A green smelting process of rare earth metals

Abstract

The present application relates to the technical field of rare earth metal smelting, and discloses a green smelting process for rare earth metals, which comprises the following steps: melting base electrolyte in an electrolytic cell by power supply, increasing the distance between the anode and the cathode, controlling the temperature of the upper layer of the molten salt to be lower than the temperature of the bottom liquid cathode working area, executing trapezoidal wave step electrolysis cycle, inputting coarse crystal slag powder in the main deposition period of the high current interval, then smoothly reducing the current to the low current interval to enter the relaxation period, synchronously spraying praseodymium neodymium oxide powder and fluxing and dedusting compound during the period, and then raising the current again after the feeding is completed. By constructing a vertical temperature gradient, the present application utilizes the high viscosity of the middle and upper layers to physically intercept waste impurities, and combines step current and time sequence feeding to generate a complex on the anode, thereby reducing the interface viscosity to empty the insulating gas film and avoiding anode effect, so that internal closed-loop recycling of high-impurity fluorine-containing solid waste and high-purity precipitation of rare earth metals are realized.

Description

Technical Field

[0001] This invention relates to the field of rare earth metal smelting technology, specifically a green smelting process for rare earth metals. Background Technology

[0002] During the electrolytic production of praseodymium and neodymium metals, the physical dispersion of raw materials and the deposition and enrichment of impurities generate dust-collecting and recovery materials and furnace slag. These materials contain rare earth elements, praseodymium and neodymium fluoride, and lithium fluoride, but also contain heavy metals such as iron and silicon, as well as non-metallic impurities. Currently, the industry typically uses hydrometallurgical processes to separate these materials and purify rare earth oxides. This processing method is lengthy and costly, and the chemical treatment process disrupts the original fluoride system, resulting in the loss of fluorine and lithium resources and a burden on wastewater treatment.

[0003] To reduce processing costs and achieve material recycling within the electrolysis system, some conventional technologies attempt to directly return the aforementioned high-impurity waste residue to the electrolytic cell. However, in conventional rare earth molten salt electrolysis processes, the electrolysis temperature is high and the system's kinematic viscosity is low. When the waste residue melts in the cell, the released impurity ions have a large diffusion coefficient and can easily penetrate the cathode diffusion layer quickly. Since the deposition potential of impurity ions such as iron and silicon is usually more positive than that of praseodymium and neodymium, they will preferentially undergo co-deposition reduction reactions at the cathode under the action of a DC electric field, resulting in the final extracted rare earth metal product having excessive impurity content and failing to meet the purity standards required for practical applications. Therefore, without adding complex external chemical extraction equipment, how to block the mass transfer of impurities to the cathode metal phase within the existing electrolytic cell is a technological obstacle limiting the direct recycling of high-impurity fluorine-containing solid waste. Summary of the Invention

[0004] The technical problem solved by this invention is that the fluorine-containing solid waste generated during the rare earth metal electrolysis production process has a high impurity content. Directly returning it to the electrolytic cell for purification will cause heavy metals and non-metallic impurities to co-deposit at the cathode, reducing the metal purity of the product. On the other hand, the traditional hydrometallurgical recycling process has technical drawbacks such as high processing costs and waste of fluorine and lithium resources.

[0005] To address the above problems, the present invention provides the following technical solution:

[0006] A green smelting process for rare earth metals includes:

[0007] A basic electrolyte is added to an electrolytic cell equipped with a variable frequency DC rectifier (we use a high frequency synchronous rectifier) ​​and an electrode gap adjustment device. The electrolyte is energized and melted to form molten salt. The vertical insertion depth of the graphite anode in the electrolytic cell is adjusted upward by a mechanical lifting mechanism. The electrode gap is set, the cell voltage is adjusted, and the temperature of the upper layer of the molten salt is controlled to be stable. The temperature of the working area of ​​the bottom liquid cathode is also controlled to be stable, and the temperature of the upper layer of the molten salt is controlled to be lower than the temperature of the working area of ​​the bottom liquid cathode.

[0008] Adjust the variable frequency DC rectifier to increase the cathode current density and maintain it at a first preset current density. During the period of maintaining the first preset current density, coarse crystal clear slag powder is fed into the middle and upper regions of the electrolytic cell through the automatic feeding port of the electrolytic cell.

[0009] The variable frequency DC rectifier is controlled to smoothly reduce the cathode current density to a second preset current density with a current drop slope, and then enters the relaxation period.

[0010] Maintain the cathode current density at the second preset current density. During the relaxation period, start the pneumatic jet feeding system to mix praseodymium oxide neodymium powder with fluxing dust removal compound and then spray it onto the molten salt surface of the area where the graphite anode is located.

[0011] After feeding is completed, the variable frequency DC rectifier raises the cathode current density back to the first preset current density with an increasing slope, thus completing a single electrolysis cycle.

[0012] By adopting the above technical solution, and through the control method that couples spatial pole spacing adjustment with time-domain trapezoidal wave step electrolysis, the following beneficial effects are achieved:

[0013] In terms of spatial mass transfer control, by increasing the electrode spacing of the graphite anode, the macroscopic thermal convection in the longitudinal direction of the electrolytic cell is cut off, thus constructing a vertical temperature gradient. The upper and middle layers of the molten salt are in a relatively low-temperature state, and the kinematic viscosity of the system increases. According to Fick's first law, the high kinematic viscosity physical environment limits the diffusion coefficient of impurities such as iron and silicon ions released from the melting of coarse crystalline slag powder, causing the impurity mass transfer process to change from activation control to slow diffusion control, and the impurities are trapped in the upper and middle layers. The bottom liquid cathode working area relies on the current focusing ear heating of the cathode bottom plate to maintain a high-temperature, low-viscosity state. Praseodymium and neodymium ions in the system penetrate to the bottom under the drive of the DC electric field and undergo electrochemical reduction, ensuring that the product metal is precipitated in liquid form.

[0014] In terms of interfacial rheology and electrical stability, a high-viscosity molten salt environment increases anode venting resistance. Introducing a trapezoidal wave step current and sequential feeding mitigates this phenomenon. During the relaxation period, reducing the cathode current density to a second preset current density weakens the anode polarization overpotential, alters the surface tension at the solid-liquid-gas interface, and loosens the gas film. Simultaneously, praseodymium-neodymium oxide powder and a fluxing and dust-removing compound containing lithium fluoride are sprayed at a low rate of electrochemical oxygen ion consumption. The powder material accumulates at the graphite anode interface and undergoes coordination reactions with fluorides within the system. The formation of complex ions disrupts the long-chain fluoride network structure within the molten salt system, causing a momentary decrease in the local microviscosity at the anode interface. The elimination of viscous resistance allows carbon dioxide and carbon monoxide gases retained on the anode surface to desorb and escape, eliminating the insulating gas film, stabilizing the electrolytic cell voltage, and preventing the occurrence of anode effects.

[0015] Preferably, the base electrolyte is composed of 55.0% to 65.0% NdF3, 15.0% to 20.0% PrF3 and 15.0% to 20.0% LiF by mass percentage.

[0016] By adopting the above technical solution, the composition ratio of the basic electrolyte meets the standard composition requirements of commercial praseodymium-neodymium alloys. The addition of a constant proportion of lithium fluoride can effectively reduce the primary crystallization temperature of the basic system, providing a thermodynamic basis for the subsequent establishment of a low-temperature, high-viscosity retention zone.

[0017] Preferably, the set electrode distance is 8.0cm to 13.0cm; the control of the upper layer temperature of the molten salt is specifically to control the upper layer temperature of the molten salt at a depth of 0 to 15cm from the surface of the molten salt to be stable at 960℃ to 980℃; the control of the bottom liquid cathode working area temperature is specifically to control the bottom liquid cathode working area at a depth of 0 to 10cm from the bottom of the tank to be stable at 1000℃ to 1020℃.

[0018] By employing the above technical solution, the defined electrode spacing range can physically isolate the interaction between the upper and lower fluid layers. The upper-middle layer temperature of 960℃ to 980℃ provides sufficient viscous resistance to inhibit the downward diffusion of heavy metal impurities. The bottom working zone temperature of 1000℃ to 1020℃ breaks through the liquidus critical value of praseodymium-neodymium alloy, preventing the solidification of the cathode product metal.

[0019] Preferably, the single electrolysis cycle is 30 to 40 minutes; the first preset current density is in the range of 4.5 A / cm². 2 ~5.5A / cm 2 The operating time maintained at the first preset current density is 23 to 32 minutes; the value range of the second preset current density is 0.8 A / cm. 2 ~1.2A / cm 2The relaxation period lasts for 2 to 4 minutes.

[0020] By adopting the above technical solution and setting specific time domain and current density parameters, the heat capacity and mass transfer rate of the electrolytic cell were matched. Sufficient cathode overpotential was provided during the first preset current density period, forcing rare earth ions to overcome viscosity resistance and complete deposition. The low current window during the relaxation period, combined with a ramp transition, mitigated the impact of large current switching on the DC power supply system and local thermal equilibrium, while providing sufficient time for powder feeding and complexation reactions.

[0021] Preferably, both the current decreasing slope and the current increasing slope are 1.5 A / (cm²). 2 ·min)~2.5A / (cm 2 ·min).

[0022] By adopting the above technical solution, the specific slope control ensures the smoothness of current switching, effectively avoids the impact of instantaneous current surges on the power grid and harmonic system, and at the same time provides time for the gas film at the anode interface to gradually loosen.

[0023] Preferably, the particle size of the coarse crystalline slag powder is between 100 and 800 μm; the coarse crystalline slag powder is added to the upper and middle layers of the electrolytic cell, specifically controlled to be 0.1 kg to 0.4 kg every 5 to 8 minutes.

[0024] By adopting the above technical solution, slag powder with a particle size of 100-800μm will not float on the surface and be carried away by the airflow, nor will it settle rapidly to the bottom cathode working area due to excessive weight. Therefore, it can be stably suspended and slowly melted in the middle and upper layers, playing a role in local heat absorption and cooling, as well as trapping impurities. The specific feeding frequency and single feeding amount are matched with the heat capacity of the molten salt, preventing drastic fluctuations in the local temperature of the system.

[0025] Preferably, the preparation steps of the coarse crystalline slag powder include: feeding the blocky slag cleaned after furnace shutdown into a jaw crusher for coarse crushing, then into a double roll crusher, and mechanically vibrating screen to retain and collect slag powder with a particle size range of 100-800μm, and discarding the slag block portion with a particle size greater than 800μm; feeding the slag block portion into a drying furnace, drying it at a temperature of 150℃-180℃ for 1-2 hours, and obtaining the coarse crystalline slag powder after discharge and cooling.

[0026] By adopting the above technical solution, through multi-stage crushing and mechanical screening, specific particle size blocks that meet the requirements of suspension melting were obtained; the subsequent drying process removed the moisture from the slag blocks, avoiding the entry of moisture into the high-temperature electrolytic cell and causing hydrolysis side reactions and safety hazards.

[0027] Preferably, the mixing of conventional industrial praseodymium-neodymium oxide powder with fluxing and dust removal compound is carried out at a mass ratio of 3:1 to 5:1; the spraying onto the molten salt surface around the graphite anode is carried out at a feeding rate of 0.5 kg / min to 1.5 kg / min.

[0028] By adopting the above technical solution, the limited mass ratio ensures that the oxide contains an appropriate amount of lithium fluoride as a cooling flux and complexation reaction promoter; the specific feeding rate matches the oxygen ion consumption rate during the low current window period, ensuring that the powder can be rapidly enriched at the anode interface and induce rheological structure collapse.

[0029] Preferably, the preparation steps of the fluxing dust removal compound include: placing the raw dust collected from the bag filter in the electrolysis workshop into a vacuum rotary kiln, and dehydrating it for 2 to 4 hours under negative pressure conditions of 200℃~250℃ and 0.05MPa~0.08MPa; then sending the dehydrated material into a homogenization bin, adding industrial anhydrous LiF particles as a flowability regulator at 2.0%~5.0% of the original dust mass, starting a forced mixer to mix for 30 to 60 minutes, and sealing and storing the material after discharge to obtain the fluxing dust removal compound.

[0030] By adopting the above technical solution, the negative pressure dehydration process removes the moisture adsorbed in the dust collector ash, preventing hydrolysis in the furnace; the addition of a small amount of industrial anhydrous LiF particles as a coarse particle flow aid breaks down the agglomeration and bridging between ultrafine dust particles, significantly improves the flowability of the powder, and ensures the smooth operation of the pneumatic injection feeding system.

[0031] Preferably, during the continuous single electrolysis cycle, the liquid praseodymium-neodymium metal at the bottom of the electrolytic cell is periodically extracted into ingots through a vacuum siphon, and the inert waste crust formed by the accumulation of impurities in the coarse crystal clear slag powder that have not been electrolytically reduced on the side wall and non-working area of ​​the electrolytic cell is concentrated and removed during the furnace shutdown and overhaul.

[0032] By adopting the above technical solution, the separation and extraction path of the product metal and the retained impurities was clarified. The liquid metal maintained at a high temperature at the bottom met the operational requirements of conventional vacuum siphon, ensuring continuous production; while the retained impurities were eventually enriched into a shell and then centrally stripped off during the shutdown period, realizing a complete process flow of closed-loop recovery of valuable elements and targeted removal of impurities within the system.

[0033] This invention provides a green smelting process for rare earth metals. It has the following beneficial effects:

[0034] 1. This invention increases the electrode spacing by adjusting the anode and constructs a vertically stratified microenvironment with a lower upper temperature (960℃~980℃) and a higher bottom temperature (1000℃~1020℃) in the electrolytic cell. The low temperature significantly increases the kinematic viscosity of the molten salt in the middle and upper layers, physically limiting the downward diffusion rate of heavy metal impurities released from the melting of coarse crystalline slag powder. At the same time, the high temperature at the bottom maintains the liquid properties of the praseodymium-neodymium alloy. Combined with the cathode overpotential provided by the high current density, rare earth ions can be selectively reduced and deposited, fundamentally preventing the co-deposition degradation caused by the direct return of high-impurity solid waste to the furnace, thus ensuring the purity of the metal product.

[0035] 2. This invention introduces a control method that couples trapezoidal wave step current with time-sequential feeding. During the window when the current density smoothly decreases to the low-frequency relaxation period, the anode polarization weakens. At this time, the conventional powder injected at the same time undergoes a coordination reaction with the lithium fluoride-containing flux dust removal compound to generate RE-OF complex ions. This local microscopic chemical reaction breaks the long-chain network of fluoride, causing the viscosity of the anode interface to drop instantaneously. The trapped gas film can be desorbed and discharged, maintaining stable tank pressure and continuous safe operation of the electrical control system.

[0036] 3. This invention eliminates the high equipment investment and chemical reagent consumption of traditional hydrometallurgical processes. It directly digests the slag and dust-recovered materials within the existing electrolytic cell system. Through step electrolysis with specific parameters and fluxing adjustment of the compound material, the valuable components of fluorine, lithium, and rare earth elements in the waste residue are reanalyzed and integrated into the electrolyte network. This replaces part of the need for adding fresh praseodymium oxide and neodymium powder and fresh lithium fluoride. From the perspective of material conservation, it achieves efficient closed-loop utilization of waste purification and fluorine and lithium resources, and reduces the unit consumption of new smelting materials. Attached Figure Description

[0037] Figure 1 This is a diagram showing the vertical depth temperature distribution characteristics inside the electrolytic cell of the present invention;

[0038] Figure 2 This is a diagram showing the stratified distribution of impurity (Fe and Si) concentrations within the electrolytic cell of the present invention at vertical depth.

[0039] Figure 3 This is a graph showing the dynamic cell pressure and the transient response characteristics of anode gas phase evacuation during a single electrolysis cycle of this invention.

[0040] Figure 4 This is a bar chart comparing the Fe and Si impurity contents in praseodymium and neodymium metals produced by each experimental group of this invention.

[0041] Figure 5 This is a comparison chart of the liquid extraction yield and the thermodynamic state evaluation of the bottom of the tank for each experimental group of the present invention.

[0042] Figure 6This is a statistical comparison chart of the electrical stability and frequency of anodic effect occurrence in the electrolysis process of each experimental group in this invention;

[0043] Figure 7 This is a comparison chart showing the absolute consumption and relative savings of producing one ton of praseodymium-neodymium metal virgin material according to the present invention. Detailed Implementation

[0044] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings, preparation examples, embodiments, comparative examples, and test examples. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0045] Preparation Examples 1-3:

[0046] Preparation Example 1:

[0047] This preparation example provides a method for preparing a customized pretreatment material (comprising component A and component B) for electrolysis, comprising the following steps:

[0048] (1) Preparation of component A (fluxing and dust removal compound):

[0049] The raw dust collected from the bag filter in the electrolysis workshop (mainly containing LiF and praseodymium-neodymium oxide powder) was placed in a vacuum rotary kiln and dehydrated for 2 hours under negative pressure conditions of 200℃ and 0.05MPa. The dehydrated material was then fed into a homogenization silo, and industrial anhydrous LiF particles were added at 2.0% of the original dust mass as a flowability modifier. A forced mixer was started and mixed for 30 minutes. After discharge, the mixture was sealed and stored to obtain component A.

[0050] (2) Preparation of component B (coarse crystal clear slag powder):

[0051] The lumpy slag collected after furnace shutdown is fed into a jaw crusher for coarse crushing, followed by a double roll crusher and mechanical vibrating screening. Slag powder with a particle size range of 100–800 μm is retained and collected, while slag lumps larger than 800 μm are discarded. This portion of slag lumps is then fed into a drying oven and dried at 150°C for 1 hour. After cooling, component B is obtained.

[0052] Preparation Example 2:

[0053] This preparation example provides a method for preparing a customized pretreatment material (comprising component A and component B) for electrolysis, comprising the following steps:

[0054] (1) Preparation of component A (fluxing and dust removal compound):

[0055] The raw dust collected from the bag filter in the electrolysis workshop was placed in a vacuum rotary kiln and dehydrated for 3 hours under negative pressure conditions of 225℃ and 0.065MPa. It was then fed into a homogenization chamber, where 3.5% (by weight of the raw dust) of anhydrous LiF granules were added as a flowability modifier. A forced mixer was started and mixed for 45 minutes. After discharge, the mixture was sealed and stored to obtain component A.

[0056] (2) Preparation of component B (coarse crystal clear slag powder):

[0057] The lumpy slag collected after furnace shutdown is fed into a jaw crusher for coarse crushing, then into a double roll crusher, and finally subjected to mechanical vibrating screening. Slag lumps with a particle size range of 30-40 mm are retained and collected, while the remaining particles are discarded. This portion of slag lumps is then fed into a drying oven and dried at 165℃ for 1.5 hours. After cooling, component B is obtained.

[0058] Preparation Example 3:

[0059] This preparation example provides a method for preparing a customized pretreatment material (comprising component A and component B) for electrolysis, comprising the following steps:

[0060] (1) Preparation of component A (fluxing and dust removal compound):

[0061] The raw dust collected from the bag filter in the electrolysis workshop was placed in a vacuum rotary kiln and dehydrated for 4 hours under negative pressure conditions of 250℃ and 0.08MPa. It was then fed into a homogenization chamber, where 5.0% (by weight of the raw dust) of anhydrous LiF granules were added as a flowability modifier. A forced mixer was started and mixed for 60 minutes. After discharge, the mixture was sealed and stored to obtain component A.

[0062] (2) Preparation of component B (coarse crystal clear slag powder):

[0063] The lumpy slag collected after furnace shutdown is fed into a jaw crusher for coarse crushing, then into a double roll crusher, and finally subjected to mechanical vibrating screening. Slag lumps with a particle size range of 40-50 mm are retained and collected, while the remaining particles are discarded. This portion of slag lumps is then fed into a drying oven and dried at 180℃ for 2 hours. After cooling, component B is obtained.

[0064] Examples 1-3:

[0065] Example 1:

[0066] This embodiment provides a green smelting process for rare earth metals, including the following steps:

[0067] (1) Initialization of vertical temperature field and flow field:

[0068] A basic electrolyte (by mass percentage: 55.0% NdF3, 25.0% PrF3, and 20.0% LiF) is added to an electrolytic cell equipped with a variable frequency DC rectifier and an electrode gap adjustment device, and then energized to melt it. The vertical insertion depth of the graphite anode is adjusted upwards using a mechanical lifting mechanism, and the electrode gap (ACD) is set to 8.0 cm. The cell voltage is adjusted to control the temperature of the upper layer of molten salt (0–15 cm from the melt surface) to be stable at 960°C, and the temperature of the bottom liquid cathode working area (0–10 cm from the bottom of the cell) to be stable at 1000°C.

[0069] (2) Trapezoidal wave step electrolysis and main deposition period:

[0070] Set a single electrolysis cycle time of 30 minutes. Initiate the main deposition stage, adjust the rectifier to increase and maintain the cathode current density at 4.5 A / cm². 2 The process continued for 23 minutes. During this period, component B (coarse crystal clear slag powder) obtained in Preparation Example 1 was fed into the upper and middle layers of the electrolytic cell at a rate of 2.0 kg / batch, every 5 minutes.

[0071] (3) Gradual descent of slope and interface drainage:

[0072] After the main deposition period, the rectifier was controlled at 1.5 A / (cm²). 2 The current drop slope (·min) smoothly reduced the cathode current density to 0.8 A / cm² within approximately 2.5 minutes. 2 .

[0073] (4) Low-frequency relaxation and starvation-style synchronous feeding period:

[0074] Maintain the cathode current density at 0.8 A / cm² 2 The relaxation period lasts for 2 minutes. During this time window, the pneumatic injection feeding system is activated to mix praseodymium oxide neodymium powder with component A (fluxing and dust removal compound) prepared in Preparation Example 1 at a mass ratio of 3:1, and spray it onto the molten surface of the area where the graphite anode is located at a feeding rate of 0.5 kg / min.

[0075] (5) Slope recovery and metal extraction:

[0076] After feeding is completed, the rectifier operates at 1.5A / (cm²). 2 The rising slope (·min) pulled the current density back up to 4.5 A / cm² within approximately 2.5 minutes. 2The liquid praseodymium-neodymium metal at the bottom of the electrolytic cell is periodically extracted into ingots through a vacuum siphon. The inert waste crust formed by the accumulation of impurities that were not reduced by electrolysis in the coarse crystal clear slag powder on the side walls and in the non-working area is removed during the furnace shutdown and overhaul.

[0077] Example 2:

[0078] This embodiment provides a green smelting process for rare earth metals, including the following steps:

[0079] (1) Initialization of vertical temperature field and flow field:

[0080] A basic electrolyte (by mass percentage: 60.0% NdF3, 20.0% PrF3, and 20.0% LiF) was added to an identical 10,000-ampere electrolytic cell and energized to melt it. The depth of the graphite anode was adjusted using a mechanical lifting mechanism, and the electrode gap (ACD) was set to 10.5 cm. The cell voltage was adjusted to maintain a stable temperature of 970°C in the upper layer of the molten salt and a stable temperature of 1010°C in the bottom liquid cathode working area.

[0081] (2) Trapezoidal wave step electrolysis and main deposition period:

[0082] The single electrolysis cycle was set to 35 minutes. The main deposition stage was initiated, and the rectifier was adjusted to increase and maintain the cathode current density at 5.0 A / cm². 2 The process was continued for 28 minutes. During this period, component B (coarse crystal clear slag powder) obtained in Preparation Example 2 was fed into the upper and middle layers of the electrolytic cell at a rate of 3.0 kg / batch every 6.5 minutes through the automatic feeding port.

[0083] (3) Gradual descent of slope and interface drainage:

[0084] After the main deposition period, the rectifier is controlled at 2.0 A / (cm²). 2 The current drop slope (·min) smoothly reduced the cathode current density to 1.0 A / cm² within 2 minutes. 2 .

[0085] (4) Low-frequency relaxation and starvation-style synchronous feeding period:

[0086] Maintain the cathode current density at 1.0 A / cm² 2 The relaxation period lasts for 3 minutes. During this time window, the pneumatic jet feeding system is activated to mix conventional industrial praseodymium-neodymium oxide powder with component A (fluxing and dust removal compound) prepared in Preparation Example 2 at a mass ratio of 4:1, and spray it onto the molten surface around the graphite anode at a feeding rate of 1.0 kg / min.

[0087] (5) Slope recovery and metal extraction:

[0088] After feeding is completed, the rectifier operates at 2.0 A / (cm²). 2 The rising slope of (·min) pulled the current density back up to 5.0 A / cm² within 2 minutes. 2 This completes a single cycle. The liquid metal at the bottom is periodically siphoned out, and the accumulated waste residue is cleaned according to regulations.

[0089] Example 3:

[0090] This embodiment provides a green smelting process for rare earth metals, including the following steps:

[0091] (1) Initialization of vertical temperature field and flow field:

[0092] A basic electrolyte (by mass percentage: 65.0% NdF3, 15.0% PrF3, and 20.0% LiF) was added to an identical 10,000-ampere electrolytic cell and melted by applying electricity. The depth of the graphite anode was adjusted using a mechanical lifting mechanism, and the electrode gap (ACD) was set to 13.0 cm. The cell voltage was adjusted to maintain a stable temperature of 980°C in the upper layer of the molten salt and a stable temperature of 1020°C in the bottom liquid cathode working area.

[0093] (2) Trapezoidal wave step electrolysis and main deposition period:

[0094] The single electrolysis cycle was set to 40 minutes. The main deposition stage was initiated, and the rectifier was adjusted to increase and maintain the cathode current density at 5.5 A / cm². 2 The process was continued for 32 minutes. During this period, component B (coarse crystal clear slag powder) obtained in Preparation Example 3 was fed into the upper and middle layers of the electrolytic cell every 8 minutes at a rate of 4.0 kg / batch through the automatic feeding port.

[0095] (3) Gradual descent of slope and interface drainage:

[0096] After the main deposition period, control the rectifier at 2.5 A / (cm²). 2 The current drop slope (·min) smoothly reduced the cathode current density to 1.2 A / cm² within approximately 1.7 minutes. 2 .

[0097] (4) Low-frequency relaxation and starvation-style synchronous feeding period:

[0098] Maintain the cathode current density at 1.2 A / cm². 2 The relaxation period lasts for 4 minutes. During this time window, the pneumatic jet feeding system is activated to mix conventional industrial praseodymium-neodymium oxide powder with component A (fluxing and dust removal compound) prepared in Preparation Example 3 at a mass ratio of 5:1, and spray it onto the molten surface around the graphite anode at a feeding rate of 1.5 kg / min.

[0099] (5) Slope recovery and metal extraction:

[0100] After feeding is completed, the rectifier operates at 2.5A / (cm²). 2 The rising slope (·min) pulled the current density back up to 5.5 A / cm² within approximately 1.7 minutes. 2 This completes a single cycle. The liquid metal at the bottom is periodically siphoned out, and the accumulated waste residue is cleaned according to regulations.

[0101] Comparative Examples 1-4:

[0102] Comparative Example 1:

[0103] Compared to Example 2, the differences are as follows: the electrode spacing was not increased (the electrode spacing was set to the conventional 5.0 cm), a vertical temperature field was not set (the global temperature was kept constant at 1050℃), and trapezoidal wave step electrolysis was not used, maintaining a constant A / cm throughout the process. 2 The constant DC electrolysis was used; instead of sequential feeding, components A and B were mixed with praseodymium-neodymium oxide powder and then added continuously and uniformly; the rest were the same.

[0104] Comparative Example 2:

[0105] Compared with Example 2, the difference is that the electrode spacing was not increased (the electrode spacing was set to the conventional 5.0 cm), and a vertical temperature field of "cold at the top and hot at the bottom" was not constructed. Instead, the global temperature of the electrolytic cell was uniformly and forcibly cooled and controlled at 970°C; all other aspects are the same.

[0106] Comparative Example 3:

[0107] Compared to Example 2, the difference lies in that: no trapezoidal wave step current and low-frequency relaxation period were set, and the current was maintained at 5.0 A / cm throughout the entire electrolysis process (including the injection feeding period). 2 The constant high current electrolysis is used; everything else is the same.

[0108] Comparative Example 4:

[0109] Compared to Example 2, the difference lies in that the feeding method is not time-synchronized with the low-frequency relaxation period. The mixture of component A and conventional praseodymium-neodymium oxide powder is not concentratedly sprayed during the relaxation period, but is continuously and uniformly added to the surface of the melt throughout the entire 35-minute electrolysis cycle; all other aspects are the same.

[0110] Test Examples 1-6:

[0111] Test Example 1:

[0112] This test case relies on the operating environment of Example 2. After the electrolytic cell reaches steady-state circulation, in-situ physical parameter measurements and fixed-point sampling analysis are performed on the melt in the cell. Specifically, the following operations are included:

[0113] In Example 2, at the 15th minute of the main deposition period, a customized integrated multi-point temperature measurement array (equipped with three S-type platinum-rhodium-platinum thermocouples, with the vertical spacing of the measuring points pre-calibrated to 5cm, 15cm, and 30cm from the liquid surface) was vertically inserted into the melt in the working area of ​​the electrolytic cell. After standing for 120 seconds until the temperature readings showed no significant fluctuations, the real-time temperature at each depth was read and recorded.

[0114] A vacuum quartz sampling tube with a valve and graduations on the outer wall was vertically lowered to a depth of 5 cm from the liquid surface. The valve was opened, and approximately 50 g of molten sample was instantly extracted using the negative pressure inside the tube. This sample was then quickly pulled from the liquid surface and cooled and quenched in an argon atmosphere to obtain a shallow solid molten salt sample. The sampling tube was replaced, and the operation was repeated to obtain solid molten salt samples at depths of 15 cm and 30 cm (near the bottom liquid metal cathode area), respectively.

[0115] The three solid molten salt samples obtained above were ground in tungsten carbide mortars and passed through a 200-mesh standard sieve. 2.000g of sample was accurately weighed from each powder group and placed in a polytetrafluoroethylene crucible. A mixture of nitric acid and perchloric acid was added, and the mixture was digested on a temperature-controlled hot plate until the solution was clear. After cooling, the solution was transferred to a 100mL volumetric flask and diluted to volume.

[0116] Start the inductively coupled plasma optical emission spectrometer (ICP-OES) and calibrate the instrument using the standard addition method. Inject the three sets of test solutions after volume adjustment sequentially, and determine the concentrations of Fe and Si elements. Combine this with the dilution factor to convert the impurity mass concentration in the original molten salt sample.

[0117] Table 1. Test data on temperature and impurity element concentration distribution at vertical depth

[0118]

[0119] Test conclusion:

[0120] Based on the data in Table 1 and the corresponding appendix Figure 1 As shown in the (vertical depth temperature distribution diagram), the melt inside the electrolytic cell exhibits a clear difference in thermal field distribution in the vertical direction. (Attached) Figure 1 The horizontal axis represents the vertical depth (cm) from the melt surface, and the vertical axis represents the measured real-time temperature (°C). The black solid line and gray-filled circular markers clearly show that the temperature curve exhibits a significant upward trend as the measuring depth increases from 5cm to 30cm. Specifically, the temperature in the shallow region at a depth of 5cm is 968.4°C, while the temperature in the bottom cathode working area at a depth of 30cm rises to 1012.7°C, resulting in a thermodynamic temperature difference of 44.3°C between the two regions. Spatial differentiation of the thermodynamic data and its attachments... Figure 1The visually confirmed characteristics demonstrate that setting the electrode spacing to increase cuts off the strong convective heat transfer between the upper and lower parts of the melt, successfully creating a "cold at the top and hot at the bottom" spatial microenvironment and a relatively stable vertical temperature gradient within a single electrolytic cell.

[0121] The element detection results for the corresponding depths are shown in Table 1 and appendix. Figure 2 (Vertical depth impurity concentration distribution diagram shown). (Attached) Figure 2 The x-axis represents vertical depth (cm), and the y-axis uses a logarithmic scale to clearly show the mass concentration (mg / kg) of impurity elements across multiple orders of magnitude. In the graph, a solid line with a dark gray filled square represents the concentration trend of Fe (values ​​are marked directly above the breakpoint), and a dashed line with a pure white filled diamond represents the concentration trend of Si (values ​​are marked directly below the breakpoint). The graph clearly shows that both curves exhibit a steep downward trend. The concentrations of Fe and Si in the shallow melt are 5842.1 mg / kg and 3415.7 mg / kg, respectively. However, moving towards the bottom layer (30cm), the concentrations of these elements experience a precipitous drop of two orders of magnitude, decreasing sharply to 45.8 mg / kg and 28.4 mg / kg, respectively. The lines are independent of each other, and the data corresponds precisely to the breakpoints.

[0122] This order-of-magnitude gradient distribution of substance concentration quantifies and confirms the local kinetic hindrance mechanism induced by the cooling and viscosity-increasing strategy. The thermodynamic state of the shallow layer at 968.4℃ leads to an increase in the kinematic viscosity of the molten salt in this region. According to Fick's first law, the diffusion coefficient of impurity ions in this microenvironment is physically inhibited. Heavy metal and non-metallic impurity ions released from the melting of the slag accumulate and are trapped here. Limited by viscous resistance and the restricted shallow magnetohydrodynamic eddies, their mass transfer changes from electrochemical control to slow diffusion control, and they cannot effectively penetrate the middle transition zone for downward mass transfer. The state of the bottom layer at 1012.7℃ maintains the low viscosity of the original molten salt composition of the system, ensuring that the product metal is in a liquid state. The data and charts objectively demonstrate that the process steps of increasing the electrode spacing and controlling the local temperature can achieve the decoupling of thermodynamics and mass transfer kinetics in the spatial domain, blocking the mass transfer path of impurities to the bottom metal phase.

[0123] Test Example 2:

[0124] This test case relies on the steady-state operating environment of Example 2 to perform high-frequency synchronous acquisition of electrical parameters and exhaust gas emission status during the operation of the electrolytic cell. The specific operation is as follows:

[0125] Connect the data acquisition port of the variable frequency DC rectifier to the industrial control computer and set the data sampling frequency to 10Hz. Record the cell voltage (inter-electrode voltage) in real time for the entire 35-minute trapezoidal wave electrolysis cycle set in Example 2, and perform moving average filtering on the raw data to eliminate high-frequency electromagnetic interference.

[0126] At the main outlet pipe of the closed exhaust hood of the electrolytic cell, a Pitot tube flow meter and a non-dispersive infrared gas analyzer (NDIR) probe are installed sequentially along the flow direction. The instrument range and baseline are calibrated, and the analog signals from the flow sensor and concentration sensor are synchronously connected to the industrial control computer.

[0127] The system algorithm is set to perform product and integral calculations on the pipe cross-sectional area, flow velocity, and instantaneous concentrations of CO and CO2, and output the instantaneous standard flow data of the mixed gas phase (CO+CO2) generated in the anode working area in real time.

[0128] Record a complete dataset for three consecutive 35-minute electrolysis cycles. Select one typical cycle and extract the corresponding data of tank pressure and gas flow rate at key time nodes such as the middle and end of the main deposition period, the gradual descent of the slope, the entry into the relaxation period, the moment of feeding, the stable section of the relaxation period, and the slope recovery period for tabular analysis.

[0129] Table 2. Dynamic cell pressure and anode gas flow rate test data within a typical trapezoidal wave period.

[0130]

[0131] Test conclusion:

[0132] Based on the data in Table 2 and the corresponding appendix Figure 3 As shown, within a single 35-minute trapezoidal wave electrolysis cycle, the real-time cell voltage and the anode gas phase venting flow rate exhibit a highly bound transient coupling and nonlinear response. (See attached image.) Figure 3 The horizontal axis (X-axis) represents the operating time points, ranging from 0 to 35 minutes. The chart is clearly divided into four operating intervals by vertical gray dashed lines: main deposition period (0-28 min), slow descent period (28-30 min), relaxation and feeding period (30-33 min), and recovery period (33-35 min). The labels for each stage are arranged at the top of the chart. The left vertical axis (left Y-axis) corresponds to the solid black line (with a square marker filled in dark gray), representing the real-time cell voltage (V), whose value trend reflects the polarization and gas film accumulation state of the anode surface; the right vertical axis (right Y-axis) corresponds to the dashed black line (with a diamond marker filled in pure white), representing the instantaneous flow rate of the CO+CO2 mixed gas phase (L / min), whose value trend reflects the physical rate of desorption and evacuation of insulating bubbles.

[0133] During the main deposition period of 0 to 28 minutes, the cathode current density was maintained at 5.0 A / cm². 2 The high position. (Refer to Table 2 and Appendix) Figure 3 As can be seen, the solid line of the cell pressure shows a slow upward trend. At 10.5, 25.2, and 27.8 minutes, the real-time cell voltage gradually increased from 5.83V to 6.42V; while the dashed line of gas flow rate remained low and continued to decline, decreasing from 42.6L / min to 35.1L / min. The monotonous increase in voltage and the low gas flow rate directly reflect the polarization phenomenon caused by the obstruction of bubble desorption under high viscosity conditions. This indicates that the high viscosity physical environment of 970℃ in the shallow layer restricts the fluid mass transfer in the anode gas generation process, causing the insulating gas to stagnate and accumulate on the graphite anode surface, increasing the gas film coverage and gradually increasing the contact resistance inside the cell.

[0134] During the 28-30 minute ramp descent period, the cell voltage dropped back to 4.85V, and the gas flow rate rose slightly to 54.8L / min, indicating that the anodic polarization overpotential weakened as the current decreased, the surface tension of the gas-liquid-solid three-phase interface changed, and the gas film showed initial loosening.

[0135] During the "relaxation and feeding" phase at 30-33 minutes, the tank pressure dropped to 3.31V. Particularly at the feeding instant at 30.6 minutes, the data in Table 2 and the attached table show this. Figure 3 The two curves in the figure exhibit a dramatic transient reversal response: the solid line of the tank voltage drops precipitously, reaching a low trough of 3.04V; while the dashed line of the gas phase flow rate forms an extremely sharp abrupt peak, with an instantaneous surge in flow rate to a peak of 215.7 L / min. The values ​​of key nodes in the figure have been clearly labeled with misaligned text, completely eliminating visual interference.

[0136] The surge in parameters at the moment of feeding confirms the intervention of coordination competition at the material level on local rheology. At 1.0 A / cm 2 At low current densities, the electrochemical reaction consumes O 2- The flow rate was lower than the feeding rate. Conventional praseodymium-neodymium oxide powder injected into the anode working area, along with the fluxing and dust removal compound, accumulated significantly at the interface and formed RE-OF complex ions with fluorides within the system. The instantaneous formation of the complex disrupted the long-chain network of fluorides, causing a structural collapse of the microviscosity in a localized area of ​​the anode interface. The graphs, through bidirectional abrupt changes in the parametric curves, quantitatively demonstrate that the flow reduction and feeding operations at a specific time sequence can instantly break the viscous constraints at the anode interface, achieving concentrated desorption and venting of the retained gas film. The sudden drop in viscous resistance completely removed the physical constraints on gas phase desorption, instantly venting previously retained bubbles, forming a huge peak in gas flow rate, and completely eliminating the additional ohmic pressure drop caused by the gas film. By 31.8 minutes, the tank pressure stabilized at 3.12V, and the gas flow rate dropped back to a low level of 26.3L / min, indicating that the gas film had been completely cleared.

[0137] The test results demonstrate from the perspective of macroscopic engineering data that the decoupling of electrolysis parameters in the time domain, combined with the addition of powder at a specific time sequence, can effectively control the rheological state of the interface microenvironment and avoid the occurrence of irreversible anodic effects under high-temperature and high-viscosity impurity trapping processes.

[0138] Test Example 3:

[0139] In this test case, after the electrolytic cells of Examples 1 to 3 and Comparative Examples 1 to 4 had been running continuously for 72 hours, destructive sampling and elemental quantitative analysis of the produced praseodymium-neodymium metal were performed. The specific procedures are as follows:

[0140] Liquid praseodymium-neodymium metal extracted from each experimental group using a vacuum siphon was collected, cast, and cooled into standard metal ingots. For Comparative Example 2, where siphon failure was caused by furnace bottom slagging, solidified metallic alloy blocks were mechanically peeled directly from the bottom of the tank during furnace shutdown as test samples.

[0141] Using a bench drill with a carbide drill bit, holes 10 mm deep were drilled at the center and at three equal points from the edge of each metal ingot (or alloy block). Metal shavings generated during drilling were collected, ultrasonically cleaned with anhydrous ethanol for 3 minutes to remove surface mechanical oil, and then dried in a vacuum drying oven. The metal shavings from the same batch were mixed thoroughly and used as the representative sample for this experimental group.

[0142] Accurately weigh 0.500g of the metal fragment sample to be tested into a Teflon dissolution vessel. Slowly add a mixture of 5mL high-purity nitric acid and 2mL hydrofluoric acid, and heat under reflux for 2 hours on a temperature-controlled heating plate at 120℃. After the sample is completely dissolved and the solution is clear, cool to room temperature, wash the inner wall of the dissolution vessel several times with ultrapure water, and transfer the solution to a 50mL volumetric flask and dilute to volume.

[0143] Turn on the inductively coupled plasma mass spectrometer (ICP-MS), adjust the instrument parameters using tuning fluid to achieve optimal sensitivity, and activate the collisional reaction cell (KED) mode to eliminate polyatomic ion mass spectrometry interference. Dilute the prepared sample solution to the appropriate volume and inject it to determine the Fe isotopes in the solution. 56 Fe) and Si isotopes ( 28 The signal response value of Si was used to calculate the mass fraction of Fe and Si in the products of each experimental group based on the internal standard method and standard curve.

[0144] Table 3. Test data of Fe and Si impurities in the metal products of each experimental group

[0145]

[0146] Test conclusion:

[0147] Based on the data in Table 3 and the corresponding appendix Figure 4 As shown, after 72 hours of continuous operation, the mass fractions of Fe and Si impurities in the praseodymium-neodymium metal produced by each experimental group in Examples 1 to 3 and Comparative Examples 1 to 4 are clearly compared. (Appendix) Figure 4 The horizontal axis represents different experimental groups, and the vertical axis represents the mass fraction (mg / kg) of impurities in the extracted samples. Given that the values ​​in Comparative Example 1 and the Examples span several orders of magnitude, a logarithmic scale was used on the vertical axis to ensure the visibility of all bars. Each group in the figure contains two parallel bars: the dark gray bar represents the mass fraction of Fe, and the light gray bar represents the mass fraction of Si. The specific detection value is precisely labeled above the top of each bar, and the staggered arrangement of the text does not cause interference.

[0148] A macroscopic comparison reveals that the columns corresponding to Examples 1 to 3 are generally in the low-level range, and the mass fractions of Fe and Si elements in the produced metal are effectively controlled. Among them, the dark gray and light gray columns of Example 2 have the lowest height and the best purification effect, with Fe content reduced to 275.2 mg / kg and Si content reduced to 126.8 mg / kg, far below the upper limit requirement of the industry's first-grade standard. The data of Examples 1 and 3 show fluctuations within a reasonable range, but also remain at a low level. The purification effect verifies that the process steps of increasing the electrode distance and constructing a vertical temperature gradient can form a stable laminar flow protective layer above the cathode region. The upper high-viscosity region performs the expected impurity trapping function, and the heavy metals and non-metallic impurities introduced by the slag cleaning do not undergo large-scale mass transfer and diffusion to the underlying metal phase.

[0149] In contrast, Comparative Example 1, employing a traditional constant short electrode gap and global high-temperature process, produced an unusually tall column, visually demonstrating the complete loss of control over impurities and its deteriorated state. The Fe and Si contents reached as high as 3642.8 mg / kg and 2154.6 mg / kg, respectively. In the traditional microenvironment without spatiotemporal decoupling, the high temperature resulted in an overall low viscosity of the melt. Combined with the intense macroscopic convection of magnetohydrodynamics caused by a constant high current, the diffusion control layer of impurity ions was disrupted. Impurities originally remaining in the molten salt phase were directly entrained into the cathode surface and underwent co-deposition reactions, leading to severe degradation of product purity.

[0150] Comparative Example 2 employed a global cooling process, and although the corresponding column was short (the Fe and Si contents of the stripped sample were 312.4 mg / kg and 148.2 mg / kg, respectively), it confirmed the physical inhibition effect of cooling and thickening on impurity diffusion. However, considering the sampling method, this process caused the underlying metal to solidify simultaneously, resulting in a destructive stripping sample that completely lost its industrial siphon value and violated the technical requirements of continuous production.

[0151] Comparative Example 3 maintained a constant high current without introducing a trough relaxation period, while Comparative Example 4 did not synchronize the feeding action with the relaxation period. The column heights of both groups significantly exceeded the baseline of the examples. The Fe content in the product of Comparative Example 3 reached 1756.3 mg / kg, and the Si content was 1085.1 mg / kg. The lack of a relaxation and venting mechanism led to frequent anode effects. The thermal and flow fields within the tank experienced periodic loss of control under voltage surges. Turbulent hydrodynamic convection broke through the original laminar boundary protection zone at the bottom layer, allowing some impurities trapped in the upper layer to be back-rolled through to the cathode region and participate in the reduction reaction. The Fe content in Comparative Example 4 was 1142.9 mg / kg, and the Si content was 823.7 mg / kg. Continuous uniform feeding failed to trigger an instantaneous coordination reaction during the low current window. The accumulation of oxide powder at the anode interface altered the local rheological structure of the upper melt. Spatiotemporal lockout caused local breaches in the hydrodynamic barrier of the high-viscosity trapping zone, reducing the system's physical barrier efficiency for downward mass transfer of impurities.

[0152] The clear quantitative comparisons in the charts and the degradation performance under different conditions objectively demonstrate that the three-dimensional synergistic linkage mechanism of spatial domain stratification, temporal domain electric field control, and feeding sequence coupling is indispensable in the process. The synergistic linkage between various parameters not only ensures the thermal balance and continuous operation capability of the electrolytic cell, but also fundamentally blocks the electrochemical deposition path of impurities to the product, which has an absolute advantage in improving product purity and solves the technical pain point of directly returning high-impurity fluorine-containing solid waste to the furnace for purification.

[0153] Test Example 4:

[0154] This test case quantitatively evaluates the thermodynamic steady-state maintenance capability and liquid metal extraction efficiency of each experimental group during a continuous 72-hour electrolysis operation cycle and the shutdown and cleaning phase. The specific operation is as follows:

[0155] Based on Faraday's law of electrolysis, the total ampere-hour integral data of the rectifiers for each experimental group during a 72-hour operating cycle were extracted. Combined with the historical Faraday current efficiency benchmark value of the electrolytic cell, the theoretical total weight of praseodymium and neodymium metal produced for each group was calculated.

[0156] During 72 hours of continuous operation, liquid praseodymium-neodymium metal accumulated at the bottom of the tank is periodically extracted using a vacuum siphon according to the standard process cycle. The weight of the metal ingot extracted each time is weighed on an electronic weighbridge, and the total weight extracted is calculated. The success rate of the liquid metal vacuum siphon is calculated by dividing the total actual weight extracted by the theoretical total output weight.

[0157] After the 72-hour operation period, disconnect the DC power supply to shut down the furnace. While still hot, use a negative pressure suction device to completely drain the liquid fluoride electrolyte from the upper layer of the tank, exposing the cathode substrate and deposits at the bottom of the electrolytic cell. Allow the tank to cool naturally to room temperature.

[0158] Using a carbide mechanical probe with millimeter-level depth graduations, five measurement coordinate points were selected at the four corners and the center of the working area at the bottom of the tank. The probe was inserted vertically downwards to pierce the residual solidified material at the bottom until it reached the surface of the graphite cathode substrate. The depth of penetration at each point was recorded, and the arithmetic mean of the five measurement points was taken as the average thickness of the solidified shell at the bottom of the tank for this experimental group.

[0159] Table 4. Test data on product yield and solidified crust thickness at the bottom of the tank for each experimental group.

[0160]

[0161] Test conclusion:

[0162] Based on the data in Table 4 and the corresponding appendix Figure 5 As shown, the coupled evaluation results of the bottom liquid metal extraction efficiency and the thermodynamic state of the tank bottom in each experimental group exhibit significant differences within a continuous 72-hour electrolysis operation cycle. (See attached image) Figure 5 The horizontal axis lists the experimental groups of Examples 1 to 3 and Comparative Examples 1 to 4 sequentially. The left vertical axis corresponds to a bar chart with medium gray fill, representing the success rate (%) of liquid metal vacuum siphoning. The higher this positive index value, the stronger the ability to maintain the liquid metal at the bottom. The right vertical axis corresponds to a thick black solid line marked with pure black diamonds, representing the average thickness (mm) of the solidified shell at the bottom of the tank. The higher this negative index value, the more severe the physical crusting phenomenon caused by local cooling. Observing the distribution of the bar chart and the line graph, it can be seen that the two show a strict negative physical correlation. The data labels of each group have been offset vertically according to the coordinate difference, and there is no visual interference with the bars or lines.

[0163] In Examples 1 to 3, the gray columns were all at a high level of over 90%, and the siphon success rate was maintained between 92.4% and 96.8% (with Example 2 reaching 96.8%). The corresponding black broken line points were suppressed to an extremely low level, almost close to the horizontal axis, and the average thickness of the hard shell at the bottom of the tank was controlled within 14.5 mm (only 4.2 mm in Example 2). The bottom metal exhibited excellent liquid fluidity, which directly proved that this process successfully maintained a liquid high-temperature microenvironment that exceeded the melting point critical value at the bottom, ensuring the continuity of industrial vacuum siphon operations. Quantitative data showed that the process intervention of increasing the electrode distance achieved thermal field separation in the spatial domain. While the upper melt retained impurities by cooling and thickening, the bottom cathode working area relied on the longitudinal physical isolation of the electrolytic cell and the current convergence at the bottom to maintain a stable high temperature of 1000 to 1020°C. The local high temperature at the bottom prevented the product from overcooling and solidifying.

[0164] Comparative Example 1, due to the use of a global high temperature of 1050℃ throughout the process, has a hard shell thickness of only 1.5mm at the bottom of the tank, achieving a siphon success rate of 97.5%. Its bottom thermodynamic indicators are good, but combined with the aforementioned test results, this state is achieved at the cost of sacrificing the mass transfer barrier against impurities, and the product impurities are in an uncontrolled state.

[0165] Comparative Example 2 employed a forced intervention involving global cooling to 970℃, resulting in an actual extracted weight of only 104.8 kg. (See attached...) Figure 5 In the process, the corresponding gray column experienced a precipitous drop, reducing the siphon success rate to 12.3%; its black zigzag points formed towering peaks, and the thickness of the solidified crust at the bottom of the tank increased dramatically to 185.6 mm. Although macroscopic global cooling physically suppressed impurity diffusion, it directly caused the cathode temperature to drop below the melting point of the biomass metal. Praseodymium and neodymium metals rapidly underwent a solid-phase transformation after precipitation, densely agglomerating at the bottom of the tank, causing conventional suction equipment to completely fail. Quantitative analysis revealed the serious engineering flaw of global cooling leading to large-area solidification and loss of siphon extraction value, confirming that single-dimensional temperature control cannot simultaneously meet the dual requirements of purification and extraction.

[0166] Comparative Example 3 lacked a low-frequency relaxation period, and Comparative Example 4 lacked feeding timing coupling. The siphon success rates of the two groups decreased to 78.6% and 81.4%, respectively, and the hard shell thickness increased to 45.2 mm and 38.7 mm, respectively. The broken line points in the graphs showed a significant upward tilt. The deterioration of these two sets of data corroborates that the frequent voltage oscillations caused by the lack of timing coupling can disrupt the steady state of the bottom thermal field. The failure to break the anodic gas film triggered a periodic anodic effect, and the violent oscillation of the cell voltage led to an imbalance in the Joule heat input power of the electrolytic cell. The thermal field disorder caused transient cold zones to appear in the originally constant high-temperature zone at the bottom. At the same time, poorly dissolved oxide powder and some high-melting-point slag blocks in the upper layer penetrated the hot layer and settled to the bottom of the cell under the action of gravity, becoming the physical nucleation sites for heterogeneous formation and inducing local eutectic solidification of the bottom metal.

[0167] Verification by comprehensive quantitative indicators shows that this process effectively eliminates the side effects of bottoming out associated with high viscosity purification processes by constructing a local high-temperature protective layer through spatial electrode spacing and maintaining a steady state of heat power input through time-domain step electrolysis. Under the premise of ensuring impurity retention rate, it achieves high-yield liquid extraction of products.

[0168] Test Example 5:

[0169] In this test case, during a continuous 72-hour electrolysis operation cycle, statistical analysis was performed on the correlation parameters characterizing the rheological state and electrical control stability based on the background data of the DC frequency converter control system of each experimental group. The specific operation is as follows:

[0170] Access the backend database of the electrolytic cell intelligent control system, export the real-time cell voltage log file during 72 hours of continuous operation, with the data sampling period set to 1 second. Clean the data and remove blank areas caused by power outages due to normal shutdowns or manual maintenance to obtain the effective electrolysis operation time.

[0171] An algorithm for setting the anodic effect trigger threshold is established. A data segment in which the instantaneous voltage jump exceeds the set reference voltage of the process stage by more than 3.5V and lasts for more than 10 seconds is defined as an anodic effect event. The total number of events that trigger this condition during the total 72-hour effective operating period is counted, and then divided by 3 to calculate the average daily frequency of anodic effects for each experimental group.

[0172] By removing the normal step current switching transition time (such as the ramp descent period and ramp return period) actively controlled by the frequency converter in the examples and some comparative examples, the cell voltage data of the steady-state electrolysis stage is extracted. The cumulative length of time for the cell voltage value to remain within the fluctuation range of ±5% of the theoretical target value is calculated. The cell voltage stability rate is calculated by dividing this cumulative time length by the total time of the steady-state electrolysis stage.

[0173] Table 5. Daily average frequency of anodic effect and tank pressure stability rate test data for each experimental group

[0174]

[0175] Test conclusion:

[0176] Based on the data in Table 5 and the corresponding appendix Figure 6 As shown, the experimental groups exhibited significant differences in electrical control stability and system rheological response. (See attached image) Figure 6 The graph shows a comparison of the correlation statistics for each group during a continuous 72-hour operation period. The horizontal axis of the graph lists Examples 1 to 3 and Comparative Examples 1 to 4 in sequence. The left vertical axis corresponds to a light gray filled bar chart, representing the cell pressure stability rate (%), which is the percentage of time the cell pressure remains within the set target value ±5% fluctuation range. The higher this positive indicator, the more stable the electrical operation. The right vertical axis corresponds to a black needle chart with a solid black circle mark, representing the average daily frequency of anode effects (times / day). The higher this negative indicator value, the more difficult it is to vent the anode interface and the more frequent the critical runaway caused by the insulating gas film. The data labels for each group have been finely misaligned according to the coordinate difference, and there is no visual interference with the bars or lines.

[0177] The charts clearly show a strong mirror-image relationship between the cell pressure stability rate (gray bars) and the anode effect frequency (black needles). In Examples 1 to 3, the gray bars are tall (stability rates all exceeding 95.7%), while the corresponding black needles are close to the bottom of the horizontal axis (effect frequency below 0.67 times / day; Example 2 even achieved zero alarms during operation). This distribution characteristic directly confirms that the combination of step electrolysis and sequential feeding in the time domain effectively eliminates the anode exhaust obstruction caused by localized cooling and thickening in the upper layer. The mechanism lies in the fact that the sudden drop in current density during the trough relaxation period reduces the polarization surface tension of the interface, allowing the synchronously sprayed oxide powder to operate at low O2 levels. 2- At the consumption rate, a coordination reaction occurs with the fluoride network, and the generated complex ions instantly cleave the long chains of macromolecules. This combination of "step current + sequential feeding" can instantly break the viscosity network. The sudden drop in local microvisual viscosity releases the physical constraints of bubble desorption, completely eliminates the exhaust obstacles caused by the high viscosity retention process, prevents the large-area accumulation of insulating gas film on the anode surface, and maintains a constant ohmic resistance in the tank.

[0178] In contrast, Comparative Example 3, which lacked a relaxation period and maintained a constant high current throughout, showed a collapse of the gray column to 45.3%, while the black needle-like dots protruded sharply, piercing the 20 mark (18.67 times / day). Due to the maintenance of a high current in a high-viscosity environment, the anode continuously generated gas at high frequencies without structural disruption of the system's viscous resistance. The bubble formation rate far exceeded the desorption and diffusion rate, and the gas formed a dense insulating film at the solid-liquid interface, triggering a critical anode effect and leading to voltage runaway.

[0179] In Comparative Example 4, which lacked synchronized feeding, the number of black needle-like spots spiked to 11.33 times / day, and the stability rate dropped to 62.7%. Continuous, uniform feeding caused the powder to be added outside the low-current window. Under the high current during the main deposition period, the added powder struggled to form localized high-concentration complexes, failing to trigger the fracture and collapse of the interfacial rheological structure. A large amount of undissolved powder accumulated in the working area, further worsening mass transfer conditions and causing blockage of the exhaust channels. The high-frequency spikes in the needle-like pattern and the precipitous drop in the histogram quantitatively demonstrate the catastrophic damage to macroscopic electrical stability caused by time-domain rheological runaway.

[0180] The data from Comparative Example 1 and Comparative Example 2 also reflect the limitations of a single-parameter environment. Comparative Example 1 uses a global high temperature, and the exhaust is relatively smooth, but due to the interference of impurities caused by the continuous addition of waste residue, the stability rate drops to 88.5%. Comparative Example 2 is limited by the shrinkage of the molten pool and the disturbance of the thermal field caused by the bottom layer, and the stability rate is only 81.2%.

[0181] Quantitative statistical data and graphical logic rigorously demonstrate the engineering necessity of spatiotemporal dual-dimensional synergistic control. The cooling and viscosity-enhancing process inevitably introduces rheological side effects that limit gas-phase mass transfer. Therefore, the spatial thermal field stratification constructed in this scheme must be deeply integrated with the decoupling of the electric field-feeding sequence in the temporal domain. This bidirectional synergistic intervention alters the microscopic coordination network and hydrodynamic boundary conditions of the anode interface, fundamentally eliminating the potential for electrolytic electrical instability caused by high-viscosity purification.

[0182] Test Example 6:

[0183] This test case aims to quantitatively evaluate the actual material savings of this process in continuous production. To ensure the rigor of the accounting baseline, a set of traditional pure virgin material processes without adding any waste materials was established as a baseline. The specific operation and accounting steps are as follows:

[0184] A baseline of traditional pure virgin materials was established. In a 10,000-ampere-level electrolytic cell of the same specifications as in the example, fresh industrial-grade praseodymium-neodymium oxide powder and fresh industrial-grade pure lithium fluoride were used (without introducing any dust-recovered materials or furnace slag), and the cells were continuously and stably operated for 72 hours under global constant temperature and constant DC conditions of 1050°C.

[0185] Collect and accurately weigh all qualified praseodymium-neodymium metal ingots produced by the baseline process. Perform a material balance calculation on all fresh oxides and fresh lithium fluoride consumed during this operating cycle to obtain the theoretical fresh material consumption per ton of metal produced under the baseline process, which will serve as a benchmark for subsequent rate of decline calculations.

[0186] During the 72-hour continuous operation of Examples 1 to 3, the total mass of the customized pretreatment materials (component A and component B) added to each group was strictly measured. At the same time, based on the liquid level sensor of the electrolyzer and the XRD composition analysis results of the molten salt sampled at regular intervals, fresh praseodymium oxide powder and fresh lithium fluoride were dynamically replenished to maintain the liquid level and component balance of the system, and the total weight of fresh material replenishment was recorded for each example within 72 hours.

[0187] Praseodymium and neodymium metals meeting Grade 1 purity standards, produced via vacuum siphoning in Examples 1 to 3, were collected within 72 hours. The total weight of the recorded fresh material replenishment was divided by the total weight of the actual qualified output for each group to obtain the actual fresh material consumption per ton of high-purity metal produced. This was compared with baseline data to calculate the percentage decrease in the consumption of each type of new material.

[0188] Table 6. Calculation data on fresh material consumption and decline rate per ton of praseodymium-neodymium metal produced.

[0189]

[0190] Note: "-" in the table indicates that it is not applicable. Because the "Traditional Pure New Material Baseline" group serves as the zero point of comparison for calculating the rate of decline, its own relative rate of decline has no mathematical meaning, and therefore no numerical calibration has been performed.

[0191] Test conclusion:

[0192] Based on the data in Table 6 and the corresponding appendix Figure 7 As shown, compared to the traditional benchmark process that relies entirely on virgin materials, Examples 1 to 3 exhibited a significant decrease in the actual consumption of fresh praseodymium oxide and fresh lithium fluoride per ton of commercially viable high-purity praseodymium-neodymium metal. (See Appendix) Figure 7 The graph quantifies the absolute consumption of fresh materials in each group and the savings rate relative to the baseline. The horizontal axis lists the traditional pure virgin material baseline and Examples 1 to 3, respectively, used in the calculation. The left vertical axis corresponds to a clustered bar chart, representing the absolute unit consumption weight (kg / t). Dark gray bars represent the unit consumption of fresh praseodymium-neodymium oxide; light gray bars represent the unit consumption of fresh lithium fluoride. To bridge the difference in values ​​and provide a clear comparison, the height of the light gray bars is magnified 50 times, but the labels at the top of the bars represent the actual measured consumption values. The right vertical axis corresponds to two marked line segments, representing the unit consumption reduction rate (%) relative to the baseline. The black solid line marked with a dark gray square represents the oxide reduction rate, and the black dashed line marked with a light gray diamond represents the lithium fluoride reduction rate. Data labels at each point are positioned in opposite directions to prevent visual overlap.

[0193] Observing the stepwise changes in the bar chart, it can be seen that compared to the towering bars of the pure virgin material baseline group (fresh oxide consumption 1218.6 kg / t, fresh lithium fluoride consumption 21.4 kg / t), the heights of both the dark gray and light gray bars in Examples 1 to 3 have significantly decreased. After applying the spatiotemporal decoupling control process, the oxide consumption in Examples 1 to 3 decreased to the range of 994.2 to 1056.3 kg / t; the lithium fluoride consumption sharply decreased to the range of 12.1 to 14.8 kg / t. The trend of the line graph more intuitively depicts the leap in savings: the rate of decrease in oxide consumption steadily climbed to 13.3% to 18.4%; the rate of decrease in lithium fluoride consumption showed a significant leap, reaching as high as 30.8% to 43.4% (with Example 2 showing the best benefits).

[0194] The strong coupling between the column height decline and the zigzag rise objectively proves that the added waste residue not only did not reduce the product purity, but also successfully resolved and reintegrated the intrinsically contained fluorine, lithium, and rare earth valuable components into the electrolyte network. The reduction in oxide consumption is mainly attributed to the in-situ resolution and reduction of the waste. The escaped rare earth dust in component A and the praseodymium-neodymium fluoride enriched in component B are incorporated into the molten salt network after being added to the electrolyzer. Relying on the vertically layered microenvironment constructed by widening the electrode spacing, not only are impurity ions successfully trapped in the upper high-viscosity region, but the release of Pr is also allowed. 3+ With Nd3+ Driven by a DC electric field, ions penetrate the bottom laminar flow zone and successfully reach the cathode to complete deposition. This selective mass transfer at the kinetic level ensures that valuable rare earth elements in the waste residue are fully extracted and converted into high-purity metals, directly offsetting the need for fresh oxide powder.

[0195] The dramatic decrease in lithium fluoride consumption confirms the mechanistic advantages of this process in suppressing volatilization and flux recycling. A large amount of previously volatilized LiF condenses in component A, which is simultaneously injected as a cooling flux during the relaxation period of the trapezoidal wave, avoiding secondary volatilization of fluorides caused by high-temperature local overheating under traditional high-current continuous feeding. Simultaneously, the simultaneous feeding under a low-current window allows LiF to rapidly and smoothly participate in the formation of a RE-OF local complex network, instantaneously disrupting the anode insulating gas film and eliminating side reactions that generate large amounts of harmful fluorine-containing gases such as HF and CF4. This effectively blocks the irreversible loss of fluorine and lithium elements from the system from both thermodynamic and electrochemical perspectives.

[0196] The intuitive cost data and material accounting in the charts objectively demonstrate that this solution eliminates the need for expensive wet acid dissolution, extraction, and wastewater treatment equipment. By simply applying thermodynamic flow field intervention and electrodynamic timing matching to the existing electrolytic cell, the technical barriers to waste separation and closed-loop resource recovery within the internal system can be overcome. The precise linkage of various parameters maintains the thermal balance of the electrolytic cell and the yield of the finished product, eliminating impurities caused by waste residue at the source. This solves the pain points of high purification and processing costs of rare earth smelting waste and serious waste of fluorine and lithium resources, perfectly replacing the high-cost wet smelting process and achieving the green smelting goal of balancing metal purity and closed-loop resource utilization.

Claims

1. A green smelting process for rare earth metals, characterized in that, include: A basic electrolyte is added to an electrolytic cell equipped with a variable frequency DC rectifier and an electrode gap adjustment device. The electrolyte is energized and melted to form molten salt. The vertical insertion depth of the graphite anode in the electrolytic cell is adjusted upward by a mechanical lifting mechanism. The electrode gap is set, the cell voltage is adjusted, and the temperature of the upper layer of the molten salt is controlled to be stable. The temperature of the working area of ​​the bottom liquid cathode is controlled to be stable, and the temperature of the upper layer of the molten salt is controlled to be lower than the temperature of the working area of ​​the bottom liquid cathode. Adjust the variable frequency DC rectifier (we use a high frequency synchronous rectifier) ​​to increase the cathode current density and maintain it at the first preset current density. During the period of maintaining the first preset current density, coarse crystal clear slag powder is fed into the middle and upper layer of the electrolytic cell through the automatic feeding port of the electrolytic cell. The variable frequency DC rectifier is controlled to smoothly reduce the cathode current density to a second preset current density with a current drop slope, and then enters the relaxation period. Maintain the cathode current density at the second preset current density. During the relaxation period, start the pneumatic jet feeding system to mix praseodymium oxide neodymium powder with fluxing dust removal compound and then spray it onto the molten salt surface of the area where the graphite anode is located. After feeding is completed, the variable frequency DC rectifier raises the cathode current density back to the first preset current density with an increasing slope, thus completing a single electrolysis cycle.

2. The green smelting process for rare earth metals according to claim 1, characterized in that, The basic electrolyte is composed of 55.0%–65.0% NdF3, 15.0%–20.0% PrF3 and 15.0%–20.0% LiF by mass percentage (actually about 85% praseodymium-neodymium fluoride + about 15% lithium fluoride).

3. The green smelting process for rare earth metals according to claim 1, characterized in that, The set electrode spacing is 8.0cm to 13.0cm; The control of the upper layer temperature in the molten salt is specifically to control the upper layer temperature in the molten salt at a depth of 0-15cm from the surface of the molten salt to be stable at 960℃-980℃. The control of the bottom liquid cathode working area temperature stability specifically involves controlling the temperature of the bottom liquid cathode working area at a depth of 0-10cm from the bottom of the tank to be stable at 1000℃-1020℃.

4. The green smelting process for rare earth metals according to claim 1, characterized in that, The single electrolysis cycle is 30 to 40 minutes; The first preset current density ranges from 4.5 A / cm². 2 ~5.5A / cm 2 The continuous operating time at the first preset current density is 23 to 32 minutes; The second preset current density ranges from 0.8 A / cm². 2 ~1.2A / cm 2 The relaxation period lasts for 2 to 4 minutes.

5. The green smelting process for rare earth metals according to claim 1, characterized in that, The current decreasing slope and the current increasing slope are both 1.5 A / (cm). 2 ·min)~2.5A / (cm 2 ·min).

6. The green smelting process for rare earth metals according to claim 1, characterized in that, The particle size range of the coarse crystalline slag powder is between 100 and 800 μm; The process of adding coarse crystallized slag powder to the upper and middle layers of the electrolytic cell is specifically controlled to add 0.1 kg to 0.4 kg every 5 to 8 minutes.

7. The green smelting process for rare earth metals according to claim 6, characterized in that, The preparation steps of the coarse crystal clear slag powder include: The blocky slag cleaned up after the furnace shutdown is fed into a jaw crusher for coarse crushing, and then into a double roll crusher. After mechanical vibration screening, slag powder with a particle size range of 100 to 800 μm is retained and collected, while slag lumps with a particle size greater than 800 μm are discarded. The slag block is fed into a drying furnace and dried at a temperature of 150℃~180℃ for 1 to 2 hours. After being discharged and cooled, the coarse crystal clear slag powder is obtained.

8. The green smelting process for rare earth metals according to claim 1, characterized in that, The mixing of conventional industrial praseodymium-neodymium oxide powder with fluxing dust removal compound is carried out at a mass ratio of 3:1 to 5:

1. The molten salt surface surrounding the graphite anode is sprayed at a feed rate of 0.5 kg / min to 1.5 kg / min.

9. The green smelting process for rare earth metals according to claim 8, characterized in that, The preparation steps of the fluxing dust removal compound include: The raw dust collected from the bag filter in the electrolysis workshop is placed in a vacuum rotary kiln and dehydrated for 2 to 4 hours under negative pressure conditions of 200℃~250℃ and 0.05MPa~0.08MPa. The dehydrated material is then fed into a homogenization silo, and industrial anhydrous LiF particles are added as a flowability regulator at 2.0% to 5.0% of the original dust removal ash mass. The forced mixer is started and mixed for 30 to 60 minutes. After discharge, the material is sealed and stored to obtain the fluxing dust removal compound.

10. The green smelting process for rare earth metals according to claim 1, characterized in that, During the continuous single electrolysis cycle, the liquid praseodymium-neodymium metal at the bottom of the electrolytic cell is periodically extracted into ingots through a vacuum siphon. The inert waste crust formed by the accumulation of impurities in the coarse crystal clear slag powder that have not been electrolytically reduced on the side wall and non-working area of ​​the electrolytic cell is concentrated and removed during the furnace shutdown and overhaul.

Images