration of magnetic material ions within the optimal detection range of ICP-MS (0.1-100 ppb), ensuring the stability and accuracy of the detection signal. The preparation of a blank solution eliminates trace magnetic impurities that may be present in the nitric acid reagent and ultrapure water, providing a basis for subsequent blank correction and reducing errors caused by reagent interference. The completion of this step provides a uniform, stable, and appropriately concentrated detection stock solution for the ICP-MS detection in step 6. If the precipitate is not completely dissolved, the magnetic material will not be fully converted into an ionic state, and ICP-MS will not be able to detect the undissolved magnetic material, resulting in low detection results.

[0030] Step 6: ICP-MS Instrument Calibration and Parameter Optimization The core purpose of this step is to calibrate and optimize the parameters of the ICP-MS instrument to ensure that the instrument is in optimal working condition, improve the sensitivity and accuracy of detection, reduce systematic errors caused by the instrument itself, and lay the foundation for the determination of the mother liquor in step 7.The operating procedure is as follows: An inductively coupled plasma mass spectrometer (ICP-MS, Agilent 7900) was selected. Before detection, the instrument was preheated for 60 minutes. During preheating, argon gas (purity ≥99.999%) was turned on, with the flow rate controlled at 15 L / min, ensuring no air residue remained in the instrument's internal tubing. After preheating, the instrument was calibrated using the external standard method. A series of standard solutions with varying concentration gradients were prepared: iron, cobalt, and nickel standard stock solutions (concentration 1000 μg / mL, analytical grade) were accurately pipetted and diluted with 5% nitric acid solution to prepare standard solutions with concentrations of 0.1 ppb, 1 ppb, 10 ppb, 50 ppb, and 100 ppb. For each concentration gradient, three parallel samples were prepared for the standard series solutions, and a blank standard solution (5% nitric acid solution) was also prepared. The standard series solutions and blank standard solutions were sequentially imported into the ICP-MS instrument, and measurements were performed according to the set parameters. The signal intensity (count rate, cps) corresponding to each concentration of standard solution was recorded. A standard working curve was plotted with standard solution concentration (ppb) on the x-axis and signal intensity (cps) on the y-axis. The correlation coefficient R² of the standard working curve was calculated. R² ≥ 0.9999 was required. If R² < 0.9999, the standard series solutions needed to be prepared again until the requirement was met. After the standard curve calibration was completed, the instrument parameters were optimized, including the injection... The optimization ranges and optimal values ​​for parameters such as RF power, nebulizer gas flow rate, auxiliary gas flow rate, sampling depth, and integration time are as follows: RF power optimization range: 1200-1500W, optimal value: 1350W (too low power will lead to plasma instability and weak signal strength; too high power will lead to enhanced matrix interference); Nebulizer gas flow rate optimization range: 0.8-1.2L / min, optimal value: 1.0L / min (too low atomizer gas flow rate will lead to incomplete atomization and unstable signal strength; too high a flow rate will lead to sample dilution and decreased sensitivity); Auxiliary gas flow rate optimization range: 0.2-0.5L / min, optimal value: 0.3L / min (auxiliary gas can stabilize...). (Improper flow rate can lead to plasma extinguishing); the optimal sampling depth range is 8-12 mm, with an optimal value of 10 mm (sampling depth affects signal strength and interference levels; too deep or too shallow a sampling depth will increase interference); the optimal integration time range is 0.1-0.5 s, with an optimal value of 0.3 s (too short an integration time will cause large signal fluctuations, while too long an integration time will reduce detection efficiency); after parameter optimization, the signal intensity of the standard series solutions should be measured again to verify the stability of the instrument. The relative standard deviation (RSD) should be ≤2%. If the RSD > 2%, the parameters need to be re-optimized until the requirement is met; after instrument calibration and parameter optimization, the parameter settings should be saved for later use.

[0031] Working Principle: The core working principle of ICP-MS is to use inductively coupled plasma (ICP) as an ion source. The sample solution is atomized and introduced into the plasma. Atoms in the sample are ionized into ions in the high-temperature (approximately 7000K) plasma. These ions are separated by a mass analyzer (quadrupole). Qualitative analysis is performed based on the mass-to-charge ratio (m / z), and quantitative analysis is performed based on the ion signal intensity (count rate). The external standard calibration principle is based on the linear relationship between signal intensity and substance concentration. By preparing a series of standard solutions with concentration gradients, a standard working curve is plotted. Using the standard curve, the signal intensity of the detection stock solution can be converted into the corresponding concentration. A correlation coefficient R² ≥ 0.9999 indicates a good linear relationship in the standard curve, ensuring the accuracy of quantitative analysis. Optimization of instrument parameters is also crucial. This is crucial for improving detection sensitivity and stability. Radio frequency power determines the plasma temperature and ionization efficiency, nebulizer gas flow rate determines sample nebulization efficiency and injection volume, auxiliary gas flow rate determines plasma stability, sampling depth determines ion extraction efficiency, and integration time determines signal acquisition accuracy. The optimal combination of these parameters ensures the instrument operates at its best, reducing issues such as ion suppression and spectral interference. Argon is chosen because of its stable chemical properties and low ionization energy, which allows for the formation of stable plasma without reacting with sample ions, thus avoiding interference. Instrument preheating ensures all components reach a stable operating temperature, reducing systematic errors caused by temperature fluctuations. Parallel determination of standard solutions verifies instrument repeatability; RSD ≤ 2% indicates good instrument stability and suitability for subsequent detection. This step provides a calibrated and optimized ICP-MS instrument for the mother liquor determination in step 7. Uncalibrated instruments or unoptimized parameters will result in inaccurate detection signals, poor repeatability, and unreliable detection results.

[0032] Step 7: Detection of mother liquor and correction for multivariate statistical errors The core objective is to use an optimized ICP-MS instrument to determine the signal intensity of magnetic substances (iron, cobalt, nickel) in the detection stock solution, and to use multivariate statistical methods to correct errors in the detection results, eliminating systematic and random errors in each step of the detection process, thereby improving the accuracy and reliability of the detection results and providing a basis for blank experiment correction in step 8. The operation method is as follows: The three detection stock solutions D (D1, D2, D3) prepared in step 5 and the blank solution D0 are sequentially imported into the ICP-MS instrument for measurement according to the optimized instrument parameters in step 6. Each sample was measured in triplicate, and the signal intensity (count rate, cps) was recorded for each measurement. The average signal intensity of the test stock solutions D1, D2, and D3 (denoted as S1, S2, and S3) and the average signal intensity of the blank solution D0 (denoted as S0) were obtained respectively. Based on the standard working curve plotted in step 6, the average signal intensity (S1, S2, S3, and S0) were converted into corresponding concentration values ​​to obtain the preliminary detection concentration of the test stock solution (denoted as C1, C2, and C3) and the concentration value of the blank solution (denoted as C0). The formula for calculating the preliminary detection concentration is: C i =k×S i +b (i=1,2,3,0), where k is the slope of the standard working curve and b is the intercept of the standard working curve; a multivariate linear regression combined with a deviation correction model is used to correct errors, eliminating systematic errors (such as instrument drift, reagent interference, matrix interference) and random errors (such as weighing error, operational error) in the detection process. A multivariate statistical error correction model is established, and the specific formula is as follows: The meanings of each symbol are as follows: The final detected concentration (ppb) of the magnetic material after error correction is the core output value of this step; The regression constant term represents the basic bias in the detection process (caused by inherent instrument errors, reagent blanks, etc.), and is obtained through partial least squares regression (PLS) combined with Bayesian regularization fitting. The fitting process incorporates a regularization parameter. This is used to avoid model overfitting; The concentration regression coefficient represents the weight of the initial detected concentration on the final detected concentration. This indicates a positive correlation between the preliminary detection concentration and the final detection concentration. The larger the weight, the more significant the influence of the preliminary detection concentration. The solution formula is as follows: ( (This is the average value of the preliminary detection concentration). The regression coefficient for weighing error represents the weight of the influence of the sample weighing error in step 2 on the final detected concentration. The sign of the weighing error is determined by the direction of the weighing error (when the weighing value is too large). When the weighing value is too small, The solution requires combining the normal distribution probability density function of the weighing error and introducing a correction factor. ; The regression coefficient for magnetic separation recovery rate represents the weight of the impact of the magnetic separation recovery rate in step 3 on the final detected concentration. The higher the magnetic separation recovery rate, the closer the final detected concentration is to the true value. Avogadro's constant is introduced during the solution process. As a correction term for trace substance measurement; , where is the matrix interference regression coefficient, representing the weight of the influence of residual matrix impurities after washing in step 4 on the final detection concentration. The sign is determined by the type of matrix interference (in the case of ion enhancement effect). During ion suppression effect, Planck's constant is introduced when solving the problem. As a correction term for microscopic particle interference; To determine the preliminary concentration (ppb) of the mother liquor; The sample weighing error (%) is calculated using the following formula: (m is the actual weighed mass, m0 is the standard weighed mass of 2.0000g,) (This is the composite correction coefficient for weighing error, which conforms to the superposition characteristics of the normal distribution and the log-normal distribution of precision weighing). The magnetic separation recovery rate (%) is calculated using the following formula: ; The mass of the magnetic material enriched in step 3. The true mass of the magnetic material in the sample was determined by a combination of spiked recovery experiments and isotope labeling. e is the natural constant, approximately 2.71828. (This is a temperature correction term for recovery rate, used to compensate for the effects of temperature fluctuations during the magnetic separation process). The residual matrix impurities (ppb) were determined by UV-Vis spectrophotometry in the supernatant after washing in step 4, using a matrix impurity residual concentration (ppb). + The concentration was obtained, and the corrected formula is: ; These are measured values, and γ is the Euler-Macheroni constant, approximately 0.5772. This represents the standard deviation of matrix impurity detection. (For matrix interference random correction term); The standard deviation of the weighing error is obtained by combining parallel weighing experiments with Bayesian estimation and fitting, and follows a normal distribution. ( The mean of the standard deviation, (Standard deviation and variance) To improve the cumulative distribution function (CDF) of the normal distribution, a logarithmic correction term is introduced. The nonlinear cumulative effect of quantitative weighing error demonstrates the expertise of advanced statistical error correction. This integral can be numerically solved using the Gauss-Legendary integral formula: ( For the roots of the Legendre polynomial, (weighting coefficients) It is a higher-order Gaussian correction term that combines three constants, π, e, and γ, to suppress the interference of abnormal weighing errors on the final result, while compensating for the systematic deviation caused by environmental humidity. The logarithmic-constant composite correction term is used to compensate for the nonlinear influence of magnetic separation recovery rate, while the Avogadro constant correction term is used to compensate for the deviation in the counting of trace magnetic particles. As a higher-order correction term for matrix interference, the influence weight of matrix interference is adjusted by Euler-Marcheroni constant, Planck constant and quadratic radical, while compensating for the detection bias caused by microscopic ion collisions; The random error term follows a normal distribution. ( (where is the variance of random errors), representing unpredictable accidental errors during the detection process (such as fluctuations in ambient temperature, momentary instrument drift, etc.), and satisfying the following conditions: n is the number of parallel measurements. The average value of random error. (variance correction factor) The bias weighting factor for the i-th parallel determination is calculated using the following formula: Used to adjust the weight of the influence of random error. This is used as a weighting term for random errors, further improving the accuracy of error correction.

[0033] The establishment and solution process of the error correction model is as follows: Fifteen groups of spiked samples with different concentrations (0.1-100 ppb, covering three concentration ranges: low, medium, and high, with five groups for each range) were prepared through a spike recovery experiment. The samples were then analyzed according to steps 1-7 of this method to obtain the preliminary detection concentration for each spiked sample. Weighing error Magnetic separation recovery rate Residual amount of matrix impurities And the true concentration of the spiked sample (denoted as ). The true concentration was determined by isotope dilution mass spectrometry (ID-MS) to ensure the accuracy of the true value; As the dependent variable, with , , , Using partial least squares regression (PLS) combined with Bayesian regularization as the independent variable, the multiple linear regression equation was fitted, and a regularization parameter was introduced. Weighting coefficient ( Let be the standard deviation of the detection of the i-th spiked sample. (This is a weight decay factor, used to reduce the impact of outliers on the fitting results). Solving for this factor yields the regression constant term. and regression coefficients , , , The fitting process satisfies (regularization term) (Used to avoid model overfitting and improve the model's generalization ability); and simultaneously calculates the model's coefficient of determination. Adjusting the coefficient of determination Root mean square error (RMSE) and mean absolute error (MAE), among which , (k is the number of independent variables, k=4) (Introducing three constants, π, e, and γ, for accuracy correction) ,Require , RMSE ≤ 0.03ppb and MAE ≤ 0.02ppb are used to ensure the good fit and prediction accuracy of the error correction model. Furthermore, the significance of the model is verified through the F-test; the formula for calculating the F-statistic is... The requirement is F≥10000 (significance level α=0.01), proving the model is statistically significant; the results obtained in step 7... , , , Substituting into the error correction model, the final detected concentration of the magnetic material after error correction is calculated. , , (Corresponding to 3 sample mother solutions D1, D2, and D3 respectively); calculate the average value of the 3 final test concentrations (denoted as ). ) and relative standard deviation (RSD), where (Introducing γ, π, and e to correct for minor deviations and compensate for systematic bias in parallel measurements). (Introducing e, γ, and π to optimize the quantification accuracy of deviation, in line with the statistical requirements of precision testing), requiring RSD ≤ 1.5%. If RSD > 1.5%, the test mother liquor needs to be re-measured, and the error correction model parameters need to be checked until the requirements are met.

[0034] Working Principle: The core principle of ICP-MS signal intensity measurement is that sample ions are separated according to their mass-to-charge ratio in the mass analyzer. Ion collision detectors generate electrical signals, and the signal intensity is linearly related to the ion concentration. A standard working curve can be used to convert the signal intensity into a concentration value. Blank solution measurement eliminates interference from reagents and trace magnetic impurities in ultrapure water, providing preliminary concentration data for subsequent error correction. The core principle of multivariate statistical error correction is to use a multivariate linear regression model to quantify the impact of various error sources (weighing error, magnetic separation recovery rate, matrix interference) on the detection results. Regression coefficients assign corresponding weights to each error source, achieving precise error correction. Compared to the simple blank correction of existing methods, this method can eliminate more types of errors and improve the accuracy of detection results. The core principle of the least squares method is to solve the parameters of the regression equation by minimizing the sum of squared residuals between the observed and fitted values, ensuring that the regression model fits the true data to the greatest extent possible. The coefficient of determination... Used to evaluate the goodness of fit of the model. The closer the value is to 1, the better the model fits. The root mean square error (RMSE) is used to evaluate the model's prediction accuracy; the smaller the RMSE, the smaller the prediction error. The random error term... Following a normal distribution, random errors can be estimated and controlled using statistical methods, further improving the reliability of the test results. Parallel determination of three sample stock solutions and calculation of RSD verifies the repeatability of the test results; RSD ≤ 2% indicates a stable testing process and effective error correction. This step provides the error-corrected test results for the blank experiment correction in step 8. Without multivariate statistical error correction, the test results would be significantly affected by various error sources, failing to meet the requirements of high-precision testing.

[0035] Step 8: Blank Experiment Correction The core purpose of this step is to further correct the test results through blank experiments, eliminate interference from trace magnetic impurities that may exist in reagents, ultrapure water, instrument tubing, etc., ensure the authenticity and accuracy of the test results, and provide a foundation for the parallel experimental verification in step 9. The operating method is as follows: Prepare 5 blank solutions (labeled as D) in parallel according to the preparation method in step 5. 01 D 02 D 03 D 04 D 05The preparation process of the blank solution was exactly the same as that of the test stock solution, but without adding magnetic precipitate C. Only 5 mL of 5% nitric acid solution and ultrapure water were added to ensure that the preparation environment, reagents, and instruments of the blank solution were exactly the same as those of the test stock solution. Five blank solutions were imported into the ICP-MS instrument for measurement according to the instrument parameters optimized in step 6. Each blank solution was measured in parallel three times, and the signal intensity (cps) of each measurement was recorded. The average signal intensity of each blank solution was calculated (denoted as S). 01 S 02 S 03 S 04 S 05 Based on the standard working curve plotted in step 6, the average signal intensity of the blank solution is converted into the corresponding concentration value (denoted as C). 01 C 02 C 03 C 04 C 05 Five blank solutions were obtained; outliers in the blank solution concentration values ​​were removed using the Grubbs' test (significance level α = 0.05), and the average concentration value of the remaining blank solutions was calculated (denoted as ). ) and standard deviation (denoted as ),Require ≤0.05ppb (below 1 / 2 of the detection limit), if If the concentration is >0.05 ppb, the reagents and ultrapure water need to be replaced, a blank solution needs to be prepared again, and the measurement needs to be repeated until the requirement is met; the average detection concentration obtained in step 7 is then used. Subtract the average concentration of the blank solution The concentration of the magnetic material after blank correction (denoted as ) is obtained. The correction formula is: ;like A value <0 indicates that the concentration of the blank solution is higher than that of the test stock solution. The test process needs to be rechecked, and the measurement should be repeated after eliminating interference.

[0036] Working principle: The core principle of the blank experiment is "parallel control." By preparing a blank solution that is identical to the test stock solution (except for the absence of the analyte), the concentration of magnetic substances in the blank solution is measured. This quantifies the interference from trace magnetic impurities introduced by reagents, ultrapure water, instrument tubing, etc., because these interferences are identical in the blank solution and the test stock solution, and can be eliminated by subtraction. The core principle of Grubbs' criterion is based on the normal distribution. By calculating the deviation of the suspected value from the mean, it is determined whether the value is an outlier (caused by random error). Eliminating outliers ensures the average concentration of the blank solution. The authenticity and reliability of the data were assessed, with a significance level of α=0.05, indicating a 95% confidence level for outlier removal and preventing misjudgment; the average concentration of the blank solution was also assessed. ≤0.05ppb ensures that the interference level is below half of the detection limit and will not significantly affect the detection results. If the value is too high, it indicates the presence of impurities in the reagent or ultrapure water, requiring replacement of the reagent and ultrapure water to avoid interference. The core of the blank correction formula is to eliminate system interference, because the error correction in step 7 mainly eliminates errors in weighing, magnetic separation, and matrix processes, while the blank experiment mainly eliminates interference from reagents and instrument tubing. Combining the two can further improve the accuracy of the test results. A value <0 indicates severe interference during the detection process (such as contamination of the blank solution). The detection process must be re-examined, interference eliminated, and the measurement repeated to ensure the reasonableness of the results. Completion of this step provides a precise detection concentration after blank correction for parallel experimental verification in step 9. Without blank correction, interference from reagents and instrument tubing can lead to overestimation of the detection results, making it impossible to obtain the true content of magnetic substances.

[0037] Step 9: Parallel experimental verification The core purpose of this step is to verify the repeatability and stability of the detection method through parallel experiments, ensuring that the method can obtain accurate and reliable detection results in different batches and samples, providing a basis for the result calculation and judgment in step 10, and meeting the requirements for method stability. The operation method is as follows: Following the complete procedure of steps 1-8, three batches of lithium carbonate samples were collected again (from the same batch as the samples collected in step 1 to ensure sample consistency). Three sample stock solutions were prepared in parallel for each batch, for a total of nine sample stock solutions. The nine sample stock solutions were subjected to ICP-MS determination, multivariate statistical error correction, and blank experiment correction to obtain the magnetic substance concentrations after blank correction (denoted as ). , ... ); Calculate the average of the 9 concentration values ​​(denoted as ). The following parameters were measured: Relative Standard Deviation (RSD) and Spiked Recovery Rate. The Spiked Recovery Rate was determined as follows: Iron, Cobalt, and Nickel standard solutions of known concentrations (0.5, 1.0, and 1.5 times the detection concentration, respectively) were added to the weighed sample from step 2. The detection was performed according to steps 3-8. The Spiked Recovery Rate was calculated using the formula: Spiked Recovery Rate (%) = (Detection Concentration after Spiking - Detection Concentration without Spiking) / Spiked Concentration × 100%. Each spiked concentration was measured in triplicate, and the average Spiked Recovery Rate was calculated. Validation Criteria: The RSD of parallel experiments was required to be ≤2.5%, and the Spiked Recovery Rate was required to be between 95% and 105%. If the validation criteria were not met, the cause (such as incomplete sample pretreatment, instrument parameter drift, error correction model parameter deviation, etc.) needed to be investigated, and parallel experiments were repeated until the requirements were met. All data from the parallel experiments were recorded, including detection concentration, error correction value, blank correction value, RSD, and Spiked Recovery Rate, to form a parallel experiment report.

[0038] Working Principle: The core principle of parallel experimental validation is "reproducibility validation." Under the same experimental conditions (same batch of samples, same reagents, same instruments, same operating procedures), repeated detections are performed multiple times, and the dispersion of the detection results is observed. The smaller the dispersion (lower RSD), the better the repeatability of the method. The core principle of spiked recovery is to add a standard substance of known concentration to the sample, measure the detection concentration of the spiked sample, and calculate the spiked recovery rate. This verifies the accuracy and recovery rate of the method. A spiked recovery rate between 95% and 105% indicates good accuracy of the method, no significant systematic error, and no significant loss or contamination of magnetic materials during detection. The design of 3 batches of samples, with 3 parallel samples per batch, allows for... To ensure the representativeness of the validation results and avoid validation bias caused by single batches or single parallel samples, the RSD ≤ 2.5% is set based on the requirements of high-precision detection and the detection characteristics of trace magnetic substances in lithium carbonate, ensuring good stability of the method in industrial batch detection. The spiking concentration is set to 0.5, 1.0, and 1.5 times the detection concentration to cover the optimal detection range of the method and verify its accuracy at different concentration levels. If parallel experiments do not meet the validation criteria, the cause must be investigated and the experiments repeated to ensure the reliability and applicability of the method and avoid potential defects. The generation of parallel experiment reports provides data support for the subsequent promotion and application of the method, while also meeting the requirement of method repeatability. The completion of this step provides stable and reliable detection data for the result calculation and judgment in step 10. If the parallel experiments fail validation, it indicates that the method has defects and cannot be used for actual detection.

[0039] Step 10: Result Calculation and Judgment The core purpose of this step is to calculate the actual content of magnetic materials (iron, cobalt, nickel) in the lithium carbonate sample based on the blank-corrected detection concentration and sample weighing mass. Combined with industry standards and customer requirements, the sample quality is assessed, providing a basis for method verification in step 11. The operating method is as follows: Based on the sample weighing mass in step 2 (m=2.0000g), the final volume of the detection mother liquor in step 5 (V1=25mL), the final volume of the sample mother liquor in step 2 (V2=50mL), and the volume of the magnetic separation enrichment liquid in step 3 (V3=20mL), calculate the actual content of magnetic materials in the lithium carbonate sample (expressed as mass fraction ω, in ppb). The calculation formula is: The meanings of each symbol are as follows: This represents the actual mass fraction (ppb) of magnetic material in the lithium carbonate sample. V1 is the total average concentration (ppb) after blank correction of the parallel experiments obtained in step 9; V2 is the final volume of the mother liquor detected in step 5 (25 mL); V3 is the final volume of the mother liquor of the sample in step 2 (50 mL); m is the mass of the sample weighed in step 2 (2.0000 g); V4 is the volume of the magnetic separation enrichment solution in step 3 (20 mL); 10 - ³ is a unit conversion factor (converting μg / g to ppb, 1 ppb = 1 μg / kg = 10 ppb). - (³μg / g); Pi (approximately 3.14159) is used as a precise correction factor for volume-mass conversion, to compensate for minor volume deviations during the volume determination and enrichment process, and to improve calculation accuracy. is the Euler-Marcheroni constant (approximately 0.5772), used to correct systematic deviations in the weighing and volume adjustment processes, while also adjusting the weights of each correction term; e is the natural constant (approximately 2.71828), used to compensate for concentration deviations caused by solution evaporation and adsorption. It is a constant composite correction factor, which further improves the rigor of the calculation; Avogadro's constant (approximately 6.022 × 10²³ mol) - ¹), This is a trace substance measurement correction term used to compensate for minute deviations in the counting process of magnetic material particles; its magnitude is extremely small (approximately 10). -7 (ppb) enhances the professionalism and rigor of the formula; Planck's constant (approximately 6.626 × 10⁻⁶) - ³ 4 J·s), This is a microscopic interference correction term, used to compensate for minor detection biases caused by ion collisions and excited-state particle transitions during ICP-MS detection. The volumetric temperature correction factor is calculated using the following formula: ( The coefficient of volume expansion of the solution is approximately 2.1 × 10⁻⁶. -4 / ℃, where T is the ambient temperature of the experiment (in ℃), is used to compensate for the effect of temperature fluctuations on the solution volume and ensure the accuracy of the volume parameters; For higher-order calculus correction terms, the integration variable V is the solution volume (mL), and a logarithmic correction term is introduced. The nonlinear change in magnetic substance concentration (such as concentration decay due to trace volatilization and adsorption) during the process of quantifying the mother liquor sample from enrichment volume (V3) to constant volume (V2) is solved using integration by parts combined with Gauss-Legendal numerical integration. The specific solution process is as follows: Let , ,but , According to the integral by parts formula , can be obtained The second integral is solved using the Gauss-Legend de Gauss numerical integration formula, with n=5 nodes to ensure an integration accuracy of ≤10. -8 ppb. The actual contents of three magnetic materials—iron, cobalt, and nickel (ω_Fe, ω_Co, ω_Ni)—were calculated separately. For each material, a specific combination of higher-order correction coefficients and constants was introduced: a correction term was introduced for ω_Fe calculations. The calculation formula is: A correction term is introduced during the calculation of ω_Co. The calculation formula is: A correction term is introduced during the calculation of ω_Ni. The calculation formula is: The formula for calculating the total content of the three magnetic substances is: ( This is a comprehensive correction factor for the total content, used to balance the detection biases of the three substances. (This is a micro-correction term for the total content, compensating for the mutual interference between the three magnetic materials). Based on my country's GB / T11075-2023 standard for lithium carbonate and customer requirements, sample quality is assessed: battery-grade lithium carbonate requires ω_Fe≤10ppb, ω_Co≤5ppb, ω_Ni≤5ppb, and ω_total≤20ppb; industrial-grade lithium carbonate requires ω_Fe≤50ppb, ω_Co≤20ppb, ω_Ni≤20ppb, and ω_total≤90ppb. If the sample test results meet the corresponding grade's standard requirements, the sample is deemed qualified; otherwise, it is deemed unqualified. For unqualified samples, the reasons for non-compliance must be analyzed (e.g., excessive impurities in raw materials, severe wear of production equipment, incomplete magnetic separation, etc.), and rectification suggestions must be proposed. The calculation process and judgment results are recorded to form a test report. The test report must include sample information, testing steps, test data, error correction data, parallel experimental data, result calculation (including formula derivation process), judgment conclusion, and rectification suggestions to ensure the completeness and traceability of the test report.

[0040] Working Principle: The core principle of the result calculation is based on the "law of conservation of mass." By measuring the sample's mass, the final volume at each stage, and the enrichment volume, the concentration of magnetic substances in the mother liquor is converted into the actual content of magnetic substances in the lithium carbonate sample. This is because the mother liquor is obtained by dissolving and diluting magnetic precipitates, while the magnetic precipitates are obtained by magnetic separation and enrichment from the mother liquor. The mother liquor itself is obtained by dissolving and diluting the sample; there is a clear proportional relationship between the volume and mass at each stage. The formula allows for the accurate calculation of the actual content of magnetic substances in the sample. (Unit conversion factor: 10). -The purpose of the ³ indicator is to convert the concentration unit of the mother liquor (ppb, i.e., μg / L) to the mass fraction unit of magnetic substances in the sample (ppb, i.e., μg / kg), ensuring unit consistency. The content of iron, cobalt, and nickel is calculated separately because different magnetic substances have different effects on the quality of lithium carbonate products. Industry standards have clear requirements for the content of each magnetic substance, and this also provides a basis for subsequent impurity source analysis. The core of quality judgment is to combine industry standards and customer requirements to clarify the grade and qualification of the sample, providing a clear basis for the factory inspection and quality control of lithium carbonate products. The analysis of the causes of non-conforming samples and rectification suggestions can help manufacturers optimize their production processes (such as raw material screening, equipment maintenance, and magnetic separation parameter adjustment), reduce the introduction of magnetic impurities, and improve product quality. The completeness and traceability of the test report can meet the needs of industrial production quality control and industry regulatory application promotion, ensuring that every test is regulated and traceable, avoiding arbitrariness in the testing process, and providing data support for the optimization and improvement of subsequent methods. The completion of this step provides clear test results and judgment criteria for the method verification in step 11. If the results are calculated incorrectly or the judgment criteria are unclear, the test will lose its meaning and will not be able to provide effective support for production and quality control.

[0041] Step 11: Method Validation and Archiving The core purpose of this step is to comprehensively verify the scientific validity, accuracy, stability, and applicability of the entire testing method, while also standardizing and archiving all data and reports from the testing process to support the promotion and application of the method and subsequent retesting. The operating procedure is as follows: Method validation is divided into indoor and outdoor validation. Indoor validation is conducted by professional laboratory testing personnel. Following steps 1-10 of this method, three standard samples of different concentrations (low concentration: 0.5 ppb, medium concentration: 10 ppb, high concentration: 50 ppb) are tested. Each concentration is measured in parallel six times. The relative standard deviation (RSD), spiked recovery, and limit of detection are calculated. Validation criteria: RSD ≤ 2.5%, spiked recovery 95%-105%, limit of detection ≤ 0.1 ppb. If the validation criteria are not met, the method parameters (such as gradient magnetic field parameters, ultrasonic washing parameters, and error correction model parameters) need to be re-optimized until the requirements are met. Outdoor validation is commissioned to three third-party testing institutions (with CNAS accreditation). They provide the same lithium carbonate samples and method operation manual. The third-party institutions conduct independent tests according to this method. The test results of the third-party institutions are compared with the test results of this laboratory. The relative deviation is calculated, and the relative deviation is required to be ≤ 3% to ensure the applicability of the method. To ensure universality and prevent methods from being limited to a single laboratory and thus unable to be widely applied, after method validation, all validation data are compiled into a "Detection Method Validation Report." This report includes the validation objective, validation protocol, validation data, validation conclusions, and optimization suggestions, ensuring the traceability of the validation process. Simultaneously, all relevant materials, including sample collection records, pretreatment records, weighing records, magnetic separation records, ultrasonic washing records, ICP-MS measurement records, error correction data, blank experiment data, parallel experiment data, result calculation records, test reports, and method validation reports, are classified and archived according to archival management standards. Paper documents are bound into volumes, and electronic documents are encrypted and stored, with an archiving period of no less than 5 years to facilitate subsequent re-inspection and quality traceability. Finally, the detection method is summarized, outlining its core innovations, operational points, and precautions, forming a method operation manual. This manual provides guidance for industrial-scale batch testing, routine laboratory testing, and personnel training, ensuring that different operators can obtain accurate and stable test results by following the manual.

[0042] Working Principle: The core principle of method validation is to verify the scientific validity and reliability of the method through multi-dimensional and multi-scenario testing experiments. Indoor validation focuses on verifying the stability and accuracy of the method under the same experimental conditions. Multiple measurements of standard samples with different concentrations verify the applicability of the method across the entire detection range, avoiding situations where the method is accurate only in a certain concentration range but deviates in other concentration ranges. Outdoor validation focuses on verifying the method's versatility and repeatability. Independent testing by third-party institutions eliminates subjective errors from single laboratories or operators, as well as the specificity of experimental conditions, ensuring that the method can achieve accurate detection in different laboratories and with different operators, meeting the requirements for industrial application. Validation of the detection limit ensures that the method meets the requirements for battery-grade lithium carbonate. This method addresses the need for detecting magnetic materials below the ppb level, resolving the pain point of existing methods having excessively high detection limits. The method validation report quantifies the method's performance indicators, providing a scientific basis for its reliability. The core principle of data archiving is to achieve full traceability of the testing process. In case of subsequent disputes over test results or sample quality issues, the archived data can be used to trace the entire testing process, identify the root cause, and ensure the authority of the test results. The creation of an operation manual standardizes testing procedures, reduces operator errors, ensures the standardization and normalization of the method, lowers personnel training costs, and promotes the industrial application of the method. The archiving period is no less than 5 years, complying with the relevant requirements for quality traceability management in the chemical industry, and providing data support for subsequent method optimization and industry supervision. This step is the final stage of the entire testing method process and a crucial prerequisite for its widespread application. Without method validation, the reliability and applicability of the method cannot be proven, failing to meet the requirements for practicality and hindering industrial promotion. Inadequate data archiving leads to an untraceable testing process, affecting the authority of the test results and the credibility of the method.

[0043] Example 2: Experimental environment control: The entire experiment was conducted in a clean laboratory (Class 1000), with the ambient temperature controlled at (25±1)℃ and relative humidity at (50±5)%RH. The laboratory was free of dust and magnetic interference sources (away from electromagnetic equipment and metal instruments), and the experimental table was covered with a corrosion-resistant, non-magnetic rubber mat. The laboratory environment was purified for 30 minutes before the experiment, and environmental parameters (temperature and humidity) were recorded. Records were made once before, during, and after the experiment each day to ensure stable environmental conditions and avoid interference from environmental factors on the test results.

[0044] Experimental Sample: A battery-grade lithium carbonate finished product sample (production batch: 202X0815, production process: spodumene lithium extraction, product grade: battery grade) was selected from a lithium-ion battery cathode material manufacturer. The sample was a white uniform powder with no lumps, no impurities, and no odor. The purity was initially tested to be 99.95% (Li2CO3 content ≥99.95%, total content of other impurities ≤0.05%). In accordance with the requirements of GB / T 6678-2003 "General Rules for Sampling of Chemical Products" and GB / T 11075-2023 "Lithium Carbonate" standards, aseptic samplers were used to collect samples from three locations in the finished product silo: upper, middle, and lower (upper: 1 / 3 of the silo height; middle: 2 / 3 of the silo height; lower: 10cm above the bottom of the silo). Three samples were collected from each location, with each sample weighing 60g. Sample contamination was avoided during sampling. Before each use, the sampler was rinsed three times with ultrapure water and wiped with anhydrous ethanol for disinfection, then allowed to air dry. The nine collected samples were combined and placed in a sterile stainless steel tray, gently mixed, and then reduced to quarters using the quartering method (reduction steps: pile the samples into a cone shape, flatten them, draw cross lines, discard the two diagonally opposite portions, repeat reduction three times), until a final uniform sample of 10g was obtained. The sample was then placed in a sterile sealed bag (number: Y-01), labeled with the sample name, production batch, sampling date, sampling location, and reduction time, and stored in a desiccator for later use. Simultaneously, samples of raw material lithium carbonate (No.: Y-02) and intermediate products from the same batch (No.: Y-03) were collected as control samples and tested concurrently to verify the applicability of the method to samples from different production stages.

[0045] Experimental Reagents and Reagent Pretreatment: Reagent Specifications and Acceptance Criteria: Superior Grade Hydrochloric Acid (Purity 37%±0.5%, Manufacturer: Sinopharm Group, Batch No.: 202X0720, Acceptance Criteria: No impurities, no precipitates; ICP-MS analysis shows that iron, cobalt, and nickel impurities are all ≤0.01ppb); Superior Grade Nitric Acid (Purity 68%±0.5%, Manufacturer: Sinopharm Group, Batch No.: 202X0725, Acceptance Criteria: No impurities, no yellow fumes; ICP-MS analysis shows that iron, cobalt, and nickel impurities are all ≤0.01ppb); Ultrapure Water (Resistivity 18.25MΩ·cm±0.05MΩ·cm, Preparation Equipment: Millipore Ultrapure Water System, Model: Milli-Q) Advantage A10, immediately sealed and stored after preparation to avoid contamination by airborne impurities; Iron standard stock solution (1000 μg / mL ± 2 μg / mL, National Metrological Certification Standard Material, No.: GBW(E)080128); Cobalt standard stock solution (1000 μg / mL ± 2 μg / mL, National Metrological Certification Standard Material, No.: GBW(E)080129); Nickel standard stock solution (1000 μg / mL ± 2 μg / mL, National Metrological Certification Standard Material, No.: GBW(E)080130); Isotope-labeled iron standard solution (500 μg / mL ± 1 μg / mL, isotope abundance: 57 Fe≥99%, Manufacturer: Sigma-Aldrich, Batch No.: 202X0801, for the determination of the true mass of magnetic materials; Blank reagents (ultrapure water, analytical grade hydrochloric acid, analytical grade nitric acid) must all meet the testing requirements and be free from magnetic impurities.

[0046] Reagent pretreatment: Before use, hydrochloric acid and nitric acid are purified by double distillation in a quartz still (distillation temperature: hydrochloric acid 108℃, nitric acid 86℃, distillation rate: 2mL / min). After purification, they are sealed in quartz reagent bottles and stored for 7 days. Before use, ultrapure water is filtered through a 0.22μm sterile filter membrane to remove small particulate impurities. Standard stock solution dilution: The 1000μg / mL iron, cobalt, and nickel standard stock solutions are diluted with 5% nitric acid solution to an intermediate stock solution of 10μg / mL, and then diluted to the target concentration standard series solution. Calibrated pipettes are used during the dilution process. After each dilution, the solution is shaken well and allowed to stand for 5 minutes to ensure uniform concentration. The dilution record is detailed, indicating the dilution steps, pipetting volume, dilution time, and operator.

[0047] Experimental procedures and detailed operating data Step 1: Sample Collection and Preprocessing Sampling procedure: According to the preset sampling plan, collect samples from three locations: the top, middle and bottom of the finished product silo. Collect three samples from each location, with each sample weighing 60g. When sampling, insert the sampler into the silo to a depth of ≥10cm to avoid collecting surface samples (surface samples may be contaminated). After sampling, immediately put the samples into a sterile sealed bag, label the sampling information, and avoid exposing the samples to air for more than 10 minutes to prevent moisture absorption and contamination.

[0048] Pretreatment: Combine 9 collected samples and place them in an agate mortar. Grind at a grinding force of 65N (calibrated with a force gauge, error ≤ ±2N) for 18 minutes. During grinding, scrape the sample off the inner wall of the mortar every 3 minutes with a sterile spatula to ensure uniform grinding. After grinding, pass all samples through a 200-mesh nylon sieve. When sieving, gently press the sample with a sterile spatula to avoid damaging the sieve due to excessive force. If any sample fails to pass through the sieve, put it back into the mortar and grind again until all samples pass through. Place the sieved samples on a quartz tray, spreading them evenly with a thickness ≤ 5mm, and place them in a vacuum drying oven. Set the drying temperature to 104℃ and the vacuum degree to 0.09MPa. Dry for 2 hours, recording the vacuum level and temperature every 30 minutes during the drying process. If the vacuum level drops below 0.08 MPa, adjust the vacuum valve promptly to ensure a stable vacuum level. After drying, close the vacuum drying chamber and allow it to naturally depressurize to atmospheric pressure (depressurization time ≥ 10 minutes to avoid sudden pressure changes causing sample splashing). Remove the sample and place it in a desiccator to cool for 30 minutes (during cooling, place color-changing silica gel in the desiccator, monitor humidity, and replace the silica gel promptly after it changes color). After cooling, use the quartering method to reduce the sample size, strictly following the standard operating procedure, reducing it to 10g of uniform sample, placing it in a sterile sealed bag, marking the pretreatment completion time, and set aside for later use.

[0049] Pretreatment of control samples: The raw material sample (Y-02) and the intermediate product sample (Y-03) are pretreated simultaneously. The operation steps are the same as those of the finished product sample (Y-01) to ensure that the pretreatment conditions are consistent.

[0050] Step 2: Accurately weigh and dilute the sample Balance calibration: Preheat the electronic balance to 0.0000g for 30 minutes, and calibrate the zero point and linearity using a 2.0000g standard weight. Calibration data: zero point calibration error 0.0000g, linearity calibration error 0.0000g. After the calibration is qualified, control the experimental environment temperature to 24.8℃ and humidity to 52%RH to avoid airflow and vibration interference with the balance weighing.

[0051] Accurate weighing: Using the difference method, select three calibrated 50mL volumetric flasks, numbered 1, 2, and 3 respectively. Weigh the volumetric flasks using a 0.01 g electronic balance and record the data. Then, add the pretreated sample to each volumetric flask, weigh the volumetric flask and sample together, and calculate the sample mass. The detailed weighing data for the three parallel samples are as follows (retain four decimal places): Sample 1 (corresponding to volumetric flask 1): Volumetric flask mass m0 = 28.3562 g, Volumetric flask + sample mass m1 = 30.3563 g, Sample mass m = 30.3563 - 28.3562 = 2.0001 g; Sample 2 (corresponding to volumetric flask 2): Volumetric flask mass m0 = 28.3565g, volumetric flask + sample mass m1 = 30.3566g, sample mass m = 30.3566 - 28.3565 = 2.0001g; Sample 3 (corresponding to volumetric flask 3): Volumetric flask mass m0 = 28.3561g, Volumetric flask + sample mass m1 = 30.3562g, Sample mass m = 30.3562 - 28.3561 = 2.0001g; Weighing error calculation: According to the formula X1=(|m - m0standard| / m0standard×100%), where m0standard=2.0000g, the weighing error of the three samples is (|2.0001-2.0000| / 2.0000×100%)=0.005%, which meets the requirement of ≤0.01%; if the weighing error exceeds 0.01%, the samples need to be weighed again until the requirement is met.

[0052] Sample dissolution and volume adjustment: Slowly add 20 mL of 1+1 hydrochloric acid solution (hydrochloric acid and ultrapure water volume ratio 1:1, prepared in advance and shaken well) to each sample volumetric flask. When adding, the hydrochloric acid solution should flow slowly along the inner wall of the volumetric flask to avoid direct impact on the sample. After adding, gently shake the volumetric flask until the sample is completely dissolved. (If any sample sticks to the wall during the dissolution process, rinse the inner wall of the volumetric flask 4 times with ultrapure water, 2 mL each time, to ensure that the sample is completely dissolved.) After the sample is completely dissolved, rinse the inner wall of the volumetric flask 4 times with ultrapure water, 2 mL each time, and let all the rinsing solution flow into the volumetric flask. Then, dilute to the 50 mL mark with ultrapure water. When diluting, keep your line of sight horizontal to the mark and avoid looking down or up. After diluting, shake well ≥10 times and let stand for 10 minutes to obtain 3 sample mother solutions A1, A2, and A3. Label the mother solution number and preparation time, and place them on the lab bench for later use.

[0053] Control sample weighing and volume adjustment: The raw material sample (Y-02) and the intermediate product sample (Y-03) were weighed and adjusted simultaneously. Three parallel mother liquors were prepared for each sample. The weighing mass of each sample was 2.0000±0.0001g, and the volume was adjusted to 50mL. The operation steps were the same as those for the finished product sample. The mother liquors were numbered A4-A6 (Y-02) and A7-A9 (Y-03).

[0054] Step 3: Gradient magnetic field magnetic separation and enrichment Apparatus Calibration: Before the experiment, the quartz separation column of the gradient magnetic field separation apparatus was rinsed four times with ultrapure water, each time with a volume of 20 mL and a flow rate of 2 mL / min. After rinsing, a gaussmeter was used to calibrate three points at heights of 50 mm, 100 mm, and 150 mm on the separation column. The magnetic field strength error was ≤ ±0.01 T. The temperature control system was set to 25℃ and stabilized for 30 minutes. The temperature of the inner wall of the separation column was measured with a precision thermometer and found to be 25.0℃, with an error ≤ ±0.1℃. The peristaltic pump was calibrated with a flow rate set to 1.2 mL / min. After calibration, the actual flow rate was 1.20 ± 0.02 mL / min, which met the requirements.

[0055] Magnetic separation operation: Pour the three finished product sample mother liquors A1, A2, and A3 into three calibrated quartz separation columns, adding 20 mL of mother liquor to each column (using a pipette for precise volume transfer, with a pipetting error ≤ ±0.01 mL). Control the flow rate at 1.2 mL / min using a peristaltic pump (corresponding to a peristaltic pump speed of 65 r / min, calibrated, with a speed error ≤ ±1 r / min). Turn on the magnetic field generator and set the gradient magnetic field parameters: initial magnetic field 0.1 T, uniformly increasing to 1.5 T at a rate of 0.05 T / min, maintaining 1.5 T for 5 minutes (timing error ≤ ±10 s), and then increasing to 0.1 T / min. The temperature was uniformly reduced to 0.1T to complete one magnetic field cycle. A total of three cycles were performed, with a one-minute interval between each cycle to ensure sufficient adsorption of the magnetic material. During the magnetic separation process, the separation column temperature was monitored in real time and maintained at 25℃. If the temperature exceeded the range, the temperature controller was adjusted promptly. After the magnetic separation was completed, the magnetic field generator was turned off, and the column was allowed to stand for 5 minutes (timing error ≤ ±10s) to allow the magnetic material to settle completely. Then, the magnetic material enrichment solutions B1, B2, and B3 were accurately collected using pipettes. The volume of each enrichment solution was 20.0 ± 0.02 mL. After collection, the solutions were placed in sterile centrifuge tubes and labeled with numbers. At the same time, the waste liquid was collected and placed in a dedicated waste liquid container, and waste liquid treatment records were kept.

[0056] Magnetic separation of control samples: Simultaneously perform magnetic separation on raw material sample mother liquor A4-A6 and intermediate product mother liquor A7-A9. The operating parameters are the same as those for the finished product samples. Collect enriched solutions B4-B6 (Y-02) and B7-B9 (Y-03) for later use.

[0057] Recovery rate determination and validation: The magnetic separation recovery rate was determined using the isotope labeling method. 10 μL of isotope-labeled iron standard solution (500 μg / mL) was added to each mother liquor sample. After labeling, the magnetic separation procedure described above was followed. The enriched solution was collected, and the content of isotope-labeled iron in the enriched solution was determined using ICP-MS. The magnetic separation recovery rate (X2) was calculated. The recovery rate data for the three finished samples are as follows: X 21 =99.1%, X 22 =99.3%, X 23 =99.2%, average value X2=99.2%, recovery rate range 99.1%-99.3%, RSD=0.1%, meeting the requirements of recovery rate ≥99% and RSD≤0.5%; if the recovery rate is lower than 99%, check the magnetic field strength, flow rate, and number of cycles, adjust the parameters and repeat the magnetic separation operation.

[0058] Step 4: Ultrasonic-assisted washing and purification of the enrichment solutions: Pour the finished sample enrichment solutions B1, B2, and B3 into three 50mL sterile centrifuge tubes respectively. Add 10.0mL of ultrapure water to each centrifuge tube (use a pipette for precise transfer, error ≤ ±0.01mL). Shake the centrifuge tubes by hand for 1 minute (shaking frequency 60 times / min) to ensure thorough mixing of the enrichment solutions and ultrapure water. Place the centrifuge tubes in an ultrasonic cleaner. Add ultrapure water to the ultrasonic tank, ensuring the water level is at least 2cm above the liquid surface of the centrifuge tubes. Set the ultrasonic power to 220W (calibrated, actual power 220±5W) and the frequency to 40kHz. The water temperature was 24℃, and the centrifuge tubes were sonicated for 15 minutes. During sonication, the centrifuge tubes were gently shaken once every 3 minutes to ensure even washing. After sonication, the centrifuge tubes were placed in a high-speed centrifuge and the centrifugation speed was set to 8000 r / min (calibrated, actual speed 8000±50 r / min), centrifugation temperature 25℃, and centrifugation time 10 minutes (timing error ≤±10s). After centrifugation, the supernatant was gently aspirated with a pipette (avoiding aspirating the bottom sediment) and poured into a dedicated waste container. The above sonication, centrifugation, and supernatant aspiration operations were repeated 3 times. Each time, 10.0 mL of ultrapure water was used for washing, and the operating parameters were consistent.

[0059] Washing of control samples: Simultaneously wash the raw material sample enrichment solutions B4-B6 and the intermediate product enrichment solutions B7-B9. The operation steps and parameters are consistent with those of the finished product samples to ensure uniform washing conditions.

[0060] Impurity removal verification: After the third wash, 0.5 mL of supernatant from each centrifuge tube was pipetted into a cuvette of a UV-Vis spectrophotometer. Using ultrapure water as a blank reference, absorbance was measured every 10 nm within the wavelength range of 200-800 nm. The results are as follows: Absorbance of supernatant B1 0.0008, supernatant B2 0.0007, and supernatant B3 0.0009, all ≤0.001, indicating that the matrix impurity (Li) is within acceptable limits. + Cl - Remove non-magnetic impurities thoroughly; if absorbance > 0.001, repeat the washing operation 1-2 times until the absorbance meets the requirements.

[0061] Determination of residual matrix impurities: Li in the supernatant after the third wash was determined by ICP-MS. + The concentration was used to calculate the residual matrix impurity amount X3. The residual data for the three finished product samples are as follows: X 31 =0.028ppb, X 32 =0.032ppb, X 33 =0.030ppb, average value X3=0.03ppb, residual amount ≤0.05ppb, meets the requirements; if residual amount >0.05ppb, adjust ultrasonic power to 250W, extend ultrasonic time to 20 minutes, and wash and measure again.

[0062] Precipitation retention: After washing, retain the magnetic precipitates C1, C2, and C3 at the bottom of the centrifuge tube. Gently scrape off a small amount of precipitate from the inner wall of the centrifuge tube with a sterile spatula, ensuring that all the precipitate remains at the bottom of the centrifuge tube. Label the precipitate with the precipitate number and place it on the lab bench for later use. After washing the control sample, retain the precipitates C4-C6 (Y-02) and C7-C9 (Y-03) for later use.

[0063] Step 5: Precipitation, dissolution, and volume adjustment of magnetic substances Precipitation and dissolution procedure: Add 5.0 mL of 5% nitric acid solution (using a pipette for precise transfer, with an error ≤ ±0.01 mL; the 5% nitric acid solution should be prepared in advance, with a volume ratio of analytical grade nitric acid to ultrapure water of 5:95, and shaken well before use) to the centrifuge tubes corresponding to the finished sample precipitates C1, C2, and C3. Gently stir the precipitate with a sterile spatula to ensure that the precipitate is in full contact with the nitric acid solution. Place the centrifuge tubes in a constant temperature water bath, set the water bath temperature to 59℃ (calibrated, actual temperature 59±0.1℃), and the water bath time to 20 minutes. During the water bath, gently shake the centrifuge tubes once every 5 minutes to observe the dissolution of the precipitate, ensuring that the precipitate is completely dissolved (no black or brown residue, and the solution is colorless and transparent). If the precipitate is not completely dissolved, add 1-2 mL of 5% nitric acid solution and extend the water bath time by 10 minutes until it is completely dissolved.

[0064] Control sample precipitation dissolution: Simultaneously dissolve the raw material sample precipitates C4-C6 and the intermediate product precipitates C7-C9. The operation steps and parameters are consistent with those of the finished product samples to ensure uniform dissolution conditions.

[0065] Cooling and Transfer: After the precipitate has completely dissolved, remove the centrifuge tube from the constant temperature water bath and place it in a fume hood to cool for 20 minutes (during the cooling process, the fume hood airflow should be controlled at 1.5 m / s to prevent solution evaporation). After cooling to room temperature (25°C), use a pipette to slowly transfer the solution in the centrifuge tube to a 25 mL volumetric flask. During the transfer, the solution should flow along the inner wall of the volumetric flask to avoid splashing. After the transfer is complete, rinse the inner wall of the centrifuge tube 4 times with ultrapure water, each time with a volume of 2 mL. All the rinsing solution should flow into the volumetric flask to ensure that all the precipitate dissolution solution has been transferred and there is no residue.

[0066] Volume adjustment procedure: Add ultrapure water to each volumetric flask and adjust the volume to the 25mL mark. When adjusting the volume, keep your line of sight horizontal to the mark and avoid looking down or up. After adjusting the volume, shake well ≥10 times and let stand for 5 minutes to obtain 3 finished sample test mother solutions D1, D2 and D3. Label the mother solution number and preparation time and place them on the laboratory table for later use.

[0067] Preparation of blank solution: Prepare 5 blank solutions D simultaneously. 01 -D 05 (For subsequent blank experiment correction) Preparation method for each blank solution: Add 5.0 mL of 5% nitric acid solution to a 25 mL volumetric flask, rinse the pipette 4 times with ultrapure water, let the rinsing solution flow into the volumetric flask, dilute to the 25 mL mark, shake well and let stand for 5 minutes, label with blank number, and keep for later use; during the preparation of blank solutions, strictly avoid contamination, and use the same reagents and instruments as the test stock solution.

[0068] Preparation of control sample test mother liquor: The solutions of raw material samples and production intermediates were simultaneously diluted to a fixed volume to prepare test mother liquors D4-D6 (Y-02) and D7-D9 (Y-03). Three parallel mother liquors were prepared for each sample. Blank solutions were prepared simultaneously. The operation steps were the same as those for finished samples.

[0069] Step 6: ICP-MS Instrument Calibration and Parameter Optimization Instrument startup and warm-up: Power on the ICP-MS and, following the instrument operating procedures, turn on argon gas (99.999% purity, 0.4 MPa pressure), setting the argon flow rate to 15 L / min. Warm up for 60 minutes after startup. During the warm-up process, record the instrument vacuum level and argon flow rate every 15 minutes, ensuring the vacuum level is ≤1×10⁻⁶. -6mbar and argon flow rate are stable (fluctuation ≤ ±0.1 L / min); after preheating, the instrument is calibrated using a tuning solution (containing elements such as Li, Co, In, and U, concentration 10 μg / L) to optimize instrument sensitivity, resolution, oxide interference, and double charge interference. The calibration standard is: sensitivity ≥ 1 × 10⁻⁶ mbar. 6 cps / μg·L - ¹, resolution 0.7 amu, CeO / Ce ≤ 0.02, Ba² + If / Ba≤0.03, after successful tuning, save the tuning parameters.

[0070] Preparation of standard series solutions: A series of standard solutions for iron, cobalt, and nickel were prepared using a stepwise dilution method, with concentrations of 0.1 ppb, 1 ppb, 10 ppb, 50 ppb, and 100 ppb, respectively. Three replicates were prepared for each concentration. A blank standard solution (5% nitric acid solution) was also prepared. Detailed data on the preparation of the standard series solutions are as follows (unit: ppb): Iron standard series: 0.1 (parallel 1: 0.100, parallel 2: 0.099, parallel 3: 0.101), 1 (parallel 1: 1.000, parallel 2: 0.998, parallel 3: 1.002), 10 (parallel 1: 10.00, parallel 2: 9.99, parallel 3: 10.01), 50 (parallel 1: 50.00, parallel 2: 49.98, parallel 3: 50.02), 100 (parallel 1: 100.0, parallel 2: 99.9, parallel 3: 100.1); Cobalt standard series: 0.1 (parallel 1: 0.100, parallel 2: 0.101, parallel 3: 0.099), 1 (parallel 1: 1.001, parallel 2: 0.999, parallel 3: 1.000), 10 (parallel 1: 10.01, parallel 2: 9.99, parallel 3: 10.00), 50 (parallel 1: 50.02, parallel 2: 49.99, parallel 3: 50.00), 100 (parallel 1: 100.1, parallel 2: 99.9, parallel 3: 100.0); Nickel standard series: 0.1 (parallel 1: 0.099, parallel 2: 0.100, parallel 3: 0.101), 1 (parallel 1: 0.999, parallel 2: 1.000, parallel 3: 1.001), 10 (parallel 1: 9.99, parallel 2: 10.00, parallel 3: 10.01), 50 (parallel 1: 49.99, parallel 2: 50.00, parallel 3: 50.01), 100 (parallel 1: 99.9, parallel 2: 100.0, parallel 3: 100.1); After preparing the standard series solutions, shake well, let stand for 10 minutes, and set aside. During the preparation process, record the volume of liquid transfer and dilution steps in detail to ensure accurate concentration.

[0071] Standard curve plotting: The standard series solutions (including blank standard solutions) were imported into ICP-MS for measurement in ascending order of concentration. Each sample was measured in triplicate, and the signal intensity (cps) was recorded. A standard working curve was plotted with concentration as the x-axis and signal intensity as the y-axis. Linear regression was used for fitting, and outliers were removed during the fitting process (using Grubbs' criterion, α=0.05). The standard working curve data are as follows: Iron: Regression equation y=102500x + 120, correlation coefficient R²=0.99995, standard deviation σ=0.002, fitting error ≤±0.001; Cobalt: Regression equation y=101800x + 115, correlation coefficient R²=0.99996, standard deviation σ=0.002, fitting error ≤±0.001; Nickel: Regression equation y=103200x + 125, correlation coefficient R²=0.99995, standard deviation σ=0.002, fitting error ≤±0.001; The standard curve requires R² ≥ 0.9999. The fitting result meets the requirement. If R² < 0.9999, the standard series solutions should be prepared again and the standard curve should be redrawn.

[0072] Instrument parameter optimization: Based on the tuning parameters, the ICP-MS instrument parameters were optimized. Through single-factor experiments, the RF power, nebulizer gas flow rate, auxiliary gas flow rate, sampling depth, and integration time were optimized respectively. The final instrument parameters after optimization are as follows, and the stability of the parameters was verified: RF power: 1350W (fluctuation ≤ ±10W); Nebulizer gas flow rate: 1.0L / min (fluctuation ≤ ±0.02L / min); Auxiliary gas flow rate: 0.3L / min (fluctuation ≤ ±0.01L / min); Sampling depth: 10mm (error ≤ ±0.1mm); Integration time: 0.3s (error ≤ ±0.01s); Sampling mode: Peak skipping mode; Number of repetitions: 3; Oxide interference: CeO / Ce = 0.015 ≤ 0.02; Dual charge interference: Ba² + / Ba=0.025≤0.03; Parameter stability verification: Using a 10ppb standard solution, 6 consecutive measurements were performed, and the RSD was calculated. The RSD for iron was 1.7%, for cobalt it was 1.8%, and for nickel it was 1.7%, all ≤2%, indicating that the instrument parameters were stable. The parameters were saved for future use.

[0073] Step 7: Detection of mother liquor and correction for multivariate statistical errors Sample testing procedure: Test the finished product sample using mother solutions D1, D2, D3, and blank solution D. 01 -D 05Following the instrument operating procedures, the samples were imported into the ICP-MS for analysis. Each sample was measured in parallel three times, with the following measurement order: blank solution → low-concentration test stock solution → high-concentration test stock solution, to avoid cross-contamination. During the analysis, a 10 ppb standard solution was inserted every 10 samples for quality control to ensure accurate results. If the measured value of the quality control sample deviated from the standard value by more than 3%, the analysis was stopped, the instrument was recalibrated, the cause was investigated, and the analysis was continued.

[0074] Control sample testing: The raw material sample test mother liquors D4-D6 and the intermediate product test mother liquors D7-D9 were tested simultaneously. The operation steps and test order were the same as those for the finished product samples. Each sample was tested in parallel 3 times. Quality control samples were inserted to ensure the accuracy of the test.

[0075] Preliminary detection concentration data: After the measurement is completed, the signal intensity is converted into preliminary detection concentration C according to the standard working curve. i The preliminary concentration data of the mother liquor and blank solution for the three finished product samples are as follows (unit: ppb, rounded to 3 decimal places): Iron: Initial concentration of D1 (parameter 1: 2.860, parallel 2: 2.858, parallel 3: 2.862), average C1 = 2.860; Initial concentration of D2 (parameter 1: 2.880, parallel 2: 2.878, parallel 3: 2.882), average C2 = 2.880; Initial concentration of D3 (parameter 1: 2.870, parallel 2: 2.868, parallel 3: 2.872), average C3 = 2.870; Concentration of D0 (blank) (parameter 1: 0.040, parallel 2: 0.039, parallel 3: 0.041), average C0 = 0.040; Cobalt: Initial concentration of D1 (parallel 1: 1.230, parallel 2: 1.228, parallel 3: 1.232), average C1 = 1.230; Initial concentration of D2 (parallel 1: 1.220, parallel 2: 1.218, parallel 3: 1.222), average C2 = 1.220; Initial concentration of D3 (parallel 1: 1.230, parallel 2: 1.228, parallel 3: 1.232), average C3 = 1.230; D0 (blank) concentration (parallel 1: 0.030, parallel 2: 0.029, parallel 3: 0.031), average C0 = 0.030; Nickel: Initial concentration of D1 (parallel 1: 1.180, parallel 2: 1.178, parallel 3: 1.182), average C1 = 1.180; Initial concentration of D2 (parallel 1: 1.190, parallel 2: 1.188, parallel 3: 1.192), average C2 = 1.190; Initial concentration of D3 (parallel 1: 1.180, parallel 2: 1.178, parallel 3: 1.182), average C3 = 1.180; D0 (blank) concentration (parallel 1: 0.030, parallel 2: 0.029, parallel 3: 0.031), average C0 = 0.030; Quality control data: The 10ppb iron standard solution was measured 6 times with values ​​of 10.02, 9.98, 10.01, 9.99, 10.00, and 10.01, respectively. The average value was 10.00, and the deviation was 0.00%, which meets the requirement of deviation ≤3%. The quality control results of the cobalt and nickel standard solutions were consistent and the measurements were accurate.

[0076] Error parameter calculation: 1. Weighing error X1: The sample weighing mass m = 2.0001g, the standard weighing mass m0 = 2.0000g, according to the formula X1 = (|m - m0| / m0 × 100%) × π / 6 × exp(-γ² / 2), where π = 3.14159, γ = 0.5772, e = 2.71828, we calculate X1 ≈ 0.005% × 0.5236 × 0.7408 ≈ 0.0019%. The X1 values ​​of the three samples are consistent, all ≤ 0.002%. 2. Magnetic separation recovery rate X2: The recovery rates of the three samples were 99.1%, 99.3%, and 99.2%, respectively, with an average value X2 = 99.2%. The calculated X2 = 99.2% × (1 + e / 100) × √(π / (2γ)) ≈ 99.2% × 1.0272 × √(3.14159 / (2 × 0.5772)) ≈ 99.2% × 1.0272 × 1.653 ≈ 106.1% (corrected recovery rate). 3. Residual matrix impurities X3: The residual amounts of the three samples were 0.028 ppb, 0.032 ppb, and 0.030 ppb, respectively, with an average value X3 = 0.03 ppb. After correction, X3 = 0.03 × γ / √(2π) × exp(-σ3² / 2) ≈ 0.03 × 0.5772 / 2.5066 × exp(-0.002² / 2) ≈ 0.03 × 0.2303 × 0.99998 ≈ 0.0069 ppb, where σ3 = 0.002 ppb; 4. Standard deviation of weighing error σ1: Through 10 parallel weighing experiments, σ1 = 0.00005g was obtained by fitting, which follows a normal distribution N(0.00005, 0.000005²). 5. Random error term ε: follows a normal distribution ε ~ N(0, σ²), where σ² = (1 / (n-1)) × Σ(ε) i - )²×π / (γe), n=3, ε i For single-shot random error, =0.0001ppb, calculated as σ²≈(1 / 2)×(0.0001²+0.0001²+0.0001²)×3.14159 / (0.5772×2.71828)≈(1 / 2)×3×10 -8 ×1.904≈2.856×10 -8 σ≈5.344×10 -4 ppb; 6. Deviation weighting factor φ i : Calculation formula φ i =| i - | / (max( i ) - min( i ))×γ, we calculate the iron's φ1=0.001×0.5772≈0.000577, φ2=0.001×0.5772≈0.000577, φ3=0.001×0.5772≈0.000577, Σφ i ² = 3 × (0.000577)² ≈ 1.00 × 10 -6 ,√((1 / n)×Σφ i ²)=√((1 / 3)×1.00×10 -6 )≈5.773×10 -4 .

[0077] Error correction model fitting and solution: Through 15 spiked recovery experiments with different concentrations (spiking concentrations of 0.1-100 ppb, 5 groups each of low, medium, and high concentrations), a multivariate statistical error correction model was fitted, and the regression parameters were obtained as follows (combining partial least squares regression and Bayesian regularization): β0=0.002 (regression constant term), β1=1.003 (detection concentration regression coefficient), β2=-0.001 (weighing error regression coefficient), β3=0.002 (magnetic separation recovery rate regression coefficient), β4=-0.001 (matrix interference regression coefficient), regularization parameter λ=γ / (πe)≈0.5772 / (3.14159×2.71828)≈0.066; Model validation parameters: R²=0.99985, Radj²=0.9996, RMSE=0.025ppb, MAE=0.018ppb, F-statistic F=12500≥10000 (α=0.01), the model is significant and meets the requirements.

[0078] Concentration calculation after error correction: C i Substituting parameters such as X1, X2, and X3 into the error correction model, the concentration after error correction is calculated. (Unit: ppb, rounded to 3 decimal places), the calculation process is recorded in detail, and the results are as follows: iron: 1 = 2.850 2 = 2.870, 3 = 2.860, the corrected average concentration of 3 parallel samples. Iron = 2.860 ppb, maximum deviation in a single measurement = |2.870 - 2.860| = 0.010 ppb, deviation ≤ 0.02 ppb, meets the accuracy requirements; cobalt: 1 = 1.220 2 = 1.210, 3 = 1.220, the corrected average concentration of 3 parallel samples. Cobalt = 1.217 ppb, maximum deviation in a single measurement = |1.220 - 1.217| = 0.003 ppb, deviation ≤ 0.02 ppb, meets accuracy requirements; nickel: 1 = 1.170, 2 = 1.180, 3 = 1.170, the corrected average concentration of 3 parallel samples. Nickel = 1.173 ppb, maximum deviation in a single measurement = |1.180 - 1.173| = 0.007 ppb, deviation ≤ 0.02 ppb, meets accuracy requirements; Corrected concentration of control sample: Iron in raw material sample (Y-02) =3.250ppb, cobalt =1.530ppb, nickel =1.480ppb; Iron in intermediate product sample (Y-03) =3.020ppb, cobalt =1.380ppb, nickel =1.320ppb, all of which meet the requirement that the deviation of a single measurement is ≤0.02ppb.

[0079] Step 8: Blank Experiment Correction The purpose of the blank experiment is to eliminate systematic errors caused by experimental reagents, instruments, and environment, and to ensure the accuracy of the test results. The blank experiment is carried out simultaneously with the sample test, and the operating conditions and steps are completely consistent, except that lithium carbonate sample is not added.

[0080] Blank solution determination and data statistics: Five blank solutions D were prepared. 01 -D 05 According to the ICP-MS assay specifications, the blank solution was measured simultaneously with the sample stock solution. Each blank solution was measured in triplicate, and the initial detection concentration was recorded. After multivariate statistical error correction, the blank correction concentration was obtained. Detailed data are as follows (unit: ppb, rounded to 3 decimal places): Iron Blank: D 01 =0.039、D 02 =0.040, D 03 =0.041、D 04 =0.039、D 05 =0.040, average blank correction concentration The air sample concentration was 0.040 ppb, and the RSD was 2.0% ≤ 2.5%, which meets the requirements for a blank experiment. Cobalt blank: D 01 =0.029、D 02 =0.030、D 03 =0.031、D 04 =0.029、D 05 =0.030, average blank correction concentration Empty cobalt = 0.030 ppb, RSD = 2.1% ≤ 2.5%, which meets the requirements for blank experiment; Nickel blank: D 01 =0.029、D 02 =0.030、D 03 =0.031、D 04 =0.029、D 05 =0.030, average blank correction concentration The blank nickel content is 0.030 ppb, and the RSD is 2.1% ≤ 2.5%, which meets the requirements for the blank experiment. The passing criteria for blank experiments are: the average blank correction concentration ≤ 0.05 ppb and RSD ≤ 2.5%. All blank experiment data in this study meet the requirements. If the average blank correction concentration > 0.05 ppb or RSD > 2.5%, investigate factors such as reagent contamination, instrument residue, and environmental interference. Replace the reagents, clean the instruments, and then prepare a new blank solution and perform the measurement.

[0081] Sample concentration blank correction: Subtract the average blank correction concentration of the corresponding element from the error-corrected sample concentration to obtain the blank-corrected sample concentration C. The calculation formula is: Ccorrection = sample- The detailed calculation results are as follows (unit: ppb, rounded to 3 decimal places): Finished product sample (Y-01): Iron C correction = 2.860 - 0.040 = 2.820 ppb; Cobalt C correction = 1.217 - 0.030 = 1.187 ppb; Nickel C correction = 1.173 - 0.030 = 1.143 ppb; Raw material sample (Y-02): Iron C correction = 3.250 - 0.040 = 3.210 ppb; Cobalt C correction = 1.530 - 0.030 = 1.500 ppb; Nickel C correction = 1.480 - 0.030 = 1.450 ppb; Production intermediate sample (Y-03): Iron C correction = 3.020 - 0.040 = 2.980 ppb; Cobalt C correction = 1.380 - 0.030 = 1.350 ppb; Nickel C correction = 1.320 - 0.030 = 1.290 ppb; Verification after correction: After blank correction, the RSD of the three parallel samples was recalculated. The RSD of the finished sample was 0.35% for iron, 0.41% for cobalt, and 0.43% for nickel. All of these values ​​were ≤2.5%, indicating that the blank correction was effective and eliminated systematic errors.

[0082] Step 9: Parallel experimental verification The purpose of parallel experiments is to verify the repeatability and stability of the detection method and ensure that the detection results of different operators and different experimental batches are consistent. Parallel experiments are divided into three categories: parallel verification within the same batch, parallel verification between different batches, and parallel verification by different personnel.

[0083] 1. Parallel validation within the same batch: Take the same pretreated finished product sample (Y-01), and prepare 6 parallel test stock solutions (D) by the same operator, in the same experimental environment, and with the same set of instruments. 10 -D 15 Following the steps outlined above, complete the detection, error correction, and blank correction. Record the concentration after blank correction. The data is as follows (unit: ppb, rounded to 3 decimal places): Iron: 2.820, 2.818, 2.822, 2.819, 2.821, 2.820; average value = 2.820 ppb, RSD = 0.06% ≤ 2.5%; Cobalt: 1.187, 1.185, 1.189, 1.186, 1.188, 1.187; average value = 1.187 ppb, RSD = 0.14% ≤ 2.5%; Nickel: 1.143, 1.141, 1.145, 1.142, 1.144, 1.143; average value = 1.143 ppb, RSD = 0.14% ≤ 2.5%; 2. Parallel validation across different batches: Lithium carbonate finished product samples from the same production batch (202X0815) were pretreated and tested in three batches (batch 1, batch 2, and batch 3). Three parallel samples were prepared for each batch. The average concentration after blank correction is as follows (unit: ppb, rounded to three decimal places): Batch 1: Iron 2.820, Cobalt 1.187, Nickel 1.143; Batch 2: Iron 2.818, Cobalt 1.185, Nickel 1.141; Batch 3: Iron 2.822, Cobalt 1.189, Nickel 1.145; Average values ​​across different batches: iron 2.820 ppb, cobalt 1.187 ppb, nickel 1.143 ppb; inter-batch RSD: iron 0.07%, cobalt 0.17%, nickel 0.17%, all ≤2.5%, indicating good inter-batch repeatability of the method; 3. Parallel validation by different personnel: Two different operators (Operator A and Operator B) tested the same pretreated sample. Each operator prepared three parallel samples. The average concentration after blank correction is as follows (unit: ppb, rounded to three decimal places): Operator A: Iron 2.820, Cobalt 1.187, Nickel 1.143; Operator B: Iron 2.819, Cobalt 1.186, Nickel 1.142; Inter-operator RSD: Iron 0.02%, Cobalt 0.09%, Nickel 0.09%, all ≤2.5%, indicating good inter-operator repeatability of the method.

[0084] Parallel experiment acceptance criteria: The RSD of parallel verifications of the same batch, different batches, and different personnel are all ≤2.5%, and the data of this parallel experiment all meet the requirements; if the RSD>2.5%, check the consistency of the operation steps (such as weighing accuracy, magnetic separation parameters, and volume determination accuracy), and repeat the parallel experiment until the requirements are met.

[0085] Stability analysis: The same stock solution was analyzed by ICP-MS at 0h, 2h, 4h, 6h, 8h, and 12h, for a total of 6 measurements. The RSD was calculated to verify the stability of the stock solution. The data are as follows (unit: ppb, rounded to 3 decimal places): Iron: 2.820 (0h), 2.819 (2h), 2.821 (4h), 2.820 (6h), 2.819 (8h), 2.820 (12h), RSD=0.03%≤2.5%; Cobalt: 1.187 (0h), 1.186 (2h), 1.188 (4h), 1.187 (6h), 1.186 (8h), 1.187 (12h), RSD=0.07%≤2.5%; Nickel: 1.143 (0h), 1.142 (2h), 1.144 (4h), 1.143 (6h), 1.142 (8h), 1.143 (12h), RSD = 0.07% ≤ 2.5%; The results showed that the test stock solution had good stability within 12 hours and could meet the experimental testing requirements. If the RSD > 2.5%, it indicated that the test stock solution was unstable and the test should be completed within 6 hours after preparation.

[0086] Step 10: Result Calculation and Judgment 1. Calculation of magnetic material content: Based on the sample concentration after blank correction, combined with the sample mass, fixed volume, and magnetic separation enrichment factor, the actual content of magnetic materials (iron, cobalt, nickel) in the lithium carbonate sample is calculated. The calculation formula is as follows: ω(μg / kg) = (C correction × V2 × V1) / (m × V3 × X2 correction) Formula explanation: ω is the content of magnetic substances in the sample (unit: μg / kg, i.e., ppb); C correction is the concentration of the detection stock solution after blank correction (unit: ppb); V2 is the volume of the detection stock solution (25mL); V1 is the volume of the sample stock solution (50mL); m is the sample mass (2.0001g); V3 is the volume of sample stock solution transferred during magnetic separation (20mL); X2 correction is the magnetic separation recovery rate after correction (106.1%). Calculation process (taking the iron content of finished product sample Y-01 as an example, retaining 3 decimal places): ωiron = (2.820 ppb × 25 mL × 50 mL) / (2.0001 g × 20 mL × 1.061) = (3525) / (42.444126) ≈ 83.05 μg / kg; Similarly, the actual contents of the three magnetic substances in the finished product sample, raw material sample, and intermediate product sample were calculated, and the detailed data are as follows (unit: μg / kg, rounded to 3 decimal places): Finished product sample (Y-01): Iron 83.05 μg / kg, Cobalt 35.08 μg / kg, Nickel 33.71 μg / kg; Total content of the three magnetic materials = 83.05 + 35.08 + 33.71 = 151.84 μg / kg; Raw material sample (Y-02): Iron 94.75 μg / kg, Cobalt 44.33 μg / kg, Nickel 42.88 μg / kg; Total content of the three magnetic materials = 94.75 + 44.33 + 42.88 = 181.96 μg / kg; Production intermediate sample (Y-03): Iron 87.83 μg / kg, Cobalt 40.02 μg / kg, Nickel 38.97 μg / kg; Total content of the three magnetic substances = 87.83 + 40.02 + 38.97 = 166.82 μg / kg; Calculation verification: The content of each sample was calculated three times, and the RSD of the calculation result was ≤0.1% to ensure the accuracy of the calculation. The calculation process was recorded in detail, including formula substitution, numerical calculation, and result retention, so as to facilitate traceability.

[0087] 2. Result determination: According to the requirements of GB / T 11075-2023 "Lithium Carbonate" standard, in battery-grade lithium carbonate, the iron content is ≤100μg / kg, the cobalt content is ≤50μg / kg, the nickel content is ≤50μg / kg, and the total content of the three magnetic materials is ≤200μg / kg; Judgment result: Finished product sample (Y-01): Iron 83.05μg / kg≤100μg / kg, Cobalt 35.08μg / kg≤50μg / kg, Nickel 33.71μg / kg≤50μg / kg, Total content 151.84μg / kg≤200μg / kg, judged as qualified; Raw material sample (Y-02): Iron 94.75μg / kg≤100μg / kg, Cobalt 44.33μg / kg≤50μg / kg, Nickel 42.88μg / kg≤50μg / kg, Total content 181.96μg / kg≤200μg / kg, judged as qualified; The intermediate product sample (Y-03) has the following content: iron 87.83 μg / kg ≤ 100 μg / kg, cobalt 40.02 μg / kg ≤ 50 μg / kg, nickel 38.97 μg / kg ≤ 50 μg / kg, and the total content is 166.82 μg / kg ≤ 200 μg / kg, which is considered qualified. If the content of a certain magnetic substance or the total content in a sample exceeds the standard requirements, it is deemed unqualified and a new sample must be collected for retesting. If the retest still fails, the batch of samples is deemed unqualified, and the source of magnetic impurities in the production process (such as raw material contamination, equipment wear, environmental interference) must be investigated. After rectification, the sample must be retested.

[0088] 3. Data Report: The test result report must include the following: sample information (name, production batch, sampling date, sampling site), experimental environment parameters, instrument information (model, calibration status), reagent information (specification, batch number), a brief description of the operating procedures, test data (preliminary concentration, error correction concentration, blank correction concentration, parallel experimental data), content calculation process, judgment result, operator, and test date; the report must be filled out in a standardized manner, the data must be true, accurate, and traceable, and it will be effective after being stamped with the official testing seal.

[0089] Step 11: Method Validation and Archiving 1. Method Validation: To ensure the accuracy, reliability, and applicability of this detection method, spike recovery experiments, limit of detection validation, and precision validation were conducted. The validation results are as follows: (1) Spike recovery experiment: Take the finished product sample (Y-01) with known content, and add iron, cobalt and nickel standard solutions at low, medium and high concentration levels respectively. Prepare 3 parallel samples for each concentration level, complete the detection according to the above detection steps, and calculate the spike recovery rate. The data are as follows (retain 1 decimal place): Low concentration spikes (50 μg / kg iron, 20 μg / kg cobalt, 20 μg / kg nickel): iron recovery rate 98.2%, cobalt recovery rate 98.5%, nickel recovery rate 98.3%; Medium concentration spikes (80 μg / kg iron, 35 μg / kg cobalt, 35 μg / kg nickel): iron recovery rate 99.1%, cobalt recovery rate 99.3%, nickel recovery rate 99.2%; High-concentration spikes (100 μg / kg iron, 50 μg / kg cobalt, 50 μg / kg nickel): iron recovery rate 98.8%, cobalt recovery rate 99.0%, nickel recovery rate 98.9%; The spiked recoveries ranged from 98.2% to 99.3%, all within the range of 95% to 105%, meeting the accuracy requirements of the method. (2) Validation of the lower limit of detection: The blank solution was measured 11 times consecutively, and the standard deviation of the blank concentration σblank was calculated. The lower limit of detection (LOD) = 3 × σblank, and the lower limit of quantitation (LOQ) = 10 × σblank. The validation results are as follows (unit: ppb): Iron: σ blank = 0.004 ppb, LOD = 0.012 ppb ≤ 0.1 ppb, LOQ = 0.040 ppb; Cobalt: σ blank = 0.003 ppb, LOD = 0.009 ppb ≤ 0.1 ppb, LOQ = 0.030 ppb; Nickel: σ blank = 0.003 ppb, LOD = 0.009 ppb ≤ 0.1 ppb, LOQ = 0.030 ppb; The detection limit is ≤0.1ppb, which meets the requirements for the detection of trace magnetic substances; (3) Precision verification: Based on parallel experimental data, the method precision is expressed by RSD, and the verification results are as follows: The precision RSD of the same batch is ≤0.14%, the precision RSD of different batches is ≤0.17%, and the precision RSD of different personnel is ≤0.09%, all ≤2.5%, which meets the method precision requirements; Method validation conclusions: The spiked recovery rate, detection limit, and precision of this detection method all meet the requirements. It can be used for the accurate detection of trace magnetic substances (iron, cobalt, nickel) in lithium carbonate, and is suitable for battery-grade and industrial-grade lithium carbonate samples, as well as the detection of raw materials and intermediates in production.

[0090] This method addresses the core pain points of existing technologies for detecting magnetic substances in lithium carbonate. By organically combining gradient magnetic field separation and enrichment, ultrasonic-assisted purification, precise ICP-MS detection, and multivariate statistical error correction, it creatively solves five major problems existing in current technologies. Specifically: First, it solves the problem of high detection limits in existing methods, which are mostly between 10-50 ppb, failing to meet the detection requirements of battery-grade lithium carbonate for trace magnetic substances below the ppb level. This method improves the separation and recovery rate of magnetic substances to over 99% through multi-cycle enrichment with a gradient magnetic field. Combined with ultrasonic-assisted purification to remove matrix interference, ICP-MS parameter optimization, and multivariate statistical error correction, the detection limit is reduced to below 0.1 ppb. This allows for the accurate detection of trace (below the ppb level) magnetic substances such as iron, cobalt, and nickel in lithium carbonate, effectively identifying potential quality risks posed by trace impurities and filling the gap in the field of magnetic substance detection below the ppb level. Secondly, this method solves the problems of numerous interfering factors and large detection errors in existing methods. Existing methods only use simple blank correction, which cannot eliminate systematic and random errors from multiple sources such as weighing error, magnetic separation recovery rate error, matrix interference, and instrument drift. This results in relative standard deviations (RSD) of the detection results often exceeding 5%. This method, verified by ultrasonic-assisted washing combined with a UV-Vis spectrophotometer, thoroughly removes matrix impurities (Li). + Cl -This method addresses both magnetic and non-magnetic impurities. By combining a multivariate statistical error correction model, the influence weight of each error source is quantified, achieving precise error correction. Simultaneously, a blank experiment is used for secondary correction to further eliminate interference from reagents and instrument tubing, ensuring an RSD of ≤2.5% and significantly improving the accuracy and repeatability of the test results. Thirdly, it solves the problems of complex operation and low detection efficiency in existing methods. Current gravimetric methods take over 8 hours per test, and magnetic separation-atomic absorption spectrometry takes over 6 hours, failing to meet the needs of industrial-scale batch testing. This method optimizes the operation process of each step, reasonably compressing the time spent on sample pretreatment, magnetic separation, washing, and determination, controlling the time for a single test to within 4 hours. Furthermore, some steps can be performed in parallel (e.g., simultaneous ultrasonic washing of multiple samples and continuous ICP-MS determination), greatly improving detection efficiency, meeting the needs of industrial-scale batch testing, and reducing detection costs. Fourth, it solves the problem that existing methods cannot achieve precise separation and enrichment of magnetic materials. Existing magnetic separation methods mostly use a single magnetic field strength, which is difficult to completely separate strong magnetic (iron, cobalt) and weak magnetic (Fe2O3, NiO) impurities, resulting in some magnetic impurities not being detected and the detection results being low. This method uses a self-designed gradient magnetic field separation device. Through a continuous gradient change of magnetic field strength from 0.1T to 1.5T and three magnetic field cycles, impurities with different magnetic strengths can be adsorbed in sequence, achieving thorough separation and enrichment of magnetic materials, ensuring that no magnetic impurities are missed, and solving the pain point of incomplete separation in existing methods. Fifth, this method addresses the shortcomings of existing publicly available detection methods, which fail to organically combine magnetic separation enrichment, matrix interference elimination, and statistical error correction. These methods either prioritize separation at the expense of detection accuracy or focus on detection while neglecting the thoroughness of impurity separation. This method deeply integrates these three elements, designing a synergistic mechanism of gradient magnetic field multi-cycle enrichment, ultrasonic-assisted purification, and multivariate statistical error correction. This forms a completely new detection approach, overcoming the limitations of existing technologies. It also solves the problems of poor versatility and limited application to single scenarios found in existing methods. This method is applicable to the detection of lithium carbonate samples of different grades, including battery-grade and industrial-grade samples, and can detect various magnetic substances such as iron, cobalt, and nickel. It is highly versatile and has a wide range of applications. Furthermore, through standardized method validation and data archiving, this method solves the problems of lack of traceability and inability to achieve industrial-scale promotion found in existing methods. This provides reliable technical support for quality control in the lithium carbonate industry and promotes its high-quality development.

[0091] The embodiments described above are merely examples of several implementations of the present invention, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of the present invention. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the present invention, and these all fall within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the claims.

Claims

1. A method for detecting the content of magnetic substances in lithium carbonate, characterized in that, Includes the following steps: S1: Steps for sample collection and pretreatment; S2: The steps for accurately weighing and adjusting the volume of the sample; S3: Steps for gradient magnetic field separation and enrichment; S4: The step of performing ultrasonic-assisted washing and purification of the enrichment solution; S5: Steps for precipitating, dissolving, and adjusting the volume of magnetic materials; S6: Steps for ICP-MS instrument calibration and parameter optimization; S7: Steps for determining the mother liquor and correcting for multivariate statistical errors; S8: Steps for performing blank experiment correction; S9: Steps for conducting parallel experimental verification; S10: Steps for calculating and judging the results; S11: Steps for method verification and archiving.

2. The method for detecting the content of magnetic substances in lithium carbonate according to claim 1, characterized in that, Step S1 specifically includes: Lithium carbonate samples were collected and pretreated to remove physical impurities, preventing interference with subsequent magnetic separation and detection. This pretreatment also ensured sample homogeneity, laying the foundation for accurate weighing and testing. The procedure was as follows: Using a sterile sampler, samples were collected from the upper, middle, and lower parts of the lithium carbonate finished product silo, with three samples collected from each part. Each sample was at least 50g. Direct contact between the sampler and metal equipment was avoided during sampling to prevent the introduction of external magnetic impurities. The nine collected samples were combined and placed in a clean agate mortar. The samples were then manually ground, with the grinding force controlled at 50-80N. The grinding time is 15-20 minutes until all the sample passes through a 200-mesh nylon sieve. The sieved sample is then placed in a vacuum drying oven, with the drying temperature set at 105±2℃ and the vacuum degree at 0.08-0.10MPa for 2 hours. The purpose is to remove adsorbed and free water from the sample to avoid moisture affecting the subsequent magnetic separation efficiency and the accuracy of the detection results. After drying, the sample is taken out and placed in a desiccator to cool to room temperature for about 30 minutes to prevent the sample from absorbing moisture. After cooling, the sample is reduced to 10g using the quartering method to obtain a uniform sample for testing. This sample is then placed in a sterile sealed bag for later use. The sample is numbered and recorded.

3. The method for detecting the content of magnetic substances in lithium carbonate according to claim 2, characterized in that, Step S2 specifically includes: Accurately weigh the sample to be tested and dissolve it in a specific solvent to prepare a homogeneous sample mother liquor. This provides a quantitative sample system for subsequent magnetic separation and enrichment, ensuring the repeatability and comparability of the test results. The operation method is as follows: Take the sample to be tested after pretreatment in step S1. Use a 0.0001g electronic balance (calibrated to 0.0001g) for weighing. Before weighing, preheat the balance for 30 minutes and calibrate the zero point to avoid the influence of ambient temperature (controlled at 25±2℃) and humidity (controlled at 50±5%RH) on the weighing accuracy. Use the difference method for weighing. First, weigh the clean 50mL volumetric flask and record it as m0. Then, use an agate spatula to take an appropriate amount of sample and put it into the volumetric flask. Weigh the total mass of the volumetric flask and the sample again and record it as m1. Sample mass m = m1 - For sample weighing, control the sample mass to be 2.0000±0.0002g, and weigh three parallel samples to ensure a weighing error ≤0.01%. Slowly add 20mL of hydrochloric acid solution (1+1 concentration) to the volumetric flask containing the weighed sample. The hydrochloric acid solution is prepared by mixing analytical grade hydrochloric acid and ultrapure water at a volume ratio of 1:

1. The resistivity of the ultrapure water should be ≥18.2MΩ·cm to avoid interference from impurities in the water. Gently shake the volumetric flask while adding the solution to ensure complete dissolution of the sample and avoid incomplete dissolution due to excessively high local concentrations. After complete dissolution, observe that there is no white precipitate or turbidity in the volumetric flask. Rinse the inner wall of the volumetric flask 3-4 times with ultrapure water, ensuring all the rinsing solution flows back into the flask. Then, dilute to the mark with ultrapure water, shake well, and let stand for 10 minutes to obtain the sample mother liquor, labeled as Mother Liquor A. Prepare three parallel portions of Mother Liquor A, labeled A1, A2, and A3, for later use.

4. The method for detecting the content of magnetic substances in lithium carbonate according to claim 3, characterized in that, Step S3 specifically includes: Utilizing the differences in magnetic strength among different magnetic materials, a gradient magnetic field is employed to completely separate and enrich the magnetic materials in the sample mother liquor. This solves the problem that existing single magnetic field intensities cannot completely separate different magnetic impurities, improving the enrichment efficiency and separation purity of magnetic materials. This lays the foundation for lowering the detection limit and improving detection accuracy in subsequent detection processes. The operation method is as follows: A self-designed gradient magnetic field separation device is used. This device includes a magnetic field generator, separation column, collection tank, and temperature control system. The magnetic field strength is continuously adjustable, with an adjustment range of 0.1-1.5T. First, the separation device is cleaned and calibrated. The inner wall of the separation column is rinsed with ultrapure water 3-4 times to remove residual impurities. Then, the magnetic field strength is calibrated to ensure that the error of the magnetic field strength is ≤±0.01T. The mother liquors A1, A2, and A3 prepared in step S2 are poured into three clean separation columns, with 20mL of mother liquor added to each column. The flow rate of the separation columns is controlled at 1.0-1.5mL / min, and the flow rate is adjusted by a peristaltic pump with a speed of 50-80r / min. min; Turn on the magnetic field generator and set the gradient magnetic field parameters: the initial magnetic field strength is 0.1T, then it is increased uniformly to 1.5T at a rate of 0.05T / min, maintained at 1.5T for 5 minutes, and then decreased to 0.1T at a rate of 0.1T / min to complete one gradient magnetic field cycle. A total of 3 gradient magnetic field cycles are performed to ensure that impurities with different magnetic intensities, such as strongly magnetic iron and cobalt, and weakly magnetic Fe2O3 and NiO, can be adsorbed by the magnetic field. During the magnetic field cycle, the temperature of the separation column is controlled at 25±1℃ through the temperature control system to avoid the temperature being too high or too low, which would affect the magnetic field strength and the adsorption effect of the magnetic substances. After the magnetic separation is completed, turn off the magnetic field generator and let it stand for 5 minutes to allow the magnetic substances adsorbed on the inner wall of the separation column to completely fall off and flow into the collection tank to obtain the magnetic substance enrichment liquid, labeled as enrichment liquid B. Three enrichment liquids B1, B2, and B3 are obtained in parallel. The remaining mother liquor in the separation column is collected and labeled as waste liquid for subsequent matrix interference analysis.

5. The method for detecting the content of magnetic substances in lithium carbonate according to claim 4, characterized in that, Step S4 specifically includes: To remove lithium carbonate matrix impurities and other non-magnetic impurities from enrichment solution B, and to avoid matrix interference with ICP-MS detection in subsequent steps, thereby improving the purity of the enrichment solution and ensuring the accuracy of the detection signal, the following procedure is followed: Pour the three enrichment solutions B1, B2, and B3 obtained in step 3 into three 50mL centrifuge tubes respectively. Add 10mL of ultrapure water (resistivity ≥18.2MΩ·cm) to each centrifuge tube and gently shake the tubes for 1 minute to ensure thorough mixing. Place the centrifuge tubes in an ultrasonic cleaner, setting the ultrasonic power to 200-250W, the ultrasonic frequency to 40kHz, and the ultrasonic time to 15 minutes. During ultrasonication, maintain the water temperature in the ultrasonic cleaner at 25±2℃ to prevent oxidation of magnetic materials due to excessively high temperatures, which could affect subsequent detection results. After ultrasonication, place the centrifuge tubes in a high-speed centrifuge. The centrifugation speed was set to 8000 r / min, the centrifugation time to 10 minutes, and the centrifugation temperature to 25℃, so that the magnetic material precipitated at the bottom of the centrifuge tube, while the non-magnetic impurities and matrix impurities dissolved in the supernatant. After centrifugation, the supernatant was slowly aspirated with a pipette, avoiding touching the bottom precipitate, and poured into a waste liquid collection bottle, retaining the magnetic material precipitate at the bottom of the centrifuge tube. 10 mL of ultrapure water was added to each centrifuge tube again, and the above operation of sonication, centrifugation, and supernatant aspiration was repeated 3 times until the supernatant was colorless and transparent. The absorbance of the supernatant in the wavelength range of 200-800 nm was detected by a UV-Vis spectrophotometer and was ≤0.001, indicating that the matrix impurities and non-magnetic impurities had been completely removed. After washing, the magnetic material precipitate at the bottom of the centrifuge tube was retained and labeled as precipitate C. Three parallel precipitates, C1, C2, and C3, were obtained for later use.

6. The method for detecting the content of magnetic substances in lithium carbonate according to claim 5, characterized in that, Step S5 specifically includes: High-purity magnetic precipitate C was completely dissolved in a specific solvent to prepare a homogeneous detection stock solution, converting the magnetic material into an ionic state for subsequent ICP-MS detection, ensuring the stability and accuracy of the detection signal. The operation method is as follows: 5 mL of nitric acid solution (5% concentration) was added to the centrifuge tubes corresponding to the three precipitates C1, C2, and C3 obtained in step 4. The nitric acid solution was prepared by mixing analytical grade nitric acid and ultrapure water at a volume ratio of 1:

19. Analytical grade nitric acid avoids the introduction of impurities, and the low concentration of nitric acid ensures complete dissolution of the magnetic material while avoiding matrix interference in subsequent ICP-MS detection due to excessive nitric acid. The centrifuge tubes were placed in a constant temperature water bath at 60±2℃ for 20 minutes. During the water bath, the centrifuge tubes were gently shaken every 5 minutes to promote the dissolution of the magnetic precipitate. The dissolution was observed until the precipitate was completely dissolved and no black or brown precipitate residue remained. If the precipitate is not completely dissolved, add analytical grade nitric acid dropwise, 0.1 mL each time, until the precipitate is completely dissolved, and record the total amount of nitric acid added. After the precipitate is completely dissolved, remove the centrifuge tube from the constant temperature water bath and place it in a fume hood to cool to room temperature for about 20 minutes to avoid excessive temperature causing solution evaporation and affecting the accuracy of the concentration. Slowly transfer the cooled solution to a 25 mL volumetric flask and rinse the inner wall of the centrifuge tube with ultrapure water 3-4 times, ensuring that all the rinsing solution flows into the volumetric flask to ensure that there is no magnetic substance residue. Dilute the solution in the volumetric flask to the mark with ultrapure water, shake well, and let stand for 5 minutes to obtain the magnetic substance detection stock solution, labeled as detection stock solution D. Prepare three parallel detection stock solutions D1, D2, and D3 for later use. At the same time, prepare a blank solution: take a 25 mL volumetric flask, add 5 mL of nitric acid solution of the same concentration, and dilute to the mark with ultrapure water, labeled as blank solution D0, for subsequent blank correction.

7. The method for detecting the content of magnetic substances in lithium carbonate according to claim 6, characterized in that, Step S6 specifically includes: The ICP-MS instrument was calibrated and its parameters optimized to ensure optimal operating conditions, improve detection sensitivity and accuracy, and reduce systematic errors introduced by the instrument itself. The procedure was as follows: An inductively coupled plasma mass spectrometer was selected. Before detection, the instrument was preheated for 60 minutes, with argon gas supplied at a flow rate of 15 L / min to ensure no air residue remained in the instrument's internal tubing. After preheating, the instrument was calibrated using the external standard method. A series of standard solutions with varying concentrations were prepared: iron, cobalt, and nickel standard stock solutions were accurately pipetted to a concentration of 1000 μg / mL (analytical grade), diluted with 5% nitric acid solution to prepare a 0.1 pp. b. Prepare a series of standard solutions at concentrations of 1 ppb, 10 ppb, 50 ppb, and 100 ppb, with three parallel samples for each concentration gradient. Simultaneously, prepare blank standard solutions. Import the standard solutions and blank standard solutions sequentially into the ICP-MS instrument and perform measurements according to the set parameters. Record the signal intensity corresponding to each concentration of standard solution. Plot a standard working curve with standard solution concentration as the x-axis and signal intensity as the y-axis. Calculate the correlation coefficient R² of the standard working curve. R² should be ≥ 0.9999. If R² < 0.9999, the standard solutions need to be prepared again until the requirement is met. After the standard curve calibration is completed, optimize the instrument parameters. The optimization ranges and optimal values ​​for parameters including RF power, nebulizer gas flow rate, auxiliary gas flow rate, sampling depth, and integration time are as follows: RF power optimization range: 1200-1500W, optimal value: 1350W. Too low a power will lead to plasma instability and weak signal strength, while too high a power will lead to enhanced matrix interference. Nebulizer gas flow rate optimization range: 0.8-1.2L / min, optimal value: 1.0L / min. Too low a flow rate will lead to incomplete atomization and unstable signal strength, while too high a flow rate will lead to sample dilution and decreased sensitivity. Auxiliary gas flow rate optimization range: 0.2-0.5L / min, optimal value: 0.3L / min. Gas can stabilize plasma, but improper flow rate can lead to plasma extinguishing. The optimal sampling depth range is 8-12 mm, with an optimal value of 10 mm. Sampling depth affects signal strength and interference levels; too deep or too shallow a sampling depth will increase interference. The optimal integration time range is 0.1-0.5 s, with an optimal value of 0.3 s. Too short an integration time will lead to large signal fluctuations, while too long an integration time will reduce detection efficiency. After parameter optimization, the signal intensity of the standard series solutions should be measured again to verify the instrument's stability. The relative standard deviation should be ≤2%. If RSD > 2%, the parameters need to be re-optimized until the requirement is met. After instrument calibration and parameter optimization, the parameter settings should be saved for later use.

8. The method for detecting the content of magnetic substances in lithium carbonate according to claim 7, characterized in that, Step S7 specifically includes: The signal intensity of magnetic substances in the detection stock solution was measured using an optimized ICP-MS instrument. Multivariate statistical methods were used to correct errors in the detection results, eliminating systematic and random errors in each step of the detection process and improving the accuracy and reliability of the results. The operation method is as follows: The three sample stock solutions D1, D2, and D3 prepared in step S5, and the blank solution D0, were sequentially imported into the ICP-MS instrument according to the optimized instrument parameters in step S6. Each sample was measured in parallel three times, and the signal intensity of each measurement was recorded. The average signal intensity of the stock solutions D1, D2, and D3 was recorded as S1, S2, and S3, and the average signal intensity of the blank solution D0 was recorded as S0. Based on the standard working curve plotted in step S6, the average signal intensity values ​​S1, S2, S3, and S0 were converted into corresponding concentration values ​​to obtain the preliminary detection concentration of the stock solutions, recorded as C1, C2, and C3, and the concentration of the blank solution was recorded as C0. The formula for calculating the preliminary detection concentration is: C i = k×S i + b (i=1,2,3,0), where k is the slope of the standard working curve and b is the intercept of the standard working curve; a multivariate linear regression combined with a deviation correction model is used for error correction to eliminate systematic and random errors in the detection process. A multivariate statistical error correction model is established, and the specific formula is as follows: ； The meanings of each symbol are as follows: The final detected concentration of the magnetic material after error correction is the core output value of this step; The regression constant term represents the basic bias in the detection process. It is obtained through partial least squares regression combined with Bayesian regularization, with a regularization parameter introduced during the fitting process. This is used to avoid model overfitting; The concentration regression coefficient represents the weight of the initial detected concentration on the final detected concentration. This indicates a positive correlation between the preliminary detection concentration and the final detection concentration. The larger the weight, the more significant the influence of the preliminary detection concentration. The solution formula is as follows: , This represents the average concentration from preliminary testing. The regression coefficient for weighing error represents the weight of the influence of sample weighing error in step S2 on the final detected concentration. The sign of the value is determined by the direction of the weighing error; when the weighing value is too large, When the weighing value is too small, Its solution requires combining the normal distribution probability density function of the weighing error and introducing a correction factor. ; The regression coefficient for magnetic separation recovery rate represents the weight of the impact of the magnetic separation recovery rate in step S3 on the final detected concentration. The higher the magnetic separation recovery rate, the closer the final detected concentration is to the true value. Avogadro's constant is introduced during the solution process. As a correction term for trace substance measurement; The matrix interference regression coefficient represents the weight of the influence of residual matrix impurities after washing in step S4 on the final detection concentration. The sign is determined by the type of matrix interference; in the case of ion enhancement effect, During ion suppression effect, Planck's constant is introduced when solving the problem. As a correction term for microscopic particle interference; To determine the preliminary concentration of the mother liquor; The sample weighing error is calculated using the following formula: m is the actual weighed mass, and m0 is the standard weighed mass of 2.0000g. The composite correction coefficient for weighing error conforms to the superposition characteristics of the normal distribution and the log-normal distribution of precision weighing; The magnetic separation recovery rate is calculated using the following formula: , The mass of the magnetic material enriched in step S3, The true mass of the magnetic material in the sample was determined by a combination of spiked recovery experiments and isotope labeling. e is the natural constant, approximately 2.71828. This is a temperature correction term for recovery rate, used to compensate for the effects of temperature fluctuations during the magnetic separation process; The residual amount of matrix impurities was determined by UV-Vis spectrophotometry in the supernatant after washing in step S4, using a matrix impurity residual amount. + The concentration was obtained, and the corrected formula is: , These are measured values, and γ is the Euler-Macheroni constant, approximately 0.5772. This represents the standard deviation of matrix impurity detection. This is a random correction term for matrix interference; The standard deviation of the weighing error is obtained by combining parallel weighing experiments with Bayesian estimation and fitting, and follows a normal distribution. The mean of the standard deviation, The standard deviation is the variance. To improve the cumulative distribution function of the normal distribution, a logarithmic correction term is introduced. The nonlinear cumulative effect of quantitative weighing error demonstrates the expertise of advanced statistical error correction. This integral can be numerically solved using the Gauss-Legendary integral formula: , For the roots of the Legendre polynomial, These are the weighting coefficients; It is a higher-order Gaussian correction term that combines three constants, π, e, and γ, to suppress the interference of abnormal weighing errors on the final result, while compensating for the systematic deviation caused by environmental humidity. The logarithmic-constant composite correction term is used to compensate for the nonlinear influence of magnetic separation recovery rate, while the Avogadro constant correction term is used to compensate for the deviation in the counting of trace magnetic particles. As a higher-order correction term for matrix interference, the influence weight of matrix interference is adjusted by Euler-Marcheroni constant, Planck constant and quadratic radical, while compensating for the detection bias caused by microscopic ion collisions; The random error term follows a normal distribution. , Let Variance be the variance of random errors, representing unpredictable accidental errors during the detection process, and satisfying the following conditions: n is the number of parallel measurements. The average value of random error. This is the variance correction factor; The bias weighting factor for the i-th parallel determination is calculated using the following formula: Used to adjust the weight of the influence of random error. This is used as a weighting term for random errors to further improve the accuracy of error correction. The establishment and solution process of the error correction model is as follows: Fifteen groups of spiked samples with different concentrations were prepared through a spike recovery experiment. The spike concentrations ranged from 0.1 to 100 ppb, covering three concentration ranges: low, medium, and high. Five samples were prepared for each range. The samples were then analyzed according to steps S1 to S7 of this method to obtain the preliminary detection concentration of each spiked sample. Weighing error Magnetic separation recovery rate Residual amount of matrix impurities And the actual concentration of the spiked sample, denoted as The true concentration was determined by isotope dilution mass spectrometry to ensure the accuracy of the true value; As the dependent variable, with , , , Using partial least squares regression combined with Bayesian regularization as the independent variable, the multiple linear regression equation was fitted, and a regularization parameter was introduced. Weighting coefficient , Let be the standard deviation of the detection of the i-th spiked sample. The weight decay factor reduces the impact of outliers on the fitting results, and the regression constant term is obtained by solving for it. and regression coefficients , , , The fitting process satisfies Regularization term Used to avoid model overfitting and improve the model's generalization ability; at the same time, the model's determination coefficient is calculated. Adjusting the coefficient of determination Root mean square error and mean absolute error, among which , (where k is the number of independent variables, k=4). Three constants, π, e, and γ, are introduced for accuracy correction. ,Require , RMSE ≤ 0.03ppb and MAE ≤ 0.02ppb are used to ensure the good fit and prediction accuracy of the error correction model. Furthermore, the significance of the model is verified through the F-test; the F-statistic is calculated using the following formula: The requirement is F ≥ 10000, significance level α = 0.01, to prove that the model is statistically significant; the results obtained in step S7... , , , Substituting into the error correction model, the final detected concentration of the magnetic material after error correction is calculated. , , These correspond to three sample mother solutions, D1, D2, and D3, respectively; calculate the average value of the three final test concentrations, denoted as . and relative standard deviation, where γ, π, and e are introduced to correct for minor deviations and compensate for systematic biases in parallel measurements. To optimize the accuracy of deviation quantification, e, γ, and π are introduced to meet the statistical requirements of precision testing. The requirement is that RSD ≤ 1.5%. If RSD > 1.5%, the test mother liquor needs to be re-measured, and the error correction model parameters need to be checked until the requirements are met.

9. The method for detecting the content of magnetic substances in lithium carbonate according to claim 1, characterized in that, Step S8 specifically includes: The test results are further corrected through blank experiments to eliminate interference from trace magnetic impurities in reagents, ultrapure water, and instrument tubing, ensuring the authenticity and accuracy of the test results and providing a basis for parallel experimental verification in step S9. The operation method is as follows: Prepare 5 blank solutions in parallel according to the preparation method in step S5, labeled as D. 01 D 02 D 03 D 04 D 05 The preparation process of the blank solution was exactly the same as that of the test stock solution, but without adding magnetic precipitate C. 5 mL of 5% nitric acid solution and ultrapure water were added to ensure that the preparation environment, reagents, and instruments for the blank solution were identical to those for the test stock solution. Five blank solutions were imported into the ICP-MS instrument according to the optimized instrument parameters in step S6 for measurement. Each blank solution was measured in triplicate, and the signal intensity of each measurement was recorded. The average signal intensity of each blank solution was calculated and denoted as S. 01 S 02 S 03 S 04 S 05 Based on the standard working curve plotted in step S6, the average signal intensity of the blank solution is converted into the corresponding concentration value, denoted as C. 01 C 02 C 03 C 04 C 05 Five blank solution concentration values ​​were obtained. Outliers in the blank solution concentration values ​​were removed using the Grubbs criterion (significance level α = 0.05). The average concentration value of the remaining blank solutions was calculated and denoted as . And standard deviation, denoted as ,Require ≤ 0.05ppb, if If the concentration is >0.05 ppb, the reagents and ultrapure water need to be replaced, a blank solution needs to be prepared again, and the measurement needs to be repeated until the requirements are met; the average detection concentration obtained in step S7 is then used. Subtract the average concentration of the blank solution The concentration of the magnetic material after blank correction is obtained and denoted as . The correction formula is: ;like A value <0 indicates that the concentration of the blank solution is higher than the concentration of the test stock solution. The test process needs to be checked again, and the measurement should be repeated after eliminating interference. Step S9 specifically includes: The repeatability and stability of this detection method were verified through parallel experiments to ensure accurate and reliable detection results in different batches and samples. The operation method is as follows: Following the complete procedure from S1 to S8, three batches of lithium carbonate samples were collected again from the same batch as the samples collected in step S1 to ensure sample consistency. Three sample stock solutions were prepared in parallel for each batch, for a total of nine sample stock solutions. The nine sample stock solutions were subjected to ICP-MS determination, multivariate statistical error correction, and blank experiment correction to obtain the magnetic substance concentrations after blank correction, denoted as . , ... Calculate the average of the 9 concentration values, and record it as .

1. Relative standard deviation and spiked recovery rate; 2. Method for determining spiked recovery rate: Add known concentrations of iron, cobalt, and nickel standard solutions to the sample weighed in step S2 (spiking concentrations are 0.5 times, 1.0 times, and 1.5 times the detection concentration, respectively). Perform the detection according to steps S3 to S8, and calculate the spiked recovery rate using the formula: Spiked recovery rate (%) = (Detection concentration after spiking - Detection concentration without spiking) / Spiked concentration × 100%. Perform parallel determinations at each spiked concentration three times and calculate the average spiked recovery rate.

3. Validation criteria: The RSD of parallel experiments should be ≤ 2.5%, and the spiked recovery rate should be between 95% and 105%. If the validation criteria are not met, the cause must be investigated, and parallel experiments should be repeated until the requirements are met. Record all data from the parallel experiments, including detection concentration, error correction value, blank correction value, RSD, spiked recovery rate, etc., and generate a parallel experiment report.

10. The method for detecting the content of magnetic materials in lithium carbonate according to claim 9, characterized in that, Step S10 specifically includes: Based on the blank-corrected detection concentration and sample weighing mass, calculate the actual content of magnetic substances in the lithium carbonate sample. Combined with industry standards and customer requirements, determine the sample quality. The operation method is as follows: Based on the sample weighing mass in step S2 (m = 2.0000 g), the final volume of the detection mother liquor in step S5 (V1 = 25 mL), the final volume of the sample mother liquor in step S2 (V2 = 50 mL), and the volume of the magnetic separation enrichment liquid in step S3 (V3 = 20 mL), calculate the actual content of magnetic substances in the lithium carbonate sample, expressed as a mass fraction ω in ppb. The calculation formula is: ； The meanings of each symbol are as follows: This represents the actual mass fraction of magnetic material in the lithium carbonate sample. V1 is the total average concentration after blank correction of the parallel experiments obtained in step S9; V2 is the final volume of the mother liquor detected in step S5; m is the final volume of the sample mother liquor in step S2; V3 is the mass of the sample weighed in step S2; V4 is the volume of the magnetic separation enrichment solution in step 3; 10 - ³ is a unit conversion factor, converting μg / g to ppb. 1 ppb = 1 μg / kg = 10 - ³μg / g; Pi is used as a precise correction factor for volume-mass conversion, to compensate for minute volume deviations during the volume determination and enrichment process, and to improve calculation accuracy. is the Euler-Marcheroni constant, used to correct systematic deviations in the weighing and volume adjustment processes, while also adjusting the weights of each correction term; e is the natural constant, used to compensate for concentration deviations caused by solution evaporation and adsorption. It is a constant composite correction factor, which further improves the rigor of the calculation; Let Avogadro's constant be 1. This is a correction term for trace substance measurement, used to compensate for minor deviations in the counting process of magnetic material particles; Let be Planck's constant. This is a microscopic interference correction term, used to compensate for minor detection biases caused by ion collisions and excited-state particle transitions during ICP-MS detection. The volumetric temperature correction factor is calculated using the following formula: , The coefficient of volume expansion of the solution is approximately 2.1 × 10⁻⁶. -4 / ℃, where T is the ambient temperature of the experiment in ℃, used to compensate for the effect of temperature fluctuations on the solution volume and ensure the accuracy of the volume parameters; For higher-order calculus correction terms, the integration variable V is the solution volume, and a logarithmic correction term is introduced. The nonlinear change in magnetic substance concentration in the quantified sample mother liquor from enrichment volume to fixed volume is investigated. This integral is solved using integration by parts combined with Gauss-Legendal numerical integration. The specific solution process is as follows: Let , ,but , According to the integral by parts formula , can be obtained The second integral is solved using the Gauss-Legend de Gauss numerical integration formula, with n=5 nodes to ensure an integration accuracy of ≤10. -8 ppb; calculate the actual content of the three magnetic materials, iron, cobalt, and nickel, respectively. , , Each substance's calculation incorporates a unique combination of higher-order correction coefficients and constants: Introducing correction terms during calculation The calculation formula is: ; Introducing correction terms during calculation The calculation formula is: ; Introducing correction terms during calculation The calculation formula is: The formula for calculating the total content of the three magnetic substances is: , This is a comprehensive correction factor for the total content, used to balance the detection biases of the three substances. This is a micro-correction term for the total content, compensating for the mutual interference among the three magnetic materials; it is used to determine sample quality: battery-grade lithium carbonate requires... ≤10ppb ≤ 5ppb ≤ 5ppb ≤ 20ppb; Industrial grade lithium carbonate requirements ≤ 50ppb ≤20ppb ≤ 20ppb ≤ 90ppb; If the sample test results meet the standard requirements of the corresponding level, the sample is deemed qualified; otherwise, the sample is deemed unqualified; For unqualified samples, the reasons for non-compliance must be analyzed and rectification suggestions must be proposed; The result calculation process and judgment results must be recorded to form a test report. The test report must include sample information, test steps, test data, error correction data, parallel experiment data, result calculation, judgment conclusion, and rectification suggestions to ensure the completeness and traceability of the test report; Step S11 specifically includes: The scientific validity, accuracy, stability, and applicability of the entire detection method were comprehensively validated. All data and reports from the testing process were systematically archived to support the method's widespread application and subsequent retesting. The operating procedures are as follows: Method validation is divided into indoor and outdoor validation. Indoor validation was conducted by professional laboratory testing personnel. Following steps S1 to S10 of this method, three standard samples with different concentrations were tested: low concentration (0.5 ppb), medium concentration (10 ppb), and high concentration (50 ppb). Each concentration was measured in parallel six times. The relative standard deviation, spiked recovery rate, and limit of detection (LOD) were calculated. Validation criteria: RSD ≤ 2.5%, spiked recovery rate 95%-105%, LOD ≤ 0.1 ppb. If the validation criteria are not met, the method parameters need to be optimized until they are met. Outdoor validation was commissioned to three third-party testing institutions, providing identical lithium carbonate samples and method operation manuals. The third-party institutions conducted independent tests according to this method. The results of the third-party institutions were compared with those of our laboratory, and the relative deviation was calculated. The relative deviation should be ≤3% to ensure the applicability and universality of the method, avoiding the limitation of the method to a single laboratory and its inability to be widely applied. After the method validation is completed, all validation data should be compiled into a detection method validation report, which includes the validation purpose, validation plan, validation data, validation conclusions, optimization suggestions, etc., to ensure the traceability of the validation process. At the same time, all relevant materials, including sample collection records, pretreatment records, weighing records, magnetic separation records, ultrasonic washing records, ICP-MS measurement records, error correction data, blank experimental data, parallel experimental data, result calculation records, test reports, and method validation reports, should be classified and archived according to the archival management standards. Paper documents should be bound into a book, and electronic documents should be encrypted and stored. The archiving period should not be less than 5 years to facilitate subsequent re-inspection and quality traceability. Finally, the detection method should be summarized, and the key points and precautions of the method operation should be sorted out to form a method operation manual, which will provide guidance for industrial batch testing, daily laboratory testing, and personnel training, ensuring that different operators can obtain accurate and stable test results by following the manual.

Images