A nuclear-grade sponge zirconium and its preparation method

Abstract

This invention discloses a nuclear-grade sponge zirconium and its preparation method, belonging to the field of rare metal metallurgy and separation technology. The invention utilizes a Mannich-type reaction between polyethyleneimine and triethyl phosphate to generate a functionalized prepolymer containing aminophosphonate groups. This prepolymer is then reacted with an epoxy-based silane coupling agent to obtain a silanized bifunctional macromolecule. Finally, it is loaded onto the surface of a mesoporous silica support via a silanol condensation reaction to prepare an N-P bifunctional synergistic solid-phase adsorbent. In an acidic medium, the aminophosphonate groups on the adsorbent surface form a stable six-membered ring chelate complex with zirconium ions through the synergistic coordination of P=O and N-H. Hafnium ions, due to ionic radius contraction, experience coordination mismatch, leading to a decrease in the stability constant of the complex, thereby achieving selective interfacial adsorption and separation of zirconium and hafnium. The nuclear-grade sponge zirconium product obtained by this invention has a consistently low hafnium content (below 0.01%), meeting nuclear-grade standards.

Description

Technical Field

[0001] This invention relates to the field of rare metal metallurgy and separation technology, and in particular to a nuclear-grade sponge zirconium and its preparation method. Background Technology

[0002] Zirconium and its alloys possess extremely low thermal neutron absorption cross-sections (approximately 0.18 bar), excellent high-temperature mechanical properties, and corrosion resistance, making them irreplaceable nuclear fuel cladding materials in nuclear reactors. However, hafnium is always present in naturally occurring zircon minerals (such as zircon), typically comprising 2–3 wt% of zirconium. Hafnium has a thermal neutron capture cross-section as high as approximately 105 bar, about 500 times that of zirconium. Therefore, zirconium used in nuclear reactors must have its hafnium content strictly controlled below 0.01% (100 ppm) to ensure neutron economy. This high-purity zirconium is known as nuclear-grade sponge zirconium.

[0003] Since zirconium and hafnium both belong to Group IVB, have the same outer electron configuration, and their ionic radii are extremely similar due to the lanthanide contraction effect (Zr... 4+ 0.72Å, Hf 4+ The two atoms (0.71 Å) have highly similar chemical properties and are known as chemical twins in the metallurgical world, making deep separation extremely difficult.

[0004] Currently, the most widely used industrial method for separating zirconium and hafnium is solvent extraction, represented by the methyl isobutyl ketone-thiocyanate (MIBK-HSCN) system. Developed by Oak Ridge National Laboratory in 1949, this system has a history of over seventy years, and more than two-thirds of the world's nuclear-grade zirconium and hafnium are produced using it. The separation mechanism of the MIBK-HSCN system is based on the following: in a sulfuric acid medium, thiocyanate (SCN) ions... - The complex formed by hafnium ions and zirconium ions has a higher partition coefficient in the organic phase of MIBK than the corresponding complex of zirconium ions. Hafnium is preferentially extracted into the organic phase, while zirconium is retained in the aqueous raffinate, with a separation coefficient of about 5 to 10. However, this system has the following serious defects: (1) Thiocyanate is easily decomposed under acidic conditions, releasing highly toxic hydrogen cyanide gas, which poses a serious safety threat to operators and the environment; (2) MIBK has high solubility in water (about 2%), a low flash point (14°C), is flammable and explosive, and has high solvent consumption and high operating costs; (3) Emulsification is prone to occur during the extraction process, especially when processing high-concentration feed solutions, making phase separation difficult and affecting process continuity and product consistency.

[0005] To overcome the environmental and safety challenges of the MIBK-HSCN system, academia and industry have long been committed to developing alternative green separation technologies. The tributyl phosphate (TBP)-nitric acid system achieves separation through the selective coordination of TBP with zirconium, but this system requires a high-acidity environment (6–8 mol / L HNO3), leading to severe equipment corrosion. Furthermore, the radiolysis product of TBP, dibutyl phosphate, forms insoluble interfacial contaminants with zirconium, posing a risk of emulsification. Amine extractants (such as trioctylamine N235) have some separation effect on zirconium and hafnium in sulfuric acid media, but the separation coefficient is low (typically 3–5), requiring multi-stage series operation and a long process.

[0006] In recent years, solid-phase adsorption separation methods—especially extraction chromatography (extraction resins)—have been regarded as a promising alternative technology due to their combination of the high selectivity of solvent extraction and the ease of operation of ion exchange, without the problem of emulsification. The core of this type of method is to load functional molecules containing specific coordinating groups onto the surface of a solid support (such as silica gel, polymer microspheres, carbon aerogels, etc.), and to achieve selective adsorption by utilizing the coordination chemical reactions that occur at the solid-liquid interface.

[0007] Existing research has shown that: (1) aminophosphonate small molecule compounds (such as BEAP, CAMP, BAMP) have the ability to separate zirconium and hafnium in sulfuric acid medium, among which BEAP has a separation coefficient of up to 6.8. The mechanism is that the amino and phosphonate groups cooperate in coordination with metal ions; (2) solid-phase extractants prepared by loading phosphate ester extractants (such as P507) onto carbon aerogel carriers have the ability to separate zirconium and hafnium. 4+ and Hf 4+ The adsorption capacities were 16.8 mg / g and 12.0 mg / g, respectively, with a separation coefficient of approximately 3.3. However, the separation coefficient of existing solid-phase adsorption materials for zirconium hafnium is still far lower than the actual requirements for nuclear-level separation (Hf < 0.01%). The fundamental reason is that neither the impregnation method, which physically loads small molecule extractants onto the pores of the support, nor the chemical bonding of single functional groups to the surface of the support can fully utilize the synergistic effect of coordination chemistry. Summary of the Invention

[0008] The primary objective of this invention is to provide a method for preparing nuclear-grade sponge zirconium. This method employs a PEI-derived NP bifunctional polymeric solid-phase adsorbent, which achieves highly selective adsorption and separation of zirconium and hafnium through an interfacial bifunctional synergistic coordination recognition mechanism. This fundamentally solves the environmental and safety issues of traditional solvent extraction methods, while also overcoming the technical bottleneck of insufficient separation coefficient in existing solid-phase adsorption materials. This method can stably prepare nuclear-grade sponge zirconium products with a hafnium content of less than 0.01 wt%.

[0009] A second objective of this invention is to provide a high-purity nuclear-grade sponge zirconium prepared by the above method.

[0010] To achieve the above objectives, the present invention adopts the following technical solution: A method for preparing nuclear-grade sponge zirconium includes the following steps: Step 1: Preparation of Zirconium oxychloride feedstock solution Industrial-grade zirconium oxychloride (ZrOCl2·8H2O) was used as the raw material. It was dissolved in hydrochloric acid with a concentration of 0.5–3.0 mol / L to prepare an acidic aqueous solution containing 0.5–2.0 mol / L zirconium, which served as the raw material solution for separation. This raw material solution contained approximately 2–3 wt% hafnium, along with trace amounts of impurities such as iron, titanium, aluminum, and silicon.

[0011] Step 2: Preparation of NP bifunctional polymeric solid-phase adsorbent (a) Preparation of PEI functionalized prepolymer: Polyethyleneimine (molecular weight 600–10000 Da) was dissolved in anhydrous ethanol, and triethyl phosphate was slowly added dropwise at 60–80 °C under nitrogen protection. The mass ratio of PEI to triethyl phosphate was 1:0.5–1:2. The reaction was carried out for 8–12 hours to generate PEI functionalized prepolymer containing aminophosphonate groups. The reaction mechanism is as follows: Triethyl phosphate undergoes a nucleophilic substitution reaction (Mannich-type reaction) with the primary or secondary amine groups on the PEI molecular chain, removing one molecule of ethanol to generate a neutral aminophosphonate structure linked by PN bonds.

[0012] (b) Silanization modification: The PEI functionalized prepolymer obtained in step (a) is mixed with an epoxy silane coupling agent (preferably 3-(glycidoxypropyl)trimethoxysilane, KH-560) at a mass ratio of 1:0.3 to 1:1.5, and reacted at room temperature to 60°C for 4 to 8 hours. The epoxy groups of the silane coupling agent undergo ring-opening addition reactions with the residual amino groups in the prepolymer to generate a silanized bifunctional macromolecule containing a β-amino alcohol structure. This step not only introduces a trimethoxysilane anchoring group that can bond with the inorganic support, but also further enriches the functional group environment of the adsorbent with the newly generated hydroxyl groups.

[0013] (c) Loading onto a support: The silanized bifunctional macromolecule obtained in step (b) is added to a mesoporous silica support (specific surface area 200–600 m²). 2In a water-ethanol mixed solvent (water:ethanol = 1:1 to 1:4 v / v) with a particle size of 5–20 nm and a pore size of 50–200 μm, the pH was adjusted to 4–6 with dilute hydrochloric acid, and the reaction was carried out at 50–70 °C with stirring for 6–12 hours. Under these conditions, the trimethoxysilyl group of the silane coupling agent hydrolyzes to generate silanol groups, which undergo a condensation reaction with the silanol groups on the surface of the silica support to form stable Si-O-Si covalent bonds, firmly anchoring the bifunctional macromolecule to the support surface. After the reaction is complete, the mixture is filtered, washed successively with ethanol and deionized water, and vacuum dried at 60–80 °C for 12 hours to obtain the NP bifunctional polymer solid-phase adsorbent.

[0014] Step 3: Interfacial adsorption and separation of zirconium and hafnium from complex The NP bifunctional solid-phase adsorbent prepared in step two is packed into a jacketed glass or stainless steel chromatographic column with a column diameter ratio of 1:5 to 1:20. The zirconium oxychloride feed solution prepared in step one is passed through the adsorption column at a flow rate of 1 to 5 BV / h (bed volume / hour) and the operating temperature is 25 to 50 °C, which is maintained at a constant temperature using a jacketed water bath. Within the optimized operating pH range (0.5–1.5), the aminophosphonate groups on the adsorbent surface form stable six-membered ring chelate complexes with zirconium ions at the solid-liquid interface through the synergistic coordination of P=O and NH, thereby achieving selective adsorption of zirconium at the adsorbent interface. Due to the slightly smaller ionic radius of hafnium ions (a difference of about 0.01 Å), the coordination bond angle shifts by about 2–3° when forming chelate rings with the above-mentioned bifunctional ligands, resulting in increased chelate ring strain and a decrease in the stability constant of the complex by about 1–2 orders of magnitude. Therefore, most of the hafnium ions are retained in the aqueous effluent and discharged with the water flow, thus achieving efficient interfacial adsorption and separation.

[0015] Step 4: Desorption and enrichment of zirconium Once the adsorption column has broken through (the zirconium concentration in the effluent reaches 5–10% of the feed concentration), stop feeding. Rapidly rinse the column with deionized water to remove any residual non-adsorbed hafnium ions. Then, use 0.5–2.0 mol / L dilute sulfuric acid or 0.1–0.5 mol / L oxalic acid solution as the desorbent, rinsing and desorbing the adsorption column at a flow rate of 1–3 BV / h. Under acidic desorption conditions, H... + The zirconium ion competes with the zirconium ion for coordination sites, and the P=O group is protonated, disrupting the zirconium-ligand complex formed at the interface and allowing the zirconium ion to re-enter the aqueous phase. Collecting the desorption solution yields a zirconium-rich, high-purity solution in which the hafnium content has been reduced to below 0.01%.

[0016] Step 5: Precipitation and calcination to prepare high-purity zirconium dioxide To the zirconium-rich desorption solution obtained in step four, concentrated ammonia (25–28 wt%) is slowly added dropwise under stirring, controlling the final pH value to 8–9. Within this pH range, zirconium completely precipitates as a white colloidal precipitate of Zr(OH)₄, while soluble ammonium salts and some residual trace impurity ions remain in the supernatant. After plate and frame filtration or centrifugation, the precipitate is repeatedly washed with deionized water until the conductivity of the washing solution is below 10 μS / cm (to remove NH₄⁺). + and SO4 2- (etc.), and then the filter cake is placed in a muffle furnace or rotary kiln and calcined in air at 800–1000°C for 3–6 hours. During calcination, Zr(OH)4 undergoes thermal decomposition and dehydration to generate monoclinic ZrO2 powder: The resulting zirconium dioxide powder has a purity of ≥99.99% and a hafnium content of ≤0.01wt%.

[0017] Step Six: Purification with Chlorination and Zirconium Tetrachloride The high-purity zirconium dioxide powder obtained in step five is uniformly mixed with petroleum coke or carbon black (fixed carbon content ≥98%) at a mass ratio of 1:0.15 to 1:0.25, and an appropriate amount of binder is added to press it into blocks. The blocks are placed in a fluidized bed chlorination furnace, and high-purity chlorine gas is introduced at a temperature of 1000–1200℃ to carry out the chlorination reaction. ; The generated crude zirconium tetrachloride gas, carrying small amounts of impurity chlorides such as FeCl3, AlCl3, and TiCl4, as well as unreacted Cl2, is cooled to 150–180°C in a condenser, causing ZrCl4 to condense into a solid powder. Low-boiling-point impurities (such as SiCl4, with a boiling point of 57.6°C) remain in the gas phase for further separation. The crude ZrCl4 is then subjected to hydrogen reduction in a hydrogen atmosphere at 400–500°C, reducing FeCl3 and other impurities to high-boiling-point low-valence chlorides or elemental metals. This is followed by purification through distillation at 350–400°C to obtain high-purity zirconium tetrachloride.

[0018] Step 7: Magnesium reduction to prepare sponge zirconium In a sealed reduction furnace under inert gas (high-purity argon) protection, high-purity magnesium ingots (purity ≥99.95%) are added beforehand, and the magnesium is heated to 850±10℃ to melt. The purified zirconium tetrachloride obtained in step six is ​​sublimated and vaporized at 300–350℃, and then introduced at a controlled rate above the surface of the molten magnesium. The magnesium and zirconium tetrachloride vapor undergo the following reduction reaction at 850℃: The amount of magnesium used is 1.1 to 1.3 times the stoichiometric amount to provide an appropriate reduction margin and ensure complete reduction of ZrCl4. The reaction time is 4 to 8 hours, during which the reaction temperature is maintained constant and the molten salt is continuously stirred. The metallic zirconium generated in the reaction is deposited in a sponge-like porous morphology on the bottom and sidewalls of the molten salt, and the byproduct MgCl2 serves as the molten salt medium to fill the pores of the sponge zirconium.

[0019] Step 8: Vacuum distillation and post-treatment After the reduction reaction, the crucible containing the sponge zirconium and MgCl2-Mg mixture was transferred to a vacuum distillation furnace. Vacuum distillation was carried out under conditions of 0.01–0.1 Pa and 950–1050 °C for 8–16 hours. Under these high-temperature and high-vacuum conditions, the vapor pressure of the residual metallic magnesium (boiling point 1090 °C, atmospheric pressure) and magnesium chloride (boiling point 1412 °C, atmospheric pressure) increased significantly, preferentially vaporizing and being condensed and collected on the surface of the condenser, thus achieving separation from the sponge zirconium. After distillation, the furnace was cooled to room temperature, and the sponge zirconium blocks were removed in an inert atmosphere glove box. After crushing, sieving, batch mixing, and sampling analysis, the final nuclear-grade sponge zirconium product was obtained.

[0020] Step Nine: Regeneration and Recycling of the Adsorbent After desorption, the adsorption column is regenerated by rinsing with 0.1 mol / L dilute nitric acid at a flow rate of 2–3 BV / h for 3–5 BV to thoroughly remove any remaining trace metal ions and organic contaminants. It is then rinsed with deionized water until the effluent pH is neutral. The regenerated adsorption column can be directly used for the next batch of adsorption operations, with a cycle life of at least 80 cycles. After 80 cycles, the adsorption capacity retention rate is no less than 85% of the initial capacity.

[0021] The coordination chemistry basis for zirconium-hafnium separation in this invention lies in the fact that, in acidic aqueous solutions, Zr... 4+ and Hf 4+ Both exist as hydrated ions, and the difference in their interactions with ligands mainly stems from a slight difference in ionic radius (only about 0.01 Å). In single-functional group systems, this difference is difficult to amplify significantly; however, in multi-functional group synergistic coordination systems, the geometric matching of coordination configurations will significantly affect the thermodynamic stability of the complex. If a bifunctional ligand with a specific spatial configuration is designed, allowing it to form a stable six-membered chelate ring with zirconium ions through the synergistic effect of a strong coordination anchor (e.g., P=O) and an auxiliary coordination anchor (e.g., NH), while the stability of the chelate ring decreases due to the slightly smaller ionic radius and slightly different coordination bond length requirements of hafnium ions, highly selective separation based on interfacial coordination configuration recognition can be achieved.

[0022] Based on the above coordination chemistry principle, this invention proposes to use polyethyleneimine (PEI) to react with triethyl phosphate to generate a prepolymer containing aminophosphonate groups, which is then modified with an epoxy silane coupling agent and loaded onto a mesoporous silica support to prepare a novel NP bifunctional polymeric solid-phase adsorbent. The adsorbent has the following design advantages: (1) The PEI polymer chain provides abundant primary and secondary amine groups as reaction sites and auxiliary coordination sites, and the flexibility of the polymer chain gives the coordination groups the freedom to adjust the spatial orientation at the interface, which is conducive to forming the optimal coordination configuration; (2) In the aminophosphonate group introduced by triethyl phosphate through the Mannich-type reaction, P=O has a strong coordination ability and is the main coordination anchor point; (3) The epoxy silane coupling agent not only provides the anchoring function of covalent bonding with the inorganic support, but also the reaction of its epoxy group with the remaining amino group on PEI further forms a β-amino alcohol structure containing hydroxyl groups. These hydroxyl groups may participate in secondary coordination interactions, which strengthens the NP bifunctional synergistic effect; (4) The high specific surface area and ordered pore structure provided by the mesoporous silica support ensure the high adsorption capacity and fast mass transfer kinetics of the adsorbent.

[0023] Compared with the prior art, the present invention has the following outstanding advantages: 1. This invention creatively immobilizes a PEI-derived aminophosphonate bifunctional coordination structure at the solid-liquid interface via chemical bonding, achieving zirconium-hafnium separation based on interfacial bifunctional synergistic coordination recognition. The core mechanism lies in the fact that the aminophosphonate, acting as a bidentate ligand, uses P=O as the first coordination anchor and NH as the second coordination anchor, and interacts with Zr... 4+ The ions form a stable six-membered ring chelate complex; while Hf 4+ Because the ionic radius is slightly smaller than that of Zr 4+ (Difference of only 0.01 Å) Bond length mismatch occurs in the rigid coordination geometry of this six-membered ring, resulting in a coordination bond angle shift of approximately 2–3°, significantly reducing the stability of the chelate ring. Density functional theory (DFT) calculations (see Example 11) show that the binding energy of the Zr-adsorbent complex is approximately 18–25 kJ / mol higher than that of the Hf-adsorbent complex, and this energy difference can be translated into a significant amplification of the separation coefficient. Experiments confirm that the NP bifunctional solid-phase adsorbent of this invention has an adsorption capacity of 25–45 mg / g for zirconium and a corresponding adsorption capacity of 1.2–3.5 mg / g for hafnium. Under optimized conditions, the zirconium / hafnium separation coefficient β(Zr / Hf) can reach 8–20, which is much higher than the 3.3 of traditional physically supported phosphonate solid-phase extractants and also exceeds the 5–10 of the MIBK-HSCN system.

[0024] 2. This invention eliminates the use of thiocyanate and the flammable organic solvent MIBK, fundamentally preventing the risk of cyanide formation and the emission of organic waste gas. The adsorption operation is carried out in the aqueous phase, without the use of an organic phase. The waste liquid is only an acidic aqueous solution containing trace amounts of metal ions, which can meet emission standards after simple neutralization and precipitation treatment. The adsorbent is an inorganic-organic hybrid solid material, non-flammable and non-explosive, and can be operated under normal temperature and pressure conditions, significantly improving the inherent safety of the process. Simultaneously, the adsorbent can be regenerated more than 80 times, greatly reducing reagent consumption and solid waste generation.

[0025] 3. Under the optimized process conditions of this invention, only single-column or dual-column series adsorption-desorption operation is required (total time is about 4 to 6 hours) to reduce the hafnium content in the feed liquid from 2 to 3 wt% to below 0.01% in one step. The number of separation stages is significantly reduced compared to the 6 to 10 stages of traditional solvent extraction, which significantly reduces equipment investment and floor space, and reduces the number of operators required.

[0026] 4. The hafnium content in the nuclear-grade sponge zirconium obtained by this invention can be stably controlled below 0.008 wt% (as low as 0.003 wt%), which is far superior to the 0.01% upper limit requirement of the nuclear-grade standard. Simultaneously, since non-radioactive impurity ions (iron, titanium, aluminum, etc.) do not easily form stable complexes with NP bifunctional ligands during adsorption, they are initially separated during the adsorption stage. The final product has an oxygen content ≤0.06% and an iron content ≤0.05%, and the contents of various impurity elements all meet the requirements of the HZr-01 grade in the YS / T397-2015 nuclear-grade sponge zirconium standard.

[0027] 5. The mesoporous silica support used in this invention is widely available, cost-effective, and chemically stable; PEI and silane coupling agents are both bulk industrial chemicals; the adsorbent preparation process is mature, controllable, and easy to scale up. The dynamic column operation mode is highly compatible with existing ion exchange industrial devices, allowing for direct industrial production using mature equipment, significantly reducing the risk of technology transfer.

[0028] 6. Based on an annual production of 100 tons of nuclear-grade sponge zirconium, compared with the traditional MIBK-HSCN process, this invention can save about 40-50% of the cost of purchasing and processing organic solvents, reduce wastewater treatment costs by about 60%, and reduce equipment investment by about 30% due to the shortened process flow, resulting in a reduction of about 25-35% in overall production costs. Detailed Implementation

[0029] The technical solutions in the embodiments of the present invention will be clearly and completely described below. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments.

[0030] Example 1: Preparation of NP bifunctional adsorbent -- PEI to triethyl phosphate mass ratio 1:1 (a) Preparation of PEI functionalized prepolymer: 10.0 g of polyethyleneimine (PEI, molecular weight 1800 Da, Aladdin reagent) was weighed and dissolved in 100 mL of anhydrous ethanol, transferred to a 250 mL three-necked flask, and protected with high-purity nitrogen. The mixture was heated to 70 °C with magnetic stirring, and 10.0 g of triethyl phosphate (AR, Sinopharm Group) was slowly added dropwise using a constant-pressure dropping funnel at a rate of approximately 1 mL / min. After the addition was complete, the mixture was kept at 70 °C for 10 hours. After the reaction was complete, the solvent and unreacted triethyl phosphate were removed by vacuum distillation to obtain 13.2 g of a pale yellow, viscous PEI functionalized prepolymer.

[0031] (b) Silanization modification: The prepolymer obtained in step (a) was dissolved in 50 mL of anhydrous ethanol, and 6.0 g of 3-(glycidoxypropyl)trimethoxysilane (KH-560, purity ≥98%, Maclean's reagent) was added. The mixture was stirred at 50 °C for 6 hours. During the reaction, the epoxy groups of KH-560 underwent a ring-opening addition reaction with the residual amino groups on the PEI prepolymer, and the viscosity of the solution gradually increased, resulting in an ethanol solution of the silanized bifunctional macromolecule.

[0032] (c) Loading onto a carrier: Weigh 15.0 g of mesoporous silica carrier (specific surface area 350 m²) 2 / g, pore size 8nm, particle size 75~150μm (pre-dried and activated at 150℃ for 4 hours), added to 150mL of ethanol-water mixed solvent (ethanol:water = 3:1 v / v), and ultrasonically dispersed for 30 minutes. The silanized macromolecular solution obtained in step (b) was slowly added dropwise to the silica suspension under stirring, the pH was adjusted to 5.0±0.2 with 0.5mol / L hydrochloric acid, and the temperature was raised to 60℃ and stirred for 10 hours. After the reaction was completed, the mixture was filtered, and the filter cake was washed thoroughly with 150mL of ethanol and 300mL of deionized water in sequence, and dried under vacuum at 70℃ for 12 hours to obtain NP bifunctional polymeric solid-phase adsorbent (denoted as Adsorbent-A), totaling 22.8g.

[0033] Example 2: Adsorption Separation Performance Test - Single-Column Dynamic Adsorption Preparation of feed solution: Weigh 416.5 g of zirconium oxychloride (ZrOCl2·8H2O, industrial grade, hafnium content 2.5 wt%), dissolve it in 1.0 mol / L hydrochloric acid, and bring the volume to 1.0 L. The zirconium concentration in this feed solution is 1.0 mol / L (approximately 91.2 g / L), and the hafnium content is approximately 2.28 g / L. The contents of Zr, Hf, Fe, and Ti in the feed solution were determined using inductively coupled plasma optical emission spectrometry (ICP-OES).

[0034] Adsorption experiment: 5.0 g of Adsorbent-A prepared in Example 1 was packed into a 1.0 cm inner diameter glass column using a wet packing method. The column bed height was approximately 12 cm, and the bed volume (BV) was approximately 9.4 mL. The column was pre-equilibrated to 3 BV with 1.0 mol / L hydrochloric acid (pH approximately 0.5). The above feed solution was passed through the adsorption column from top to bottom at a flow rate of 3 BV / h (approximately 28 mL / h). The eluent was collected using an automatic fraction collector, with one sample collected for every 1 / 3 BV. The Zr content in each sample was determined by ICP-OES. 4+ and Hf 4+ Concentration. When Zr in the effluent... 4+ When the concentration reaches 10% of the feed concentration, it is considered the zirconium penetration point, and feeding is stopped.

[0035] Desorption procedure: After feed is stopped, rinse rapidly with 1 / 3 BV deionized water. Then desorb with 1.0 mol / L sulfuric acid solution at a flow rate of 2 BV / h, collect the eluent, and determine the Zr content. 4+ and Hf 4+ concentration.

[0036] result: Dynamic adsorption capacity of zirconium: 38.2 mg / g (calculated based on breakthrough point); The dynamic adsorption capacity of hafnium is 2.4 mg / g. The separation coefficient β(Zr / Hf) = (38.2 / 91.2) / (2.4 / 2.28) = 15.9; Hafnium content in zirconium-rich desorption solution: 0.0068 wt% (relative to zirconium); Zirconium single-adsorption-desorption recovery rate: 92.5%.

[0037] Example 3: Dual-column tandem adsorption Adsorbent-A was prepared according to the method in Example 1, and two chromatographic columns (column A and column B) of the same specifications were packed according to the method in Example 2. The columns were then operated in a dual-column tandem mode.

[0038] Operating procedure: The feed solution (Zr 1.0 mol / L, Hf content 2.5 wt%) is passed through column A at a flow rate of 3 BV / h. The effluent from column A is directly connected in series to column B for secondary adsorption. When column A reaches its breakthrough point, column A is desorbed and regenerated offline (desorption conditions are the same as in Example 2). Simultaneously, column B is used as the first column, and the newly regenerated column A continues to operate in series (simulating a moving bed switching mode). The zirconium-rich desorption solution is collected separately.

[0039] Results (average of three runs): Column A adsorption capacity: 37.8 mg / g; Column B adsorption capacity: 15.3 mg / g (as a purification column, the feed zirconium and hafnium concentrations have been significantly reduced); Hafnium content in the aggregated desorption solution: 0.0032 wt%; Total zirconium recovery: 98.5%; The system ran continuously for 10 cycles, with adsorption capacity fluctuations of <3% and desorption liquid hafnium content consistently below 0.004%.

[0040] Example 4: Effect of different PEI / triethyl phosphate feed ratios on adsorption performance Following the method of Example 1, with a fixed PEI content of 10.0 g, three adsorbents with different P / N ratios were prepared by varying the amount of triethyl phosphate: Table 1. Effect of different PEI / triethyl phosphate feed ratios on adsorption performance

[0041] As shown in Table 1, when PEI:P(OEt)3=1:0.5 (B1), the amount of triethyl phosphate is insufficient, and some amino groups on the PEI chain fail to be converted into aminophosphonate structures, resulting in a low density of bifunctional coordination sites, unsatisfactory adsorption capacity and selectivity, and the hafnium content (0.015%) in the desorption solution exceeds the standard.

[0042] When PEI:P(OEt)3 = 1:1 (A), the number of aminophosphonate groups is sufficient and appropriate, forming an optimal spatial ratio with the unreacted amino groups on the PEI skeleton. P=O and NH coordinate with Zr in a cooperative manner. 4+ A stable six-membered ring chelate configuration was formed. The separation coefficient reached 15.9, which was the best among all groups.

[0043] When PEI:P(OEt)3=1:2 (B2), the excess triethyl phosphate almost completely converts the reactive amino groups on PEI into aminophosphonate groups. At this time, the auxiliary coordination effect of NH is weakened, and the bifunctional coordination degenerates into monofunctional phosphonate coordination, which reduces the selectivity to a certain extent (the separation coefficient drops to 11.1).

[0044] Example 5: Effect of different desorbent concentrations on desorption efficiency Using the Adsorbent-A column saturated with adsorption in Example 2 as the object, desorption was performed using sulfuric acid solutions of different concentrations at a desorption flow rate of 2 BV / h. The zirconium recovery rate was determined by collecting the desorbate.

[0045] Table 2. Effect of different desorbent concentrations on desorption efficiency

[0046] Table 2 shows that the lower the desorbent concentration (0.5 mol / L), the higher the H+ concentration. +The concentration was insufficient to completely destroy the Zr-ligand complex at the interface, resulting in a desorption recovery rate of only 85.2%, which led to some zirconium residue remaining in the column. Although the Hf content in the eluent was the lowest at this point (0.0052%), the recovery rate did not meet the requirements for industrial production.

[0047] When the desorbent concentration is 1.0–1.5 mol / L, the desorption recovery rate is >96%, and the Hf content in the desorbate is still below 0.01%, which is the optimal operating window.

[0048] When the desorbent concentration is increased to 2.0 mol / L, excessively high H+... + The concentration not only desorbed Zr, but also excessively disrupted the weak interactions between Hf sites, leading to an enhanced co-desorption effect of Hf. The Hf content in the desorbent increased to 0.012%, exceeding the nuclear grade standard.

[0049] Conclusion: The optimal concentration of the desorbent is 1.0–1.5 mol / L H2SO4.

[0050] Example 6: Effect of different adsorption pH on separation efficiency Using Adsorbent-A as the stationary phase, the pH of the feed solution was adjusted with sodium hydroxide solution (while keeping the zirconium concentration and Hf / Zr ratio constant), and dynamic adsorption experiments were conducted under different pH conditions, with the same operating conditions as in Example 2.

[0051] Table 3. Effect of different adsorption pH on separation efficiency

[0052] As shown in Table 3, the effect of pH on the separation coefficient exhibits a single-peak curve, with the optimal pH range being 1.0–1.5.

[0053] Under extremely low pH (0.2) conditions, the P=O and NH coordination groups are severely protonated (P=O...H + and N-H2 + ( ), losing coordination ability, adsorption capacity and selectivity drop sharply.

[0054] At pH 1.0–1.5, the protonation degree of the ligands is moderate, and the synergistic effect of the free P=O anchor and NH-assisted coordination is optimal. Meanwhile, Zr… 4+ and Hf 4+ Differences in the degree of hydrolysis are beginning to emerge -- Zr 4+ The hydrolysis tendency is slightly lower than that of Hf 4+ This further contributes to selective separation.

[0055] At pH 2.0, the degree of hydrolysis of metal ions is aggravated, and some zirconium and hafnium may form polynuclear hydroxyl-bridged species, which complicates the coordination chemistry. The increase in adsorption capacity is mainly attributed to non-selective physical adsorption and surface precipitation, resulting in a significant increase in Hf co-adsorption and a sharp decrease in the separation coefficient (down to 8.7).

[0056] Example 7: Effect of different silane coupling agent dosages on adsorbent stability Three adsorbents were prepared by fixing the amounts of PEI (10.0 g), P(OEt)3 (10.0 g), and SiO2 support (15.0 g), and varying the amount of KH-560. The adsorption capacity retention rate after 80 cycles was used as the stability evaluation index. Table 4. Effect of different silane coupling agent dosages on adsorbent stability

[0057] As shown in Table 4, the amount of KH-560 was too small (PEI prepolymer:KH-560 = 1:0.3, C1), and the silane coupling agent could not provide sufficient anchoring sites for the functionalized PEI macromolecules. Some functional macromolecules were attached to the support surface only through physical adsorption or weak hydrogen bonding, and gradually detached after multiple desorption and regeneration (acid rinsing), resulting in a decrease in adsorption capacity of more than 30% after 80 cycles.

[0058] When KH-560 is used at a ratio of 1:0.6 (A), the degree of silanization is appropriate, ensuring sufficient covalent anchoring without diluting the functional site density due to excessive coupling agent. After 80 cycles, the capacity retention rate remains above 90%, meeting the lifespan requirements for industrial applications.

[0059] The dosage of KH-560 was further increased to 1:1.5 (C2). Although the capacity retention rate was further improved to 95.9%, the initial adsorption capacity per unit mass of adsorbent decreased (from 38.2 to 32.1 mg / g) due to the increased proportion of silane coupling agent (inactive component) in the total mass of adsorbent. In other words, there is a trade-off between stability and capacity.

[0060] The physical impregnation method in comparison results in a large loss of functional molecules during acidic desorption because there are no chemical bonds between the functional molecules and the carrier. After 80 cycles, the capacity retention rate is only 31%, making it unsuitable for industrial recycling.

[0061] Example 8: Effect of different carrier pore sizes on adsorption kinetics Prepolymers and silanized macromolecules were prepared using the same chemical formulation as in Example 1, and then loaded onto three different mesoporous silica supports with different pore sizes (specific surface areas controlled at 300–350 m²). 2 / g range (particle size 75–150 μm). Dynamic adsorption was performed using the above three adsorbents, breakthrough curves were tested, and half-breakthrough times were calculated.

[0062] Table 5. Effect of different support pore sizes on adsorption kinetics

[0063] As shown in Table 5, although the support with a pore size of 3 nm (D1) has a considerable specific surface area, a large portion consists of micropores (<2 nm) and narrow mesopores. These pores are difficult for PEI-derived macromolecular functional groups to fully enter and modify; the functional groups are mainly distributed on the outer surface and within the macropores, resulting in a low density of effective functional sites and thus the lowest adsorption capacity (22.5 mg / g). Simultaneously, the narrow pores restrict the adsorption of Zr in the solution. 4+ The diffusion mass transfer of hydrated ions has the longest half-penetration time (8.5 min / BV).

[0064] The 8 nm (A) pore size support balances sufficient pore size to allow for the full penetration and modification of macromolecular functional groups, and appropriate pore size to enable rapid diffusion and mass transfer of zirconium ions, resulting in the highest adsorption capacity and reasonable adsorption kinetics.

[0065] The carrier with a pore size of 20 nm (D2) has a relatively low specific surface area and a relatively large pore volume. Although it diffuses the fastest (shortest half-penetration time), the density of the loaded organic functional groups is slightly lower, and the capacity is slightly lower than that of A (35.8 vs 38.2 mg / g).

[0066] Example 9: Comparison of adsorbent processing capacity for raw materials with different initial hafnium contents Adsorbent-A was used in single-column operation mode to compare the adsorption treatment of three feed solutions with different hafnium contents: Table 6. Comparison of adsorbent processing capacity for feedstocks with different initial hafnium contents

[0067] As shown in Table 6, for raw materials with Hf contents of 1.0% and 2.5%, the Hf content in the desorption solution after single-column adsorption can be reduced to below 0.008%, directly meeting the nuclear grade standard requirements. For raw materials with an Hf content as high as 5.0%, the Hf content after single-column treatment is 0.016%, which is close to the nuclear grade standard but still slightly exceeds the limit, requiring further purification using a dual-column series (as in Example 3). Regardless of the hafnium content of the raw material, the separation coefficient β of this adsorbent remains stable at 15–18, exhibiting good robustness.

[0068] Example 10: Optimization of reduction process parameters The zirconium-rich desorption solution prepared by the method in Example 2 was subjected to precipitation and calcination to obtain high-purity ZrO2 (Hf content 0.0068wt%), which was used as the raw material in this example.

[0069] Chlorination and purification: 200g ZrO2 and 40g carbon black were mixed and briquetted, and reacted in a fluidized bed chlorination furnace at 1150℃ with Cl2 for 6 hours to produce 305g crude ZrCl4. The crude ZrCl4 was reduced in a H2 atmosphere at 400℃ for 2 hours, and then distilled at 380℃ to obtain 285g refined ZrCl4. ICP-MS analysis showed that the iron content decreased from 158ppm to 12ppm.

[0070] Magnesium thermal reduction parameter optimization experiment: The refined ZrCl4 was divided into 4 equal parts, and the reduction temperature and magnesium excess coefficient were changed to study the effects on the quality and yield of sponge zirconium.

[0071] Table 7. Effects of reduction process parameters on the quality and yield of sponge zirconium

[0072] Table 7 provides a preliminary selection criterion for the following parameters: Reduction temperature: At 800℃ (R-1), the reaction rate was slow, the Mg-ZrCl4 contact reaction was incomplete, the yield was only 85.2%, and the product texture was uneven. At 850℃ (R-2), the reaction rate and mass transfer efficiency were well matched, the yield was 96.8%, and the product hardness was moderate. When the temperature was increased to 900℃ (R-3), the yield increased slightly (97.2%), but the excessively high temperature accelerated the dissolution of trace amounts of furnace material into the product, and energy consumption increased. Therefore, 850±10℃ was selected as the optimal reduction temperature.

[0073] Magnesium Excess Factor: When the magnesium excess factor is 1.05 (R-4), the stoichiometric ratio is close, but some ZrCl4 is not fully reduced, and the yield decreases to 91.5%. When the magnesium excess factor is 1.30 (R-5), reduction is complete, but although the excess magnesium can be removed in the subsequent vacuum distillation stage, the initially high Mg residue requires a longer distillation time (or a higher vacuum), and trace amounts of Mg (1850 ppm) may still exist in the sponge zirconium channels after distillation. When the magnesium excess factor is 1.15 (R-2), the optimal balance between yield and product purity is achieved, and the Mg residue of 860 ppm in the product can be further removed in subsequent distillations.

[0074] Optimized parameters for distillation and final product: The reduction product of R-2 was vacuum distilled at 1000℃ and 0.05Pa for 12 h to obtain sponge zirconium. Final product test results: Hf 0.0068wt%, O 0.052wt%, Fe 0.038wt%, Mg 0.0042wt%, C 0.008wt%, Cr 0.0025wt%, Brinell hardness HB145, loose bulk density 1.3 g / cm³. 3 Purity ≥ 99.95%. All indicators meet the requirements of YS / T 397-2015 HZr-01 grade.

[0075] Comparative Example 1 (Traditional MIBK-HSCN Solvent Extraction Method) The feed solution (Zr 1.0 mol / L, Hf 2.5 wt%) was treated using a typical industrial MIBK-HSCN extraction process, similar to that in Example 2. Extraction conditions: MIBK containing 3.0 mol / L HSCN, O / A ratio of 3:1, 6-stage extraction, 4-stage washing. Four parallel batches were tested.

[0076] Results: Hafnium content in zirconium-rich raffinate: 0.008–0.025 wt% (significant batch-to-batch fluctuations, RSD = 38%); Zr direct recovery rate: 89.5%; single-batch run time (including phase separation waiting time): approximately 12 hours (4 hours for single-column adsorption in this invention); approximately 1.2 tons / batch of thiocyanate-containing wastewater is generated during operation, requiring specialized harmless treatment; approximately 15% of the degraded organic phase needs to be treated after each batch.

[0077] The core difference from this invention is that traditional liquid-liquid extraction is based on distribution equilibrium, with a separation coefficient of only about 7. It requires 6 stages of extraction and 4 stages of washing to achieve the same effect as the single column of this invention. Backmixing and emulsification between stages lead to poor consistency between batches.

[0078] Comparative Example 2 (Physical Impregnation P507 / Carbon Aerogel Solid Phase Extraction) Following existing methods, P507 (2-ethylhexylphosphonic acid monoester) was loaded onto carbon aerogel by impregnation, with a P507 loading of 30 wt%. Experiments were conducted under the same column conditions (column diameter 1.0 cm, bed height 12 cm, flow rate 3 BV / h, and same feed solution).

[0079] Results: Zr adsorption capacity: 16.8 mg / g (only 44% of that in Example 2); Hf adsorption capacity: 5.1 mg / g; β(Zr / Hf) = 3.3 (far lower than 15.9 in this invention); Hf content in the desorption solution: 0.045 wt%, which cannot directly meet the nuclear grade standard; after 40 cycles, the Zr adsorption capacity decreased to 6.2 mg / g (a decrease of 63%), and a large number of functional groups were lost.

[0080] The mechanism difference compared with the present invention: the physically impregnated P507 is a monofunctional phosphonate group, lacking the auxiliary coordination of NH, and cannot form the configuration recognition effect of the double-anchor six-membered chelate ring, and lacks covalent bond anchoring, resulting in poor cycle stability.

[0081] Comparative Example 3 (single amino-modified silica gel, without phosphonate groups) N-monofunctional adsorbents were prepared by grafting PEI onto a silica gel support using only KH-560 (completely omitting the triethyl phosphate reaction step). Tests were conducted under the same conditions.

[0082] Results: Zr adsorption capacity: 22.3 mg / g; Hf adsorption capacity: 4.8 mg / g; β(Zr / Hf) = 4.8.

[0083] Compared with Example 2 (β=15.9): Lacking the strong P=O coordination anchor, the PEI relies solely on the weak coordination between its amino group and the metal ion (mainly through the N→M coordination bond), resulting in weak coordination strength and the inability to form a rigid six-membered chelate ring. The lack of strong coordination anchoring from the phosphate ester group leads to significantly lower adsorption capacity and selectivity compared to the NP bifunctional adsorbent. This directly demonstrates the indispensability of the P=O / NH bifunctional synergy for achieving highly selective separation.

[0084] Comparative Example 4 (Single phosphonate-modified silica gel, without amino groups) A P-monofunctional adsorbent was prepared by directly reacting triethyl phosphate with KH-560 (without PEI component) and then loading it onto silica gel.

[0085] Results: Zr adsorption capacity: 28.5 mg / g; Hf adsorption capacity: 4.0 mg / g; β(Zr / Hf) = 7.3.

[0086] Compared with Example 2 (β=15.9): In the absence of NH auxiliary coordinating groups, relying solely on P=O monodentate or bidentate coordination, the coordination configuration is more flexible, making it impossible to precisely distinguish Zr based on the configuration matching principle. 4+ and Hf 4+ The 0.01 Å radius difference limits selectivity. Compared to N-monofunctional (β=4.8) and P-monofunctional (β=7.3), the separation coefficient of the NP bifunctional adsorbent (β=15.9) is greater than about 62% of the sum of the two, exhibiting a significant synergistic effect.

[0087] Comparative Example 5 (Small molecule aminophosphonate impregnated silica gel, non-polymer covalent bonding) The PEI functionalized prepolymer obtained in step (a) of Example 1 was directly impregnated onto mesoporous silica without silanization (no chemical bonding) and tested under the same conditions.

[0088] Results: Initial Zr adsorption capacity: 35.5 mg / g (close to 38.2 mg / g in Example 2); after 40 cycles, the Zr adsorption capacity decreased to 11.8 mg / g (a decrease of 66.8%).

[0089] Compared with Example 2 (91.9% capacity retention after 80 cycles), the polymerized aminophosphonate functional groups, lacking covalent anchoring, were gradually eluted with each cycle under acidic desorption conditions, resulting in continuous capacity decay. This demonstrates that the anchoring effect of the chemical bonds in epoxy silane coupling agents is crucial for the long-life operation of adsorbents in the industry.

[0090] Comparative Example 6 (PEI molecular weight is too low) The PEI with a molecular weight of 600 Da was used instead of the 1800 Da PEI in Example 1, while other operating steps remained unchanged.

[0091] Results: Zr adsorption capacity: 32.5 mg / g; Hf adsorption capacity: 3.0 mg / g; β(Zr / Hf) = 11.0 (a decrease of approximately 31% compared to 15.9 in Example 2).

[0092] Analysis: The low molecular weight of PEI results in insufficient chain length, leading to a high density of aminophosphonate groups after reaction with triethyl phosphate, but a restricted spatial distribution, making it difficult to adjust to the optimal bifunctional coordination configuration. Furthermore, the short-chain polymer loading results in insufficient spatial freedom of functional groups on the silica surface, hindering the ability to fully adjust the configuration to accommodate Zr. 4+ - The optimal coordination geometry of the chelate ring leads to a decrease in selectivity.

[0093] Comparative Example 7 (PEI molecular weight is too high) Replace 1800Da with PEI with a molecular weight of 10000Da, while keeping other operating steps unchanged.

[0094] Results: Zr adsorption capacity: 20.1 mg / g; Hf adsorption capacity: 2.8 mg / g; β(Zr / Hf) = 12.5.

[0095] Analysis: When the molecular weight of PEI is too high, the long-chain polymer is prone to self-entanglement during loading, with a large number of amino sites being embedded inside the chain clusters and unable to participate in the reaction with triethyl phosphate, resulting in a decrease in the density of effective functional sites (manifested as a Zr adsorption capacity of only 20.1 mg / g). At the same time, the long-chain entanglement makes the local rigidity and spatial configuration of some effective sites unable to match the optimal requirements of six-membered ring chelation, and the separation coefficient also decreases.

[0096] When the molecular weight of PEI is preferably in the range of 1800–5000 Da, the optimal balance between adsorption capacity and selectivity can be obtained.

[0097] Comparative Example 8 (Product failure due to uncontrolled desorption acidity) In Example 2, desorption was performed directly on the column that was saturated with adsorption using 4.0 mol / L sulfuric acid (outside the preferred range of this invention).

[0098] Results: Desorption recovery rate: 99.3% (high); however, the Hf content in the desorption solution increased to 0.028 wt%, and the Hf content of the sponge zirconium product obtained after nitrogen calcination was 0.030 wt%, which seriously exceeded the standard and could not be used in the nuclear industry.

[0099] Note: Although an overly concentrated desorption solution increases the recovery rate, it exacerbates the non-selective co-desorption of hafnium, resulting in substandard products. This demonstrates the necessity of specifying the desorbent concentration of 0.5–2.0 mol / L in the claims.

[0100] Example 12: Scale-up preparation and pilot-scale verification of adsorbent To further verify the industrial feasibility of the present invention, a pilot-scale amplification test was conducted.

[0101] Scale-up preparation of adsorbent (10x volume): Dissolve 100g of PEI (Mw approximately 1800) in 1.0L of anhydrous ethanol, purge with nitrogen, and add 100g of triethyl phosphate dropwise at 70℃, reacting for 10h. Add 60g of KH-560, reacting at 50℃ for 6h. Add the resulting silanized macromolecular solution dropwise to a mixture containing 150g of mesoporous silica (specific surface area 350m²). 2 The NP bifunctional adsorbent was prepared by reacting a suspension of ethanol-water (g) at pH 5.0 at 60°C for 10 hours. After filtration, washing, and drying, approximately 228 g of the adsorbent was obtained, which exhibited physicochemical properties consistent with those of the small-scale production in Example 1.

[0102] Pilot-scale adsorption separation: A Φ50×500mm stainless steel column was used, packed with 200g of adsorbent (bed height 40cm, bed volume approximately 785mL). The feed solution was also scaled up to 10L (Zr 1.0mol / L, Hf 2.5wt%), adsorption was performed at a flow rate of 3BV / h, and desorption was performed with 1.0mol / L H2SO4 under the same conditions.

[0103] Pilot-scale results: Zr dynamic adsorption capacity: 37.5 mg / g (only 1.8% lower than the 38.2 mg / g in the small-scale test); Hf content in the desorption solution: 0.0072 wt%; total Zr recovery rate: 96.2%; sponge zirconium product analysis (precipitation-calcination-chlorination-magnesium reduction-distillation as in Example 10): Hf 0.0072%, O 0.055%, Fe 0.042%, other impurities all meet nuclear grade standards; good reproducibility in the pilot-scale test, with Hf contents of 0.0072%, 0.0069%, and 0.0075% in the three batches of product, respectively.

[0104] This embodiment demonstrates that the method of the present invention can maintain performance comparable to that of the pilot-scale test at the pilot-scale, and has good prospects for industrial scale-up.

[0105] Based on the above embodiments and comparative data, the following summary and analysis are provided: 1. Confirmation of the co-coordination mechanism: Comparative Example 3 (pure N-functional) had a β value of 4.8, and Comparative Example 4 (pure P-functional) had a β value of 7.3. If the two were simply linearly added together, the expected β value for the bifunctional group would be approximately the sum of the two (12.1) or lower. However, Example 2 had a β value of 15.9, significantly higher than the expected sum. This indicates that the N and P bifunctional groups do not simply coexist, but rather exert a synergistic effect by forming a six-membered ring chelate configuration with P=O anchoring and NH auxiliary coordination—one group provides a strong coordination anchor, while the other group restricts the flexibility of the coordination configuration through auxiliary coordination, thereby amplifying the Zr function. 4+ and Hf 4+ Thermodynamic coordination differences.

[0106] 2. Key impacts of process parameters: Example 4 demonstrates that PEI:P(OEt)3=1:1 is the optimal ratio (β=15.9).

[0107] Example 6 demonstrates that the optimal pH window for adsorption is 1.0–1.5. If the pH is too low (<0.5), the ligands are overprotonated and lose their coordinating function; if the pH is too high (>2.0), the hydrolysis of metal ions interferes with selectivity.

[0108] Example 5 demonstrates that the optimal desorption range for the desorbent H2SO4 concentration is 1.0–1.5 mol / L, which can balance high recovery rate with low Hf co-desorption.

[0109] Example 10 demonstrates that the reduction temperature and magnesium excess coefficient must be strictly controlled within the following ranges: reduction temperature 850±10℃, magnesium excess coefficient 1.15±0.05; exceeding these ranges will result in a decrease in product yield or an increase in Mg residue.

[0110] 3. Confirmation of industrial competitive advantage: Compared to Comparative Example 1 (MIBK-HSCN), the purity level achieved by single-column adsorption in this invention, which involves 6 stages of extraction and 4 stages of washing, is comparable to (or even surpasses) that of Comparative Example 2 (physical impregnation). Furthermore, the process is shorter, produces no organic waste, and exhibits high batch-to-batch consistency. Compared to Comparative Example 2 (physical impregnation), the covalently bonded solid-phase adsorbent of this invention has a cycle life exceeding 80 cycles, far surpassing the 30-40 cycles of the impregnation method.

[0111] Table 8. Summary of Key Performance of Examples and Comparative Examples

[0112] As shown in Table 8, the above examples and comparative examples fully demonstrate the nuclear-grade sponge zirconium and its preparation method proposed in this invention. Through the interfacial synergistic coordination recognition mechanism of the PEI-derived NP bifunctional polymeric solid-phase adsorbent, deep separation of zirconium and hafnium can be achieved efficiently and environmentally under mild conditions. The resulting nuclear-grade sponge zirconium product fully meets the stringent requirements of nuclear reactor cladding materials. Any modifications, equivalent substitutions, or improvements made within the spirit and principles of this invention should be included within the scope of protection of this invention.

Claims

1. A method for preparing nuclear-grade sponge zirconium, characterized in that, Includes the following steps: (1) Dissolve zirconium oxychloride in dilute hydrochloric acid to prepare an acidic raw material solution with a zirconium concentration of 0.5-2.0 mol / L; (2) The raw material liquid is passed through an adsorption column filled with NP bifunctional solid phase adsorbent and dynamically adsorbed under the conditions of pH 0.5 to 1.5 and flow rate 1 to 5 BV / h. Zirconium ions are selectively adsorbed at the adsorbent interface, and hafnium ions are discharged with the effluent. (3) Use 0.5-2.0 mol / L dilute sulfuric acid or 0.1-0.5 mol / L oxalic acid solution as desorbent to desorb the adsorption column that is saturated with adsorption, and obtain zirconium-rich high-purity desorption solution; (4) Add ammonia to the desorption solution to adjust the pH to 8-9, generate zirconium hydroxide precipitate, filter, wash and calcine at 800-1000℃ to obtain high-purity zirconium dioxide; (5) Zirconium dioxide is mixed with carbon powder and reacted with chlorine gas in a chlorination furnace at 1000-1200℃ to produce crude zirconium tetrachloride. After purification by hydrogen reduction and distillation, refined zirconium tetrachloride is obtained. (6) The pure zirconium tetrachloride vapor is passed into molten magnesium at 850±10℃ to carry out a reduction reaction, producing spongy metallic zirconium and magnesium chloride; (7) Spongy metallic zirconium and magnesium chloride are vacuum distilled at 950-1050℃ and pressure ≤0.1Pa to remove residual magnesium and magnesium chloride, and then cooled and crushed to obtain nuclear-grade spongy zirconium.

2. The method for preparing nuclear-grade sponge zirconium according to claim 1, characterized in that, The preparation of the NP bifunctional solid-phase adsorbent in step (2) includes the following steps: (a) Polyethyleneimine is dissolved in anhydrous ethanol, and triethyl phosphate is slowly added dropwise under an inert atmosphere. The mixture is reacted at 60-80°C for 8-12 hours to generate a PEI-functionalized prepolymer containing aminophosphonate groups, wherein the mass ratio of PEI to triethyl phosphate is 1:0.5-1:

2. (b) The prepolymer obtained in step (a) is mixed with an epoxy silane coupling agent at a mass ratio of 1:0.3 to 1:1.5 and reacted at room temperature to 60°C for 4 to 8 hours to obtain a silanized bifunctional macromolecule; (c) The product obtained in step (b) is added to a water-ethanol mixed solvent containing a mesoporous silica support, the pH is adjusted to 4-6, and the mixture is stirred at 50-70°C for 6-12 hours. After filtration, washing and drying, the NP bifunctional solid-phase adsorbent is obtained, wherein the mass ratio of silica support to silanized macromolecules is 1:0.2-1:

1.

3. The method for preparing nuclear-grade sponge zirconium according to claim 2, characterized in that, The epoxy silane coupling agent is selected from one or more of 3-(glycidoxypropyl)trimethoxysilane, 3-(glycidoxypropyl)triethoxysilane, or 3-(glycidoxypropyl)methyldimethoxysilane.

4. The method for preparing nuclear-grade sponge zirconium according to claim 2, characterized in that, The specific surface area of ​​the mesoporous silica carrier is 200–600 m². 2 / g, pore size 5-20nm, particle size 50-200μm.

5. The method for preparing nuclear-grade sponge zirconium according to claim 1, characterized in that, The adsorption separation operation in step (2) adopts a dual-column series mode. The effluent from the first column directly enters the second column for secondary adsorption, and the zirconium-rich desorption liquid from the first column is collected as the product.

6. The method for preparing nuclear-grade sponge zirconium according to claim 1, characterized in that, After desorption is completed in step (3), the adsorption column is regenerated by rinsing with 0.1 mol / L dilute nitric acid, and then washed with deionized water until neutral before being recycled. The cycle life is no less than 80 times.

7. The method for preparing nuclear-grade sponge zirconium according to claim 1, characterized in that, In step (6), the amount of magnesium used is 1.1 to 1.3 times the theoretical amount, and the reduction reaction time is 4 to 8 hours.

8. The method for preparing nuclear-grade sponge zirconium according to claim 1, characterized in that, In step (7), the vacuum distillation time is 8 to 16 hours and the vacuum degree is 0.01 to 0.1 Pa.

9. A nuclear-grade sponge zirconium obtained by the preparation method according to any one of claims 1 to 8, characterized in that, The sponge zirconium has a hafnium content of ≤0.008% by weight, an oxygen content of ≤0.06%, an iron content of ≤0.05%, and a purity of ≥99.9%.

10. The nuclear-grade sponge zirconium according to claim 9, characterized in that, The sponge zirconium has a Brinell hardness of HB120–180 and a loose packing density of 1.0–1.5 g / cm³. 3 .

Images