Selective recovery of vanadium pentoxide from spent catalysts of sulfuric acid production: Sustainable approach

Spent vanadium catalysts of sulfuric acid production (main elemental composition in wt%: 7.5 V, 9.1 K, 10.2 S, 23.2 Si and 1.4 Fe) can be used as a secondary source of vanadium. Extraction of vanadium was studied using two-step leaching (acidic and reductive) of spent vanadium catalysts with further oxidizing of leaching solutions. The factors leaching and hydrolysis temperature, concentration of leaching (H2SO4, Na2SO3) and oxidizing ((NH4)2S2O8) reagents, solid/liquid ratio, mixing parameters, and time of leaching and thermohydrolysis were systematically investigated. The solubility of V2O5 was investigated as a function of temperature, pH of sulfuric acid solutions, and concentration of Na2SO3. The kinetics of V2O5 solubility and reduction were also studied. The vanadium leaching yield after a two-step recovery was 98 wt% after acidic (H2SO4, pH 1.2–1.3) leaching with ultrasonic treatment for 5 min at ambient temperature, followed by reductive leaching in 0.01 Mol/L Na2SO3 solution for 15 min at ambient temperature. The highest vanadium extraction yield from leaching solutions was 98 wt% obtained through oxidizing of leaching solutions by 30 wt%. (NH4)2S2O8 with a molar ratio n(V2O5)/n ((NH4)2S2O8) of 5/1 for a reaction time of 5 min at 80–90 C. the extracted vanadium product was V2O5 with a purity of 85–87 wt%. The technological scheme has been developed to recycle all obtained products and sub-


Introduction
Vanadium is a valuable metal widely dispersed in the Earth's crust (the 22nd most abundant element). It is found in over 50 different minerals (Habashi 1998;Habashi 2002;Moskalyk and Alfantazi 2003), however, it is never found in its pure state (Perron 2001). Nowadays, vanadium is recovered as a by-product or a co-product from mineral resources where its content ranges from 0.01-0.2 wt% to 2 wt% (Gupta and Krishnamurthy 1992;Nikiforova et al. 2017). The current primary resources are not sufficient to satisfy the vanadium demand (Liu and Yu, 2003) because of the increasing industrial interest in vanadium applications.
The steel industry is the largest consumer of vanadium. Today, its consumption in metallurgy is up to 85% of total vanadium consumption; vanadium is used as an alloying component introduced into steel as ferro-vanadium (Erust et al. 2016;Nikiforova et al. 2016). Furthermore, vanadium is widely used for the producing of vanadium redox flow batteries (VRBs) (Wang et al. 2011;Skyllas-Kazacos et al. 2011;Cheng et al. 2011). VRBs have unique advantages for large-scale application and a long cycle life in comparison with other energy storage technologies (Zhang et al. 2013;Aaron et al. 2013). Vanadium is also widely used within the chemical and petrochemical industry, where it is employed for the production of a variety of vanadium compounds, used for instance to prepare catalysts (Khorfan et al. 2001). Vanadates are also widely used as a less hazardous substitute of chromium compounds in the production of yellow-colored pigments (Zharskiy et al. 2015) and as a corrosion inhibitor for aluminum alloys (Kharitonov et al. 2017(Kharitonov et al. , 2018(Kharitonov et al. , 2019. Growing industrial interest to vanadium requires to other ecological and economic sources to produce vanadium and its compounds. One source is vanadium-containing waste: spent vanadium catalyst (SVC), fly ash, converter and smelter slag Zharski et al. 2012). SVC is one of the most preferable secondary raw materials for vanadium extraction, as it contains 5-10 wt% of vanadium along with other valuable components, such as Cu, Ni, Mo, and Co, in the form of oxides or sulfates (Akcil et al. 2015;Erust et al. 2016). The average service life of this catalyst is about 2-5 years (up to 10 years) (Ullmann 's 1994). During production of sulfuric acid, about forty thousand tons of SVC are produced annually worldwide (Mauskar, 2007), which would cause several pollution problems if wasted. There are several advantages with the recycling of vanadium-containing waste. Costs of vanadium extraction from these raw materials are lower compared with natural deposits. Also, ecological arguments become increasingly important (Tadao et al. 2001;Rajendran et al. 2016;Liu et al. 2019). The United States Environmental Protection Agency classifies spent catalysts as hazardous wastes (USEPA, 2003). Vanadium is considered the major source of contamination (WHO, 2011;Navarro et al. 2007;Montiel-Davalos et al., 2012).
A number of hydro-and pyro-metallurgical processes have been proposed for metal recovery from secondary raw materials. The metals are recovered as mixed solutions and then separated through conventional separation techniques (solvent extraction, selective precipitation and ion-exchange) (Ognyanova et al. 2009).
Alkaline leaching is selective for vanadium (solubility of vanadium in alkaline media is more than 10 times higher than in acidic media) over iron but dissolves some silica and is more expensive in terms of reagents (Ho et al. 1994). Khorfan et al. (2001) described a three-step process of SVC utilization involving acidic leaching, oxidation and precipitation to recover vanadium pentoxide but the efficiency of this method was rather low -70 wt%.
It is obvious that the method of SVC utilization must be highly efficient from technological, economic, and ecological points of view. Because of that, our approach was to develop a hydrometallurgical method of SVC utilization using reagents with a composition close to the composition of SVC to prevent contamination. Moreover, reagents were used that are the main product of plants producing SVC, such as sulfuric acid, sulfites and persulfates. This study aimed to generate a technological scheme that could be used directly on plant. This work focused on selective metal extraction, and aimed to understand different extraction/utilization parameters of leaching and oxidation to recover vanadium in an efficient way and with high purity of obtained V 2 O 5 .

Materials and sample preparation
Spent vanadium catalyst sulfovanadate on silica gel type, used in this study, was provided by JSC Grodno Azot, Belarus. The catalyst samples (cylindrical, with an average diameter of 6 mm and 20 mm in length) were ground in the planetary ball mill PULVERISETTE 6 (FRITSCH, Germany) with a crushing time of 30 s to improve leaching efficiencies. Then, they were dried at 100 • C for 1 h.
Sulfuric acid was used as a leaching agent and it was prepared by diluting concentrated acid (Sigma-Aldrich, 98%) with distilled water to the desired concentration or pH. All acids, sodium sulfite, ammonium persulfate, and vanadium pentoxide were of analytical grade (Merck).

Analysis of materials
Morphology and elemental composition of the sample surface were investigated by Scanning Electron Microscopic (SEM) analysis using a Scanning Electron Microscope JSM 5610 LV equipped with the Energydispersive X-ray spectroscopy system EDX JED 2201 JEOL (Japan). At least three samples were investigated for each condition. The surface analyses were conducted with magnifications up to 5000× for a minimum of 10 different locations.
The phase compositions of the initial SVC samples and the synthesized V 2 O 5 and solid leaching residues were determined by means of Xray diffraction analysis with a Panalytical X'PERT PRO diffractometer (Netherlands) [wavelength Cu Kα (1.5405 A) and software Philips X'PERT suite]. The PDF2 database was used as the reference data. The software HighScore Plus was used for Rietveld refinements. The peak profile was refined by pseudo-Voigt function.
A Confotec MR350 instrument was used for confocal Raman spectroscopy measurements, using a 532 nm laser (no filter). The sample was inspected before and after analysis to ensure no laser-induced oxidation.
Thermogravimetric analysis (TGA) of SVC and obtained V 2 O 5 was done with a TA Q500 instrument. 40 mg of the sample was placed into an alumina pan without a lid and heated from ambient temperature to 1000 • C (SVC) and to 600 • C (V 2 O 5 ) at 5 • C/min under a nitrogen purge flow of 100 mL/min.
Particle size distribution was determined by Fritsch Particle Sizer Analysette 22 (Germany) in distilled water suspension.

Solubility and kinetics
The solubility testing of V 2 O 5 and SVC was performed in a glass flask with a magnetic stirrer and heater (Velp, Arec). The solubility and kinetics of SVC were also studied in an ultrasonic (US) installation with piezoelectric emitter IL 100-6/1 (Inlab, Russia), an installation power of 630 W, an operating frequency of 22 ± 10% kHz, a vibration amplitude of at least 40 μm, and a 50 mL volume of the processed suspension. The concentration of the aqueous suspension of SVC during processing in the ultrasonic installation was up to 20 wt%. The pH was continuously monitored by a controller connected to a computer. The final solution was filtered using an ashless syringe filter paper and analyzed for weight loss using a mass balance by weighing the remaining solid after each test and calculating the extent of leaching based on the initial weight. The composition and quantity of extracted vanadium compounds and solid residues after the leaching were determined by SEM and XRD methods. The triplicate experiments for every point have been done.
The effect of pH, temperature, sodium sulfite and ammonia persulfate (peroxydisulfate) concentration on kinetics was studied. Gravimetric studies, Atomic Absorption Spectroscopy (AAS), Electronic Paramagnetic Resonance (EPR), and UV-Visible Spectroscopy (UV-Vis) were applied to analyze vanadium (IV/V) concentration in solution (Kanamori et al. 1999;Yang and Gould 2003). For the study of reduction kinetics, sulfuric acid solution of V 2 O 5 with adding of 0.01 mol/L Na 2 SO 3 was used. The рН of the reduction solution was changing with time (0-1500 s) from 1.30 to 1.13.
The content of vanadium (V) after reduction by Na 2 SO 3 was determined by AAS analysis using a spectrometer Avanta GBC Scientific Equipment (Australia).
EPR spectra were recorded at 1000 G and 298 K in X-band range (operating frequency 9.3 GHz) in a magnetic field up to 7 T, with a power of 5 mW, by means of a Varian E112 spectrometer. The RF Modulation Amplitude (100 kHz) was 1.0 G. Intense signals were attenuated 8-256 times. EPR was used to determine the vanadium oxidation state in the reduction solutions of V 2 O 5 and SVC, because it is known that the vanadyl ion VO 2+ shows an EPR spectruman octet with a constant on vanadium of the order of 10.8 mT (Hanson and Berliner 2009;Krzystek et al. 2015).
UV-Vis absorption spectra of V 2 O 5 in reduction solutions were recorded at λ = 277.6 nm (maximum light absorption by sulfuric acid solution of V 2 O 5 ) (Santini et al., 1952;Samchuk and Pilipenko, 1987) and 298 K by means of a SPECORD 200 PLUS (Analytik Jena AG, Gemany). The initial concentration of V 2 O 5 in sulfuric acid solutions was 6.3•10 − 5 mol/L, the рН was 1.5, the measuring range was 190-1100 nm, the wavelength step was 0.1 nm, the speed was 5 nm/s, the integration time was 0.02 s, and the points of data fixation was 0.2 s. The molar coefficient of light absorption ε (L / (mol ⋅ cm)) was calculated as: where А -absorption or optical density; С -concentration of V 2 O 5 in solution; lsolution layer thickness.
The substance conversion rate was determined as: where ΔC -concentration difference of V 2 O 5 in solution before and after reaction; treaction time.

Synthesis
It was hypothesized that iron could be present in SVC not only as a part of its phase composition but also as solid iron particles from contact absorber shelves. To remove solid iron particles, SVC samples were handled with magnet before leaching and SVCm was obtained.

Leaching tests
The acidic and reductive leaching tests were performed in glass shake flasks (100 mL) with a magnetic stirrer at ambient temperature. The acidic leaching tests were also performed in the ultrasonic installation. The S/L ratio and the concentration of sulfuric acid and sodium sulfite were selected according to optimal experimental conditions. The S/L ratio was 10 ± 0.0005 g of SVCm in 50 mL of H 2 SO 4 (S/L = 1/5) and the pH of the solution was 1.2-1.3. The contact time was 5 min, after which the samples were filtered. Magnani et al. (2000) used sodium thiosulfate to reduce vanadium (V) to vanadium (IV) and reported that it could be dissolved in the form of oxy-sulfate (VOSO 4 ) by means of sulfuric acid leaching of spent sulfuric acid catalyst. In the present work, the solid residue from acidic leaching was used with the addition of the reducing agent sodium sulfite (Na 2 SO 3 ) at 0.01 mol/L to reduce vanadium (V) to vanadium (IV), the S/L ratio was 1/5, the solution pH was 2.6-2.8, and the reaction time was 15 min. The solid phase was filtered and dried. Filtrates from acidic and reductive leaching were combined and were used for extraction of vanadium.

Extraction tests
It is known that vanadium in acidic leach liquor can be presented in the form of VO 2 + as (VO 2 ) 2 SO 4 and in the form of VO 2+ as VOSO 4 (Eatough et al. 1984). To extract vanadium from leaching solutions in form of V 2 O 5 , filtrates were oxidized with 1.0 mL of ammonium peroxydisulfate ((NH 4 ) 2 S 2 O 8 ) at 30 wt% in solution. Oxidized solutions were processed by thermohydrolysis: they were heated to the boiling point, the time of boiling was 5 min, and the precipitated V 2 O 5 was filtered and dried. The filtrate after the vanadium extraction was recycled and included in the acidic leaching step. The filtrate was neutralized periodically with ammonia solution (25 wt%) to obtain Al(OH) 3 . NH 3 •H 2 O was added in an optimal amount to increase the pH from approximately 3 (pH of leaching solution after vanadium extraction) to 5-6 (pH of Al(OH) 3 precipitation). The precipitated Al(OH) 3 was separated by filtration and was dried. The filtrate after Al(OH) 3 separation was vaporized to reach concentration of 20 g/L K + .

Compositional analysis
XRD patterns and SEM images of SVC are presented in Supplementary Information (SI, Figs. S.1a-b). The main phases of SVC were cristobalite and quartz. Apart from the support phase, crystalline phases of potassium sulfates, disulfates, pyro-sulfato-vanadates, and pyrosulfates were identified, according to (Nikiforova et al. 2016). The elemental composition of the initial SVC sample is listed in SI, Table S.1. Vanadium was present in SVC in the following compounds: 41.6 wt% -anorthic crystal system V 10 O 18 ; 34.1 wt% -monoclinic crystal system V 4 O 10 ; 14.7 wt% -anorthic crystal system V 20 O 36 ; 8.6 wt% -other phases. DTA and TG analysis of SVC showed the number of SVC decomposition stages with total mass loss 21.1 wt% and endothermic thermal effects corresponded to the removal of physically bound water (Figueiredo et al. 1999) and sulfate decomposition (SI, Fig. S.2).
The amount of iron in the SVCm as compared to the SVC sample (SI, Fig. S.1c) after magnetic separation decreased from 1.41 to 0.85 wt% (SI , Table S.1). XRD patterns (SI, Fig. S.1a) show the same phase composition for the SVCm sample as in the initial SVC sample but with a small left shift of peaks because of the smaller amount of magnetic phases (less than 5%). The median particle size of the SVCm sample after magnetic separation decreased by 2.3 times in comparison with the initial SVC sample from 17.867 μm (SI, Fig. S.3) to 7.764 μm (SI, Fig. S.4). This supports our hypothesis on the presence of iron both in the phase composition of the catalyst and as solid iron particles from contact absorber shelves in the SVC. Such solid iron particles can be easily recycled after magnetic separation from SVC.

Solubility and kinetics analysis
The solubility and kinetics study were done to obtain optimal extraction/utilization parameters of SVC leaching. The V 2 O 5 solubility in water was studied as a function of temperature. As Fig. 1a shows, the content of V 2 O 5 in the saturated solution (pH = 2.8) is 0.2 g/L at a temperature of 20 • С. With an increase in temperature from 20 to 90 • С, the solubility of V 2 O 5 in solution was 4-fold higher and reached 0.81 g/ L. A further increase in temperature led to boiling of solution with an increase in hydrolysis processes and a decrease in the V 2 O 5 content in solution. Obtained data corresponded to (Woolery 1997;O'Neil, 2001) showing an increase in V 2 O 5 solubility from 0.30-0.35 g/L at 25 • С to 0.7 g/L at 100 • С (pH = 2.8). Upon dissolution in water, V 2 O 5 forms crystalline hydrates with three, two or one molecule of water, which correspond to the forms of ortho-, pyro-and methavanadic acid (Jahr et al. 1963). The process of electrolytic dissociation of methavanadic acid is equal for both directions (Jahr et al. 1963): It was reported (Ivakin 1966) that in acidic solutions (рН < 1.5) methavanadic acid could be mostly present in the form of VО 2 + or in the form of [H n V 10 O 28 ] (6-n)-. The ion VО 2 + is characterized by its high oxidation properties. The content of [H n V 10 O 28 ] (6-n)is increasing with increasing pH. The solubility of V 2 O 5 in sulfuric acid solutions is presented in Fig. 1b. The experimental data shows a significant increase for V 2 O 5 solubility in sulfuric acid solutions. The most probable forms of vanadium (V) in such solutions are (Pletnev et al. 1986): [Н 2 V 10 O 28 ] 4− = [HV 10 O 28 ] 5− + H + , K = 10 -3.6 (5) [HV 10 O 28 ] 5− = [V 10 O 28 ] 6− + Н + .K = 10 -6.8 In a non-complexing aqueous acidic solution, the pentavanadyl ion (VO 2 + ) is the only dominant cation in the solution. It was found that the solubility of V 2 O 5 reaches its maximum point (5.8 g/L) at рН < 1 (Fig. 1b). This fact could be explained by the formation of a monosulfate complex (VO 2 HSO 4 ) according to the following reactions (Ivakin 1966;Rakid and Durand, 1996;Puigdomenech 2004): An increase of pH in sulfuric acid media led to a significant decreasing of V 2 O 5 solubility to less than 0.3 g/L at рН 5.5-7.0. In this solution, the most probable form of vanadium (V) is [Н n V 10 O 28 ] (6-n)- (Ivakin 1966). Hence, our study suggests that the most optimal pH of sulfuric acid solution for the acidic leaching of vanadium (V) from SVC is рН 1.2-1.3 in terms of the ratio of V 2 O 5 solubility and quantity of sulfuric acid.
Since vanadium (IV) has a higher solubility than vanadium (V) in acidic media (Pourbaix 1974;Puigdomenech, 2004;Ivankovic et al. 2006), the solubility of V 2 O 5 in reducing sodium sulfite solutions was studied (Fig. 1c). The addition of sodium sulfite in aqueous solutions of V 2 O 5 at a quantity of 0.01-1.0 mol/L led to a change of the solution color from yellow to dark green, which can be explained by the formation of blue vanadium (IV) compounds or green vanadium (III) compounds according to: As shown in Fig. 1c, the solubility of V 2 O 5 in acidic solutions with Na 2 SO 3 contents ranging from 0.01 to 0.20 mol/L (constant рH = 2.8) increased from 1.05 to 2.07 g/L. The solubility of V 2 O 5 in sodium sulfite solutions is hence higher than in water and in sulfuric acid solutions because of the formation of vanadium (IV) and vanadium ( (Pletnev et al. 1986). However, the increase of the Na 2 SO 3 concentration to more than 0.2 mol/L led to a decrease of the solubility, Fig. 1c. This might be explained by the formation of sparingly soluble sulfates and sulfites of vanadium. In order to prevent reagents overconsumption and based on obtained data (Fig. 1c), it was suggested to use 0.01 mol/L solution of sodium sulfite for the reductive leaching of vanadium (V) from SVC.
The kinetics of V 2 O 5 reduction in sodium sulfite solutions are shown in Fig. 2a. The highest degree of V 2 O 5 reduction (92 wt%) was reached after the first 5 min of reaction time. During the next 40 min, the content of reduced V 2 O 5 in the solution was reduced to 61 wt%. After 90 min of the reduction process, the quantity of reduced V 2 O 5 reached 78 wt% and did not change further.
As the obtained data showed, there was a reversible reaction for the reduction of V 2 O 5 by sodium sulfite in sulfuric acid solutions. During the first 5 min, the maximum of the reduction of V 2 O 5 was detected and the color of the solution was changed from yellow (vanadium (V) compounds) to blue (vanadium (IV) compounds). Then, formed vanadium (IV) compounds were oxidized to vanadium (V) compounds and the solution color changed to light green (a mixture of vanadium (IV/V) compounds). Such fact when reduction is caused by an oxidizing agent and, conversely, oxidation is caused by a reducing agent is mentioned by Remy (1973).
The EPR analysis of reductive solutions showed the spectrum of VO 2+ in sulfuric acid, proving the existance of vanadium (IV) in solution in the form of vanadyl sulfate VOSO 4 . Fig. 2b shows the intensity of the 4th peak of the spectrum EPR (I 4 ) of VO 2+ in acidic sodium sulfite solutions as a function of reaction time. In agreement with the AAS and gravimetric results, a maximum of reduction occurred after 5 min of the process, a minimum of reduction after 45 min, and a plateau was reached after 90 min.
The rate of V 2 O 5 reduction by sodium sulfite was calculated for the first 5 min of reaction time from Fig. 2c (detailed calculation is presented in SI). An average molar absorption coefficient was calculated by (1) ε = 2535 L/(mol•cm). An average conversion rate of the was calculated by (2) v = 1.97•10 − 7 mol/(L•min).

Leaching of vanadium
As Fig. 3a shows, the S/L ratio had a significant effect of SVCm mass loss in the interval from 1/2 to 1/5. A further increase of water volume increases the SVCm mass loss slightly: in the interval of S/L ratio from 1/ 5 to 1/50 it increases by 10% and from 1/200 to 1/1000 it remains almost constant -54.5 ± 0.2%. Obtained results of the dependence of SVCm mass loss on the S/L ratio assisted in choosing the minimum solvent volume, at which the most rational ratio "water consumptionvanadium extraction" was achieved.
The increase of temperature did not influence the SVCm mass lost and its value varied from 32 wt% to 45 wt% at any temperature of the interval from 20 to 90 • C (Fig. 3b).
The results of the kinetics studies of SVCm mass loss as a function of different types of mixing are presented in Fig. 3c,d. The data showed that the first 5 min of mixing of the SVCm water solution with a magnetic stirrer led to 32 wt% of SVCm mass loss and after 60 min the maximum of SVCm mass loss (37 wt%) was reached (Fig. 3c), after which the leaching was constant. US treatment increased the mass loss of SVCm significantly as compared to the magnetic stirrer and it reached the maximum value after the first 5 min of US mixing -51-52 wt% (Fig. 3d). Such SVCm mass loss is comparable with a S/L ratio of 1/100-1/1000 (Fig. 3a). Moreover, it led to the increase of vanadium leaching from SVCm to 60%. This effect is explained by the formation of particles with smaller sizes (9.466 μm, SI,Fig. S.5) in comparison with the size of particles formed in the solutions with magnetic stirrer mixing (12.455 μm, SI, Fig. S.6). The further decrease of SVCm solubility under US treatment can be explained by the formation of particle agglomerates with larger sizes, which correlates to our previous experimental data (Zharskiy et al. 2015) and literature data (Khan et al. 2013).
Obtained results showed that the S/L ratio 1/5 and 5 min of US treatment at ambient temperature led to salvation of all soluble SVCm components. The high solubility of SVCm in water (in comparison with V 2 O 5 ) is explained by the presence of highly soluble compounds such as potassium and sodium sulfates and polysulfates in the SVCm composition (SI, Fig. S.1). Furthermore, the presence of disulfates in the SVCm leads to sulfuric acid formation in the water solution of SVCm. The solution pH becomes 1.5-1.6, which facilitates the solvation of both vanadyl sulfate VOSO 4 and V 2 O 5 in the solution. The main component of the SVCm solid residue after leaching was insoluble SiO 2 in sulfuric acid solution.
As was shown before, the presence of a reducing agent in water solutions led to the formation of vanadium compounds in low oxidation states that are significantly more soluble than V 2 O 5 (Figs. 1c and 2). Fig. 4 shows that the mass loss of SVCm and the degree of its vanadium leaching depends on the presence of sodium sulfite in water. The concentration of Na 2 SO 3 did not significantly influence the SVCm mass loss (44-46 wt%) and the vanadium leaching degree (33-35 wt%). However, an increase of Na 2 SO 3 in the solution led to their slight decrease due to an induced pH increase. This study proved that using 0.01 mol/L Na 2 SO 3 for the technological process of reductive leaching of vanadium (V) from SVC is possible and advantageous. Fig. 4b shows the same trend of the vanadium reduction degree as shown for V 2 O 5 reduction (Fig. 2). The maximum of vanadium (V) reduction was after 5-15 min of the process, the minimum of vanadium (V) reduction was after 45-90 min of the process, and a plateau was reached after 180 min.
The influence of pH on the SVCm mass loss and degree of vanadium leaching was investigated in highly acidic solutions with pH values lower than 1.56 (pH of SVCm water solution). Fig. 5 shows the results for different steps of the leaching with developed optimal conditions as a function of the pH of the sulfuric acid leaching solutions. Represented data prove the increase of the SVCm mass loss in strong acidic solutions. It reaches maximum values at a pH lower than 1.3 for all samples. At the first step of SVCm leaching in sulfuric acid solutions, mixing with a magnetic stirrer, the mass loss increased to 47-48 wt% (SVCm-1 L-MX). The use of US treatment for mixing the sulfuric acid leaching solutions of SVCm raised the mass loss to 54-56 wt% (SVCm-1 L-US). The second step of SVCm leaching, reductive leaching in 0.01 mol/L solutions of sodium sulfite, led to the increase of SVCm mass loss for solutions with both types of mixing: to 50-51 wt% for SVCm-2 L-MX and to 58-60 wt% for SVCm-2 L-US. The increase of mass loss led to the increase of vanadium leaching from SVCm, however, at a pH lower than 0.7, it was impossible to extract V 2 O 5 from the leaching solutions. The composition of solid residues at every step of leaching is presented in Fig. 6 and Table S.2 (SI). The difference in particle sizes was found only for the first step of leaching with different types of mixing (SI, Fig. S.5 and Fig. S.6) and, after the reductive leaching, the particle sizes were comparable for both samples: 8.484 μm for SVCm-2 L-MX (SI, Fig. S.7) and 8.108 μm for SVCm-2 L-US (SI, Fig. S.8).
Thus, this study showed that the optimal condition for vanadium recovery from SVCm was a two-step leaching process. The first step was leaching in sulfuric acid solutions with a constant рН interval from 1.2 to 1.3 and 5 min of ultrasonic treatment (22 ± 10% kHz). It led to the leaching of more than 90 wt% of vanadium of its initial content in SVCm (Fig. 6b, SVCm-1 L-US). The second step was reductive leaching in 0.01 mol/L sodium sulfite with a reaction time of 15 min, leading to the leaching of more than 98% of vanadium of its initial content in SVCm (Fig. 5b, SVCm-2 L-US). The residual amount of vanadium in solid residues after the two-step leaching did not exceed 0.16 wt% of its initial content in SVCm. As the solid residues were mainly SiO 2 with small impurities of sulfur, potassium and iron (Fig. 6a, SI, Table S.2), it could be used in different industrial processes. Our study highlights the possibility to use obtained solid residues in the production of building materials, such as ceramic bricks (Kryshilovich et al. 2017).

Extraction of V 2 O 5
It is well known that vanadium can be extracted from solution by hydrolysis as V 2 O 5 (Muzgin and Khamzina, 1981). Some authors (Vegliծ et al. 2006;Ognyanova et al. 2009;Erust et al. 2016) reported about using hydrogen peroxide (H 2 O 2 ) as an oxidizing agent for vanadium leaching, leading to an intensification of the V 2 O 5 hydrolytic extraction. Our previous studies showed the possibility to use H 2 O 2 for oxidation of SVC leaching solutions and further extraction of V 2 O 5 from solutions by thermohydrolysis with boiling time 5 min (Orehova et al. 2013). The vanadium extraction degree was about 92 wt% and the molar ratio of extracted V 2 O 5 to used H 2 O 2 was 1 to 2 (Orehova et al. 2013). To increase the amount of extracted V 2 O 5 and to decrease the oxidizing agent consumption, ammonium peroxydisulfate ((NH 4 ) 2 S 2 O 8 ) was used for SVC leaching solutions oxidation. Ammonium peroxydisulfate has the same O-O bond as H 2 O 2 , why it was decided to use (NH 4 ) 2 S 2 O 8 as an oxidizing agent. Furthermore, ammonium peroxydisulfate is used in water and wastewater treatment, as it has high oxidizing properties, is nontoxic, and a cheap reagent (Wacławek et al. 2017). Fig. 7 shows that using (NH 4 ) 2 S 2 O 8 helped to extract 98 wt% of vanadium from the   leaching solutions. The optimal molar ratio of extracted V 2 O 5 to used (NH 4 ) 2 S 2 O 8 was 5 to 1 and the maximum of vanadium extraction was reached after 5 min of thermohydrolysis at 80-90 • С. The residual concentration of vanadium in terms of V 2 O 5 was not less than 0.0033 mol/L and did not depend on the initial vanadium concentration or quantity of the oxidizing agent.
The elemental composition and morphology of extracted vanadium compounds is presented in Fig. 8 and Table S.2 (SI). As shown, for both methods of leaching, the obtained product was V 2 O 5 with some impurities of potassium, silica, iron and sulfur. XRD patterns and SEM images of the obtained product show that after the thermohydrolysis the amorphous hydrated V 2 O 5 •nH 2 O formed with a small contamination of VO 2 . The average size of the particles of V 2 O 5 extracted from SVCm leaching solutions prepared with magnetic stirrer mixing was 84.323 μm (SI, Fig. S.9). The average size of particles for V 2 O 5 extracted from SVCm leaching solutions prepared with ultrasonic treatment was almost in 7 times lower -12.157 μm (SI, Fig. S.10).
DTA/TG analysis of extracted V 2 O 5 shows two decomposition stages with total mass loss 11 wt% and thermal effects corresponded to the removal of physically and crystalline bound water (SI, Fig. S.11).
V 2 O 5 •nH 2 O was heated to 500 • С for 1 h. After the heating, V 2 O 5 with a purity of 85-87 wt% was obtained that is higher in comparison with other known technologies (Khorfan et al. 2001;Peng 2019). XRD analysis (Fig. 8d) showed the presence of V 2 O 5 phase with a small impurity of nonstoichiometric vanadate K 0.23 V 2 O 5 in both samples.
The crystalline quality of obtained V 2 O 5 was also evaluated using Raman spectroscopy. Fig. S.12 (SI) shows the Raman bands corresponding to the characteristic phase of V 2 O 5 samples (Shvets et al. 2019;Sundeep et al. 2019;Wachs 2013) (see the full peak explanation in SI). Fig. 9 shows the developed optimal technological scheme of V 2 O 5 recovery from spent vanadium catalyst of sulfuric acid production. The mass balance of V 2 O 5 recovery from SVC of sulfuric acid production is presented in SI, Fig. S.13.
With time, K + , Na + and Al 3+ ions accumulate in the leaching solutions. These ions are the raw material for the production of fertilizers and Al(OH) 3 for coagulant production. In order to obtain these products, the stage of periodical neutralization of leaching solutions by NH 3 • H 2 O was included in the technological scheme.

Conclusion
As a result, a new resource and energy-saving, low-cost and environmentally friendly method of vanadium recovery from SVC of sulfuric acid production has been developed. The method includes grinding; magnetic separation of iron; two-step leaching of vanadium at S/L ratio 1/5: the first stepvanadium leaching in sulfuric acid solutions with pH = 1.2-1.3, ultrasonic treatment for 5 min; the second stepvanadium reductive leaching in 0.01 mol/L Na 2 SO 3 solutions at ambient temperature for 15 min; oxidation of leaching solution by 30 wt% solution of (NH 4 ) 2 S 2 O 8 ; thermohydrolytic extraction of V 2 O 5 for 5 min at temperature 80-90 • С; periodic neutralization of leaching solutions by NH 3 •H 2 O solution; drying; recycling of leaching solutions. The degree of vanadium recovery from SVC was 98 wt%. The purity of obtained V 2 O 5 was 85-87 wt%. Using of (NH 4 ) 2 S 2 O 8 as an oxidizer helped to obtain less corrosive solutions compares with using of H 2 O 2 . All sub-products can be used in other industrial processes: solid residues (SiO 2 ) for building materials production, Al(OH) 3 for coagulant production, and concentrates of Na + and K + as fertilizers in agriculture. The technology was developed in a way to prevent any secondary waste. All spent solutions are recycled back to the process. The evaluated capital cost of the technology is about 120-125 k$ (depends on using equipment), payback period is not more than 1 year. Such a sustainable approach could be adapted for the recovery of vanadium from other wastes, such as SVCs of different types, ash residues of fuel oil, petroleum coke.
CRediT authorship contribution statement.

Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. Fig. 9. Technological scheme of vanadium recovery from spent vanadium catalyst of sulfuric acid production: 1 -mill; 2 -magnetic separator; 3 -hopper; 4ultrasonic installation with a piezoceramic emitter; 5, 11-13 -filter; 6, 7 -storage tank; 8 -reactor for leaching; 9 -reactor for thermohydrolysis; 10 -reactor for filtrate neutralization; 14 -evaporator; 15, 16 -drum dryer.