Introduction

The importance of rare (RMs) and rare earth (REMs) metals for modern technologies and the economy is widely recognised1. Today, RMs and REMs are known as critical metals. They are widely used in many spheres of industry and daily life, for example, nuclear energy, metallurgy, medicine, chemical engineering, electronics, and computer manufacturing2. Because these metals occur in multicomponent deposits, their acquisition is complex and involves many stages. Moreover, the RM and REM extraction processes are unique for each deposit3. In this context, the question of their recovery is of interest for both scientists and specialists4. There are many methods, such as precipitation and filtration, for extracting these kinds of metals. Unfortunately, they are not economically attractive. Among the available methods for removing rare and rare earth elements, solvent extraction and ion exchange are widespread2. In particular, currently operating rare metal production facilities use liquid extraction processes involving organic compounds to extract the main elements. During re-extraction, organic substances migrate into intermediate products and process solutions. This significantly complicates downstream processing in the production chain5,6 and causes environmental contamination when organic impurities enter wastewater, soil, and the atmosphere. To avoid this, it is necessary to purify rare metal production solutions to remove organic compounds.

Rhenium is among the most sought-after rare metals because of its ductile, refractory, corrosion-resistant, and oxidation-resistant properties. The main method of extracting rhenium from Kazakhstani raw materials involves conversion into technogenic products during the processing of copper and molybdenum ores and concentrates7,8,9,10. The transition and concentration of rhenium into technogenic products during the high-temperature roasting and/or smelting of copper or molybdenum concentrates proceeds through the formation of volatile rhenium oxide (VII). Depending on the technological regime, the form of rhenium in technogenic products may vary. The metal may be present in the form of perrhenate, sulfide, or lower oxide forms formed as a result of interactions with gas phase components or sublimates. In this context, methods for extracting rhenium include oxidative processes for converting it into water-soluble forms through aqueous leaching. During oxidation at low temperatures (300–350 °C), the complete removal of organic compounds is not achieved. Consequently, during water leaching, organic compounds are transferred to the rhenium-containing production solution in significant quantities7,8,9,10. To prevent organic compounds from entering the rhenium-containing solution, prolonged (4–6 h) oxidative roasting of lead cakes from copper production at temperatures up to 650 °C has been proposed11. Organic compounds adsorbed by the Purolite A170 ion exchange resin, reduce its full dynamic exchange capacity for rhenium. However, this method of reducing the content of organic compounds in the processing products of rhenium-containing raw materials is costly, energy intensive, and environmentally unfriendly.

Adsorption on new carbon adsorbents is considered the most effective method for removing organic impurities in nonferrous metal hydrometallurgy12,13. The advantages of adsorption include efficiency, selectivity, and low process costs14. However, since hydrometallurgical solutions differ in composition in each specific case, different sorbents are needed for the selective extraction of a particular metal1. In addition, limited information exists regarding the behaviour of the adsorbents obtained in12,13 in rare metal hydrometallurgy. In addition, the synthesis of these adsorbents involves the use of expensive carbon black and pyrolytic carbon obtained from natural hydrocarbon raw materials. The solution may be found in the use of inexpensive adsorbents obtained from readily available raw materials4. In recent years, the use of low-cost adsorbents for various purposes has become increasingly popular15,16. Adsorbents based on industrial and plant waste are used for wastewater treatment and metal concentration in hydrometallurgy17,18,19. Rice husk (RH) is currently considered one of the most promising renewable sources of raw materials for the production of adsorbents20. There are multiple approaches to the preparation of adsorbents from RH. Singha et al.21, Gupta et al.22, Lattuada et al.23, Rocha et al.24, Kumar et al.25, and others tested unmodified and water-washed, crushed and noncrushed RH. Aluyor et al.26, Masoumi et al.27, Akhtar et al.28, Qu et al.29, Nabieh et al.30, and others modified RH with different chemical compounds. Singh et al.31, Liu et al.32, Xu et al.33, and others have prepared biochar from RH in different ways. Meretin et al.34, Roha et al.35, and others activated RH-based biochar to produce effective adsorbents.

The main focus of research on the use of RH adsorbents is the purification of aqueous solutions from contaminants such as toxic metals16,35,36,37,38,39 and radioactive elements40,81. Research on the adsorption processes of nonferrous41 and noble42,43,44,45 metals was also conducted to concentrate metals in hydrometallurgical processes. Perhaps the smallest number of studies using RH–based adsorbents have investigated the extraction of rare and rare-earth metals46,47,48,49,50,51. However, in our opinion, the problem is not the small number of studies on the adsorption of rare metals but the lack of environmental and economic studies of proposed sorbent production methods, many of which involve the use of large quantities of expensive and aggressive chemicals52. Moreover, methods for processing adsorbents after metal adsorption remain insufficiently explored.

Several pilot production facilities have been established at our institute (the National Center on Complex Processing of Mineral Raw Materials of the Republic of Kazakhstan). These include a refractory production facility, an ammonium perrhenate production facility, a special coke production facility, and a RH processing facility. Each facility faces its own challenges. The goal of the refractory production facility is to reduce the cost of the produced products. In the ammonium perrhenate production facility, as mentioned above, we face the issue of organic impurities contaminating the production of rhenium-containing solutions. In the production of special coke, special coke fines (FSC, up to 5 mm) accumulate in the dumps, as special coke with a size of 5–25 mm is used as a reducing agent in metallurgical processes. Moreover, the processing of RH produces a biochar that is useful in the production of effective carbon adsorbents. The byproducts can be used in refractory production. In this context, the idea arose to use the challenges and waste of one production to solve problems and as sources of raw materials for others.

Previously, we obtained carbon adsorbents active towards rare, noble, and heavy metal ions from RH and special coke fines53,54. Unfortunately, their adsorption characteristics have been evaluated only in aqueous systems free of organic compounds. Since, as shown above, the synthesis of activated carbons from RH is a fairly popular scientific direction52,55,56, the literature contains information about the adsorption capacity of these carbons towards numerous organic substances15,57,58,59,60,61,62,63,64,65. However, no studies on the selective extraction of rare metal ions from solutions containing organic substances have been reported.

Given the characteristics of adsorbents produced from RH52,62,63 and FSC54, it appears unlikely that either material alone can provide selective rhenium adsorption from production solutions containing organic substances. Nevertheless, RH combined with special coke fines (waste materials whose processing is currently under consideration54,66,67,68,69 represents a promising feedstock for developing a composite adsorbent for rhenium hydrometallurgy.

In this context, the present paper focuses on the selective removal of organic impurities and rhenium from realistic organic impurity-containing production solutions using a mixture of RH, pyrolysed RH (RHp), activated pyrolysed RH (RHp−850VA), and FSC as adsorbents. The choice of these materials as objects of study was indicated by a number of factors. First, these materials have shown positive results in various adsorption systems. Second, the creation of effective adsorbents must satisfy the principles of sustainability and cost effectiveness. Finally, Kazakhstan has these resources. Moreover, RH and special coke fines are not yet processed on an industrial scale. The disposal of these materials is necessary to improve the environment in the places where they form. Therefore, this study is a continuation of research on the creation of sustainable, low-cost adsorbents using agricultural and industrial waste to solve a specific problem, namely, extracting rhenium from solutions with a high content of organic impurities.

Materials and methods

Solutions and adsorbents

The production solutions for this research were provided by specialists from the pilot ammonium perrhenate production facility of the National Center on Complex Processing of Mineral Raw Materials of the Republic of Kazakhstan RSE (Almaty). The concentrations of rhenium and organic impurities in the solutions varied depending on the composition of the initial raw material (lead production cakes), reaching values of up to 640 mg dm− 3 and 789 mg dm− 3, respectively.

RH was obtained from an agricultural enterprise in Bakanas village, Almaty region.

To produce pyrolysed RH, the RH was washed with water, dried and pyrolysed for 30 min at 450 °С in a pilot plant.

RHp characteristics: BET surface area, 69 m2 g−1; average pore diameter, 2.3 nm; total pore volume, 0.08 cm3 g−1; and iodine adsorption capacity (AI), 9%.

RHp was activated with water vapour for 30 min at 850 °С. The prepared product was boiled for 90 min with 70 g dm−3 sodium hydroxide solution at a solid (g): liquid (cm3−) (S: L) ratio of 1:10. Afterwards, it was washed with distilled water until the wash water was neutral and dried for 1 h at 150 °С. The final product was designated as RHp−850VA.

RHp−850VA characteristics: BET surface area – 665 m2 g−1; average pore diameter − 1.6 nm; total pore volume – 0.53 cm3 g−1; iodine adsorption capacity (AI) – 71.8%; static exchange capacity for NaOH (SECNaOH) – 1.12 mg-eq g−1.

Fines with a fraction of −5.0 + 2.0 mm of special coke produced at the thermal oxidation coking plant at the Center of Metallurgy branch of the National Center on Complex Processing of Mineral Raw Materials of the Republic of Kazakhstan RSE (East Kazakhstan Region) were characterised elsewhere54.

Analytical methods

The chemical composition of the solid samples was determined by analysing the ash fraction after volatile matter was removed at 1000 °C. Ash decomposition was carried out with aqua regia. In the resulting solutions, Ca, Mg, Mn, Pb, Cu, and Zn were determined using atomic absorption spectrometry (Agilent AA240FS, Agilent Technologies, USA); SiO2 was determined by gravimetric analysis; Al₂O₃ and Fe were determined by volumetric analysis; and TiO₂ and P were determined by colorimetric analysis (KFK-3-01, Sergiev Posad, Russia). K and Na were determined using flame photometry (FPA-2-01; Sergiev Posad, Russia). The analyses were performed at least twice. Permissible discrepancies between primary and control determinations were within the standards of internal laboratory quality control at the 90% confidence level.

Elemental analysis was performed with a CHNOS elemental analyser Vario MICRO Cube (Elementar Analysensysteme GmbH, Hanau, Germany).

SEM and X-ray spectral microanalysis were performed using a Superprobe 733 (JEOL Ltd., Tokyo, Japan) and JEOL JXA 8230 (JEOL Ltd., Tokyo, Japan) microanalyzer. Microanalysis was carried out in energy-dispersive (EDS) and wavelength dispersive (WDS) spectroscopy modes. EDS and WDS elemental mapping were also conducted as described in70,71,72. SEM images, EDS and WDS elemental mapping and spectra are presented in Figures S1–S6 in the Supplementary Materials.

GC–MS analysis of the production solution before and after contact with the adsorbents was conducted on a 7890 A/5975 C Triple-Axis Detector diffusion pump-based GC‒MS system (Agilent, Wilmington, DE, USA) equipped with a split/splitless inlet as described previously73. The inlet was equipped with a 0.75 mm ID SPME liner (Supelco, Bellefonte, PA, USA) and operated in splitless mode. For separation, a 60 m × 0.25 mm DB-624 (Agilent, Santa Clara, CA, USA) column with a film thickness of 1.4 μm was used at a constant He flow rate of 1.0 cm3 min−1. The oven temperature was programmed to increase from an initial 35 °C (held for 10 min) to 250 °C (held for 20 min) at a heating rate of 10 °C min−1. The MS detector was operated in selected ion monitoring mode (m/z 33–550). Prior to analysis, 10 cm³ of the sample was preincubated for 10 min at 40 °C with the addition of 3 g of sodium chloride, following the procedure described in74. Extraction was carried out by solid-phase microextraction (SPME) using a 50/30 µm DVB/Car/PDMS fibre (Supelco, Bellefonte, PA, USA) in headspace mode for 10 min at 250 rpm and 40 °C. The desorption time was 5 min. The chromatograms are shown in Figures S7–S9 in the Supplementary Materials.

The mass fraction of rhenium in the production solutions before and after sorption was determined using X-ray fluorescence (XRF) with a portable X-ray fluorescence spectrometer Niton XL3t (Thermo Scientific Portable Analytical Instruments Inc., Tewksbury, MA, USA).

The content of organic impurities was determined from the chemical oxygen demand using the dichromate method. Excess potassium dichromate after the oxidation of organic substances in solution under acidic conditions was titrated with an iron(II) sulfate solution. The concentration of organic substances (CO2, mg dm− 3) was calculated from the change in the iron sulfate volume with the concentration of CFeSO4 (g eq) dm–3), which was used for titrating excess potassium dichromate in the blank test (V1, cm3) and in the test sample (V2, cm3) and was taken for analysis in volume Vs (cm3), using a coefficient of 8000 for conversion to mg dm–3 of O2 (calculations here and below were performed using Microsoft Excel 2016)75:

$$\:{C}_{O_{2}}=\frac{({V}_{1}-{V}_{2}){C}_{FeSO_{4}}8000}{{V}_{s}}$$
(1)

IR spectra of solid samples were obtained on an FT/IR-6X Fourier transform IR spectrometer (JASCO, Hachioji, Japan) in the spectral range of 4000–400 cm−1 from preparations in tablet form after dilution at 1:200 with KBr. The KBr spectrum was recorded as a reference spectrum. Spectra Manager Ver. 2.5 software was used. Spectrum analysis was carried out using specialised literature and IR spectral databases in KnowItAll (JASCO Edition Ver. 23.1) software. To more accurately determine the appearance of new bands in the spectra of the adsorbents after contact with the solution, spectra were obtained by subtracting the IR spectra of the loaded and fresh adsorbents.

Re and organic impurity adsorption experiments

Batch adsorption experiments were performed at room temperature using artificial and realistic aqueous solutions with different concentrations of Re and organic impurities. The adsorbent was mixed with solutions at a ratio of S: L = 1:100 in an Erlenmeyer flask. The experiments were carried out under constant stirring conditions (150 rpm) for 0.5–1 h at different pH values (2; 5.3). The solutions were filtered after the experiments, and the residual concentrations of Re ions and organic impurities in the filtrate were analysed as described above.

The degree of Re ions and/or organic impurities removed from the solutions was calculated using the following expression76,77,78:

$$\:\text{\%}\:removal=\frac{({c}_{0}-{c}_{e})\cdot\:100}{{c}_{0}},$$
(2)
$$\:{a}_{e}=\frac{{c}_{0}-{c}_{e}}{m}\cdot{V}$$
(3)

where c0 and ce are the initial and equilibrium liquid-phase concentrations (mg dm−3), respectively; V is the solution volume used (dm3); and m is the adsorbent mass (g).

Re and organic impurity adsorption was carried out under dynamic conditions using a cascade of four series-connected glass sorption columns with a diameter of 16 mm in continuous mode. The first three purification columns, connected in series, were filled with special coke fines (m = 10 g, V = 20 cm3). The last adsorption column was filled with carbon adsorbent from RH (m = 2–3 g, V = 20 cm3). Adsorption was carried out at specific loading rates from 2 to 10 h− 1.

The full dynamic exchange capacity of the adsorbents for Re ions and/or organic impurities (FDEC, mg g−1) was determined with a focus on rhenium adsorption. Adsorption was conducted until the concentration of rhenium ions in the filtrate was equal to the concentration of rhenium ions in the initial solution. In the case of special coke fines, the process was conducted until the residual concentration of rhenium ions or organic impurities in solution plateaued after filtration. The FDEC (mg g−1) for Re ions and/or organic impurities was calculated by the following formula79:

$$FDEC=\frac{V_{fe}\cdot{c}-\sum{V}_f\cdot{c_f}1000}{m{s}}$$
(4)

where Vfe is the total volume of the filtrate that passed through the adsorbent layer until the rhenium concentrations in the filtrate and the initial solution were equal (dm3); c is the concentration of Re or organic impurities in the initial solution (mg dm− 3); Vf is the total volume of the filtrate that passed through the adsorbent layer until the appearance of the Re ions or organic impurities, respectively (dm3); cf is the concentration of Re ions or organic impurities, respectively, in the filtrate after the appearance of Re ions or organic impurities, respectively (g dm− 3); and ms is the adsorbent mass (g).

The degree of Re ions and/or organic impurity removal in the column experiments was determined according to expression (2).

All the experiments described in this paper were performed at least twice.

Results

Carbon material characteristics

The chemical and elemental compositions of the initial RH, pyrolysed RH (RHp) and special coke fines are presented in Tables 1 and 2.

Table 1 Chemical composition of RH, RHp, and FSC.
Full size table
Table 2 Elemental analysis of RH, RHp, and FSC (%, minimum accuracy 0.1% abs.).
Full size table

The results of examining RH, RHp, and FSC by scanning electron microscopy are shown in Fig. 1.

A comprehensive characterisation of the investigated samples indicates that RH, pyrolysed RH, and special coke fines are carbon materials containing mineral components. The principal elements of the initial and pyrolysed RH are carbon, silicon, and oxygen31. According to the chemical and elemental analysis data (Tables 1 and 2), all the samples have the same qualitative composition and fairly similar quantitative compositions. The primary difference lies in the carbon content and silicon dioxide concentration. In the special coke fines, the mass fraction of carbon is two times greater than that of RH and RHp.

Therefore, while the materials exhibit similar chemical and elemental compositions, their surface morphologies and near-surface layer structures differ significantly. As shown in the SEM image (Fig. 1a), the RH has a fibrous structure31,39. The fibres are covered with surface layers on both sides. The inner surface layer appears as a protective film. The outer surface layer resembles the structure of a corn cob, as it is covered with globular protuberances39. The surface topology of the RH is preserved after pyrolysis (Fig. 1b). At the fracture site of a pyrolysed particle (Fig. 1b), voids between the carbon fibres are clearly visible, forming a porous structure of the material. According to the SEM data (Fig. 1c), the special coke fines consist of porous and dense particles of different shapes. In accordance with the results of our previous studies and the data presented in53,54,66, the specific surface area (up to ∼70 m2 g−1) and sorption pore volume (0.08 and 0.03 cm3 g−1, respectively) of RHp and FSC are similar.

Fig. 1
figure 1

SEM images of RH (a), RHp (b), and FSC (c).

Full size image

Batch studies

Organic impurity and rhenium ion removal from the production solution by different adsorbents

The adsorption activity of the carbon materials (RH, RHp, FSC) towards organic impurities and rhenium was evaluated under static conditions using a production solution containing 220 mg dm− 3 organic impurities and 355 mg dm− 3 rhenium over a 1-hour contact period at different pH values. The degree of purification from organic impurities when RH was used was 23% at pH 5.3 and 14% at pH 2 (Fig. 2a, b). Moreover, rhenium was essentially not extracted and remained in solution. The degree of organic impurity removal with respect to RHp was 7% at pH 5.3 and 25% at pH 2. However, the degree of target metal (rhenium) extraction was high (42%) at pH 5.3. At pH 2, however, rhenium was essentially not adsorbed (1%). The degree of organic impurity removal by the FSC sample at pH values of 5.3 and 2 was 14% and 34%, respectively. Rhenium removal by this sample varied depending on the solution pH within the range of 0−1%.

Fig. 2
figure 2

Adsorption of rhenium ions and organic impurities by RH, RHp, and FSC from a realistic production solution containing 355 mg dm⁻³ rhenium and 220 mg dm⁻³ organic impurities over 1 h in batch mode at pH 5.3 (a) and pH 2 (b).

Full size image

Since the best results for selective organic impurity extraction were obtained by using special coke fines in an acidic solution, further experiments were conducted using rhenium-containing solutions with a pH of 2. The organic impurities and rhenium adsorption isotherms of special coke fines from the production solution with higher concentrations of organic impurities (789 mg dm⁻³) and rhenium (490 mg dm⁻³) over time at a ratio of 1 g of adsorbent per 100 cm³ of solution are shown in Fig. 3. The organic impurity adsorption value by special coke fines increases (up to 1.4 mg g−1) with increasing contact time of the adsorbent with the solution up to 100–120 min, after which a decrease in this indicator is observed. In contrast, rhenium adsorption tends to increase consistently with increasing solution–adsorbent contact time. In general, these data confirm the adsorption activity of FSC towards organic impurities and demonstrate its low adsorption capacity for rhenium ions. Accordingly, FSC can potentially be used for the prepurification of rhenium-containing solutions to remove organic impurities.

Fig. 3
figure 3

Changes in the аdsorption of organic impurities and rhenium by FSC from realistic production solution containing 490 mg dm⁻³ rhenium and 789 mg dm⁻³ organic impurities over time.

Full size image

Among all the materials tested, RHp showed the greatest activity towards rhenium. However, the removal rate of 42% is not efficient enough from an industrial point of view. To increase RHp activity, this material was subjected to activation by analogy with another plant raw material48, as described above. A comparative analysis of the textural characteristics of RHp and RHp−850VA revealed that activation with water vapour followed by treatment with a sodium hydroxide solution increased the specific surface area of the adsorbent by a factor of 10 and total pore volume by a factor of more than 6, respectively, while the size of the sorbing pores decreased (from 2.3 nm to 1.6 nm). As a result, the adsorption activity index for iodine increased by a factor of 8 (72% versus 9%), as confirmed by the data presented in53.

RHp−850VA was tested as an adsorbent to remove rhenium ions from production solutions purified from organic impurities. The degree of rhenium extraction was more than 60% when the experiment was carried out in batch mode for 1 h using a solution with a rhenium concentration of 500 mg dm− 3 and a pH of 2.

Rhenium adsorption by RHp−850VA

The capacity of RHp−850VA to adsorb rhenium ions was evaluated using artificial ammonium perrhenate solutions with rhenium concentrations ranging from 20 to 100 mg dm⁻³ at an adsorbent-to-solution ratio of 0.2 g:20 cm³ over contact times of 30 and 60 min (the contact times were chosen on the basis of the results of previous studies48. As shown in Fig. 4, the RHp−850VA adsorption capacity increases with the initial concentration of the adsorbate. At low initial rhenium concentrations (up to 60 mg dm⁻³), the contact time of the adsorbent with the solution has virtually no effect on the adsorption value. The effect of contact time becomes noticeable at higher concentrations of the adsorbate (above 60 mg dm⁻³). The adsorption capacity increases with increasing contact time and reaches almost 9 mg g−1 at 60 min when the initial Re concentration is 100 mg dm⁻³.

In contrast to the adsorption capacity of the adsorbent, the removal efficiency of the adsorbate decreases slightly with increasing initial concentration (20–100 mg dm⁻³) in the initial solution. The rhenium removal efficiency at a contact time of 60 min was greater than that at 30 min, particularly at high concentrations of the adsorbate. It was greatest (90%) at low initial rhenium concentrations (up to 60 mg dm⁻³) at both contact times and decreased to 70% for 30 min of contact when the initial rhenium concentration reached 100 mg dm⁻³. With increasing adsorbate concentration (20–100 mg dm⁻³) in the initial solution at a short contact time between the solution and the adsorbent, the removal efficiency of the adsorbate decreased because of the saturation of the available adsorption sites on the RH-based adsorbent. The authors21 also reported a similar trend in the change in the percentage of removal of adsorbate ions by various biosorbents with increasing initial concentration in the solution. With prolonged contact time, the adsorbate can penetrate into the deeper layers of the adsorbent. As a result, the adsorption capacity of the adsorbent increases, while the removal efficiency of the adsorbate also remains consistently high (∼90%).

Fig. 4
figure 4

Influence of the initial Re concentration and contact time on RHp−850VA adsorption capacity and removal efficiency.

Full size image

The presented data are consistent with the trend in the distribution coefficient (Kd), which is calculated according to the formula in80,81.

$$\:{K}_{d}=\frac{q}{c}$$
(5)

where q is the equilibrium concentration of the substance on the adsorbent (mg g−1), and c is the equilibrium concentration of the dissolved substance (mg dm−3).

K d is an important indicator of the ability of the adsorbent to take up ions in quantities as large as possible when their content in the medium is minimal81. As shown in Table 3, the calculated distribution coefficient is sufficiently high, indicating a strong affinity of the adsorbate for the adsorbent. As the available adsorption sites on the adsorbent become filled with increasing initial solution concentration, the adsorption efficiency decreases, and the distribution coefficient slightly decreases.

Table 3 Distribution coefficient values for various concentrations of rhenium ions in solution at pH 6.
Full size table

Isotherm models

The Langmuir (6) and Freundlich (7) isotherm models are typically used to describe the adsorption of an adsorbate on an adsorbent surface82:

$$a=\frac{a_m\cdot{K_L}\cdot{c}}{1+K_L\cdot{c}}$$
(6)
$$a=K_F\cdot{c}^{1/n}$$
(7)

where а is the adsorption capacity (mg g− 1), am is the adsorption monolayer capacity or maximum adsorption (mg g− 1), c is the equilibrium concentration of the dissolved substance (mg dm−3), KL is the adsorption equilibrium constant, and KF and 1/n are empirical constants for the adsorbent‒adsorbate system at a given temperature.

To determine the isotherm model parameters of rhenium adsorption by RHp−850VA, the above equations were linearised as follows78:

$$\:\frac{1}{a}=\frac{1}{{a}_{m}}+\frac{1}{{a}_{m}{K}_{L}}\frac{1}{c}$$
(8)
$$\:loga=log{K}_{F}+\frac{1}{n}logc$$
(9)

Additionally, the dimensionless parameter RL as the separation factor was calculated using the following equation78:

$$\:{\:R}_{L}=\frac{1}{1+{K}_{L}{c}_{0}}$$
(10)

where c0 is the initial concentration of the adsorbate (mg dm−3).

The Langmuir and Freundlich isotherms are shown in Fig. 5, and the corresponding parameters of the process under study are listed in Table 4.

Fig. 5
figure 5

Langmuir (a) and Freundlich (b) isotherms for Re adsorption by RHp−850VA.

Full size image
Table 4 Langmuir and Freundlich isotherm parameters for Re adsorption by RHp−850VA.
Full size table

According to the R2 determination coefficients (Table 4), the parameters of both the Langmuir (0.998) and Freundlich (0.994) isotherms correspond very well to the experimental data of Re adsorption by RHp−850VA for 60 min. It is believed78,82,83 that the Langmuir equation describes monolayer adsorption, whereas the Freundlich isotherm equation applies to multilayer adsorption on a heterogeneous surface. Evidently, the RHp−850VA adsorbent surface is neither strictly homogeneous nor completely heterogeneous.

According to the RL factor (0<0.185−0.278<1), Re adsorption by RHp−850VA is favoured77,78,84. The heterogeneity parameter 1/n (1 < 0.5−0.8 < 10) also characterises Re adsorption on the RHp−850VA surface as favourable77,78,82,83,84,85. Moreover, the calculated adsorption capacity values (11.4−20.7 mg g−1) were found to match the experimental values (7−8.7 mg g−1).

Both the Langmuir and Freundlich models are adequate, and the difference between them in the investigated concentration range is not critical. However, according to the R2 (0.998), the Langmuir model better conforms to the experimental data. The approximation errors RMSE (root mean square error) and χ² (goodness of fit), which are calculated using Eqs. (11) and (12) and listed in Table 4, also indicate that the Langmuir model provides a better fit31. Obviously, monolayer adsorption is somewhat predominant27.

$$\:RMSE=\:\sqrt{\frac{1}{p}\sum\:_{i=1}^{p}({a}_{exp,\:i}-{a}_{cal,\:i}{)}^{2}}$$
(11)
$$\:{\chi}^{2}=\sum\:_{i=1}^{p}\frac{({a}_{exp,\:i}-{a}_{cal,\:i}{)}^{2}}{{a}_{cal,\:i}}$$
(12)

where p is the number of data points, aexp, i are the experimental values, and acal, i are the calculated values.

Adsorption kinetics

Different kinetic models have been used to determine the mechanism of the interaction between an adsorbate and an adsorbent. The pseudo-first-order and pseudo-second-order models are typically employed to describe the adsorption process. Using a scientific approach based on the analysis of published data and presented in78, a pseudo-second-order model was used to describe the behaviour of rhenium on the RHp−850VA surface. The following equation was used as the linear representation of this model78:

$$\:\:\frac{t}{{q}_{t}}=\frac{1}{{K}_{2}{q}_{e}^{2}}+\frac{t}{{q}_{e}}$$
(13)

where qe is the equilibrium adsorption capacity (mg g− 1), qt is the adsorption capacity at time t (mg g− 1), K2 is the rate constant from the pseudo-second-order (g mg−1 min−1), and t is the contact time (min).

The pseudo-second-order kinetic model for rhenium adsorption on the RHp−850VA surface is presented in Fig. 6. The corresponding linear parameters are shown in Table 5. Analysis of the R2 (0.998) and RMSE (0.069) values indicates that the pseudo-second-order kinetic model fits the experimental data very well. The calculated adsorption capacity value (9.425 mg g− 1) almost completely matches the experimental value (8.7 mg g− 1). According to27,31,78,85,86,87,88,89 and numerous other studies cited in the literature90, when an adsorption process is described by the pseudo-second-order equation, chemisorption is considered the dominant mechanism. In90, however, mathematical analysis revealed that the pseudo-second-order model is not associated with any rate-limiting mechanism. Its use does not require assumptions regarding chemisorption or kinetics governed by chemical reaction or diffusion. Therefore, an adsorption process modelled by the Langmuir equation may also be represented by the pseudo-second-order equation90.

Fig. 6
figure 6

Pseudo-second-order model for Re adsorption by RHp−850VA.

Full size image
Table 5 Linear parameters of the pseudo-second-order model for Re adsorption by RHp−850VA.
Full size table

Column modes

Adsorption of organic impurities and rhenium in a cascade sorption installation

The possibility of selective rhenium and organic impurity adsorption was studied on a cascade installation of four series-connected columns with a diameter of 16 mm. Three purification columns were filled with FSC for trapping organic compounds. The fourth sorption column was filled with the RH sorbent RHp−850VA for rhenium extraction.

A production solution with an organic impurity concentration of 764 mg dm⁻³ and a rhenium concentration of 640 mg dm⁻³ was used for adsorption. The solution was fed to the first column of the installation at specific loadings of 2, 4, and 10 sorption (specific) volumes per hour (2, 4, and 10 h−1, respectively) and then allowed to flow under gravity into each subsequent column. The filtration process was conducted until the adsorbent bed was completely saturated with rhenium, as determined by the solution concentration at the outlet of the last column. The process was terminated when the rhenium concentration in the solution at the installation outlet reached the initial value. The experimental results are shown in Fig. 7. As shown in this diagram, the best results were achieved at a specific loading of 4 h−1. Evidently, the specific loading of 2 h−1 is too slow and 10 h−1 is too fast to provide maximum extraction of the adsorbates because of the diffusion processes and short contact time between the adsorbent and solution, respectively. The authors of91, while studying the adsorption of heavy metals on a RH-based biomatrix in column experiments, also determined that the optimal filtration rate is 0.2–1.0 cm³ min⁻¹. The extraction degree of the adsorbate in this flow rate range was 65–100%. When the filtration rate increased to 2 cm³ min⁻¹, the degree of extraction decreased significantly. In this context, further experiments were carried out at a specific loading of 4 h−1.

Fig. 7
figure 7

Influence of the specific loading on the removal efficiency of organic impurities and rhenium.

Full size image

The data in Fig. 8a demonstrate that throughout the entire period of operation of the cascade installation, the degree of organic impurity removal remained sufficiently high, fluctuating only slightly within 74–80%. Moreover, the degree of rhenium ion removal decreased from 97% (after the first 4 specific solution volumes) to 5% (after 52 specific solution volumes). Overall, the full dynamic exchange capacity for organic impurities was 22.5 mg g⁻¹, whereas the full dynamic exchange capacity for rhenium exceeded 120 mg g⁻¹ (Fig. 8b), which is significantly greater than the adsorption capacity of the adsorbent determined under static conditions91.

Fig. 8
figure 8

Dependence of the degree of rhenium and organic impurity removal (a) and FDEC for rhenium and organic impurities (b) on the specific volume of the filtered solution containing 640 mg dm⁻³ Re and 764 mg dm⁻³ organic impurities.

Full size image

GC–MS

SPME in combination with GC–MS was used to determine the organic impurity composition in the production solutions before and after contact with the FSC and RHp−850VA adsorbents. This approach is an innovative alternative for the determination of volatile organic compounds (VOCs) in water. SPME has gained widespread acceptance in environmental, pharmaceutical and food analyses because of its sensitivity, selectivity and ability to automate. Recently, SPME-based methods have been successfully used in many other official methods74. As shown in Table S1 and Figure S7, the initial production solution contains compounds belonging to the classes of alkanes and their derivatives, aldehydes, aromatic hydrocarbons, carboxylic acids, and heterocycles. According to the data in Tables S1–S3, the peak area of the initial production solution (41,031,607 a.u.) is substantially greater than that after contact with the special coke fines (9,711,223.8 a.u.) and the RHp−850VA adsorbent (7,969,004.9 a.u.). Considering that the peak area is proportional to the substance content in the sample (equal doses of solutions were used in the GC–MS analysis)92, it can be inferred that the organic compound content in the solutions after adsorption on the FSC and the RHp−850VA adsorbents decreases by factors of approximately four and five, respectively. Analysis of the data in Table S2 and Figure S8 reveals that after the solution is passed through the column packed with the special coke fines, 2-ethyl-2,3,3-trimethylbutanoic acid and 2-methylpropanoic acid completely disappear. The contents of heptanal and hexanal decrease. Consistent with the reduction in the corresponding peak areas, these compounds together account for almost 50% of the total organic impurity content. After passing the purified solution through the column packed with the RHp−850VA adsorbent (Table S3, Figure S9), the residual heptanal disappears, and the peak area of hexanal decreases, indicating a reduction in its content. The comparative results of the semiquantitative determination of the organic impurity content in the initial production solution and in the solutions after passage through the columns packed with the special coke fines and the RH–based adsorbent (Tables S1–S3, Figures S7–S9) indicate the potential of using FSC to remove organic impurities from rhenium-containing solution prior to rhenium adsorption.

Distribution of Re ions on the RHp−850VA adsorbent surface

To study the distribution of rhenium ions on the surface of the carbon adsorbent RHp−850VA, samples of the adsorbent before (Re content 0 mg g⁻¹) and after rhenium saturation (Re content >100 mg g⁻¹) were examined using scanning electron microscopy (SEM) and X-ray spectral microanalysis with the application of energy‒dispersive spectroscopy (EDS) and wavelength‒dispersive spectroscopy (WDS)70,71,72. According to the EDS mapping results (Figure S1a) and the pseudocolour method for determining concentration distributions (which range from the dark to the minimum element content to violet, blue, light blue, green, orange, red, and finally white, corresponding to the maximum element content), the main element in the composition of the studied material is clearly carbon93. Silicon, oxygen, and aluminium are localised in certain areas of the carbon sorbent, and their surface distribution areas are identical. The distribution patterns of potassium, sodium, calcium, and, to a much lesser extent, iron, copper, and zinc repeat the distribution patterns of silicon, oxygen, and aluminium. This finding indicates that residual silicon in the carbon adsorbent (after alkali treatment) can be present in the form of alkali and alkaline earth metal aluminosilicates.

The distribution pattern of oxygen, zinc, and copper on the adsorbent surface changes after contact with the production solution (Figure S1b), which corresponds to the carbon distribution pattern. After contact with the solution, the distributions of sulfur, chlorine, iodine, and rhenium on the carbon adsorbent surface were recorded. The strong correspondence between the oxygen and rhenium distribution patterns, particularly in areas of increased rhenium accumulation, suggests that rhenium is distributed on the surfaces of the carbon particles in combination with oxygen. This fully corresponds to the co-phase distribution of elements in areas of their highest concentration during EDS mapping70. This conclusion is valid since the analyses were performed with a high degree of reliability, which excludes false or ambiguous interpretations. The observed range of rhenium concentrations in the adsorbent particles under the condition of point EDS microanalysis was 2–33% ms (Figures S2–S3).

This conclusion is confirmed by the results of the more sensitive wavelength-dispersive microanalysis method. As shown in Figure S4, the distributions of oxygen and rhenium occur in strict accordance with the orientation of the carbon fibres. The distributions of the other elements are not associated with the carbon surface since they also occupy the pore space. The limit of the rhenium concentrations recorded by the WDS method was 3–4% ms (Figures S5–S6).

IR spectroscopy of adsorbents before and after contact with the production solution

Changes in the vibration frequencies of functional groups after the interaction of an adsorbate with an adsorbent are among the key factors for explaining the adsorption mechanism. Accordingly, infrared spectroscopy studies of the FSC and RHp−850VA adsorbents before and after their contact with the production solution were carried out.

According to the IR spectroscopy data (Fig. 9a), in the spectrum of the special coke fines after contact with the rhenium solution containing organic impurities, a shift in the stretching vibration band of the hydroxyl groups in the 3400 cm⁻¹ region is observed, and its intensity significantly decreases. The bending vibration band of the OH group in the 1630 cm⁻¹ region disappears. However, a new band appears at 1637 cm⁻¹, which can be explained by the appearance of carbonyl (C = O) groups of acids, aldehydes, and ketones. These groups are characterised by vibrations in the 1850–1640 cm⁻¹ region78,91,94,95,96. A vibration band at 1795 cm⁻¹ corresponds to aldehyde and ketone groups56,77,84,97. The subtraction spectrum of the IR-spectra of spent and fresh FSC in the range of 1850–1600 cm−1 (half of the fresh FSC sample spectrum was subtracted from the spent FSC sample spectrum, after which the baseline was corrected) is shown in Figure S10a. In this fragment of the spectrum, the noted changes in the region of 1800–1600 cm⁻¹ are clearly visible. These changes may be caused by the adsorption of organic impurities. This is consistent with the GC–MS data (Tables S1–S3), which show that after the solution passes through the FSC bed, the carboxylic acids completely disappear, while the aldehyde content decreases. The sulfate group (SO₄²⁻) also settles on the FSC adsorbent surface. This is confirmed by the appearance of bands (Fig. 9a) at 1091 cm⁻¹ and 621 cm⁻¹98.

In the obtained IR spectrum of the RHp−850VA sample (Fig. 9b), after contact with the production solution, the appearance of a vibration band of the aldehyde group at 1730 cm⁻¹ was recorded. This suggests that the adsorbent extracts organic impurities from the solution. As mentioned above, GC–MS revealed that after the purified solution is passed through the RHp−850VA bed, the peak area of the aldehydes decreases significantly. Conversely, in the IR spectrum of this adsorbent, practically no changes appear in the −OH group vibration region. In the fresh adsorbent, the characteristic band of hydroxyl group stretching vibrations appears at 3434 cm− 1, and in the spent sorbent, it appears at 3435 cm− 1. The band width and intensity are preserved as well. Bands at 1114 cm⁻¹ and 604 cm⁻¹, characteristic of SO₄²⁻ group vibrations98, are also present (Fig. 9b). In contrast to the spectrum of the special coke adsorbent, a band appears at 905 cm⁻¹, corresponding to stretching vibrations of the rhenium–oxygen bond98,99. This band is clearly visible in the subtraction spectrum in the range of 1400–400 cm−1 (the spectrum of the fresh RHp−850VA sample was subtracted from the spectrum of the spent RHp−850VA sample, after which the baseline was corrected (Figure S10b)). This finding confirmed that rhenium adsorption occurred.

Rhenium desorption from the RHp−850VA adsorbent

A rhenium-loaded column packed with the RHp−850VA adsorbent was used to study the possibility of rhenium removal. The adsorbent in the column was preliminarily washed with water to remove the sulfuric acid residue from the rhenium-containing production solution. The desorption process was conducted with a hot (40 °C) ammonia solution with a concentration of 8%. The eluent was passed at a rate of 100 cm³ h⁻¹ at a specific loading of 5 h⁻¹. Under these conditions, the desorption efficiency of the target metal reached ∼90%.

After rhenium removal, the RHp−850VA adsorbent was washed with water until the wash water reached neutrality. The moist adsorbent was then placed in a stainless steel reactor equipped with a vent for volatile removal and heated in a furnace to 650 °C. The adsorbent was held at this temperature for 30 min. Subsequently, the reactor was removed and cooled in a water reservoir. The adsorbent was sieved, and the fraction > 2 mm was collected. The losses amounted to 10%. When the regenerated adsorbent was reused (at least once), it retained its initial adsorption capacity (the difference between the extraction efficiencies of the fresh and regenerated adsorbents did not exceed 2%).

Fig. 9
figure 9

IR spectra of the FSC (a) and RHp−850VA (b) adsorbents before and after contact with the production solution.

Full size image

The complete cycle from the production to the reuse of spent sorbents is presented in the process flow diagram (Fig. 10). This diagram shows that the proposed method of producing and applying carbon adsorbents based on RH and special coke fines is environmentally sustainable. In accordance with the approach implemented in this method, the byproducts generated during the pyrolysis and activation of RH (organic condensate and sodium silicate solution) are used in the production of refractory materials. This has been previously verified, proven, and described in detail100. Spent special coke fines are also used as a burnout additive in the production of refractory materials. Wash and wastewater (filtrate) are returned to the corresponding stage of the technological process. Overall, the proposed technology is practically waste-free.

Fig. 10
figure 10

Technological scheme for RH and FSC processing.

Full size image

Comparative analysis of the ability of the RH-based adsorbent to extract re ions

Various RH–based adsorbents were tested for their ability to extract rhenium from model and realistic production solutions in batch and column modes (Table 6). The adsorbents differ in their production methods. Korabayev et al.49 produced activated carbons from RH and rice husk lignin using a combined pyrolysis and steam-activation process. The plant feedstock was first carbonised at 650 °C for 20 min. The temperature was then increased to 850 °C, and activation with water vapour was carried out for 35 min. Yefremova et al.50 and Zharmenov et al.51 prepared both high-ash and deashed granular sorbents. For these materials, the RH was carbonised at 650 °C for 30 min and then milled to a − 0.25 + 0.04 mm fraction. The resulting feedstock was mixed with a 35% sugar solution and granulated to a − 2.5 + 0.63 mm fraction. The granules were dried at 105–110 °C for 2 h, subsequently heated at 650 °C for 30 min, and finally activated with live water vapour for 30 min at 850–900 °C. To produce deashed granular carbon, pyrolysed RH was boiled for 2 h in an alkaline solution containing 70 g dm⁻³ NaOH at a S(g): L(cm³) ratio of 1:15. All subsequent processing steps were identical to those described above. Kablanbekov et al.48 produced a sorbent from rice husk cellulose. For this purpose, cellulose pyrolysed at 600 °C and activated at 850 °C (each for 30 min) was treated with a sodium hydroxide solution as described previously, but with an S: L ratio of 1:10 and a treatment duration of 90 min.

Table 6 Re adsorption by RH-based adsorbents.
Full size table

The data in Table 6 show that the most active adsorbents towards rhenium ions are granular deashed activated carbons from RH. The deashed sorbent is characterised by high iodine adsorption (81.4%) and an exchange capacity for rhenium of 47.4 mg g⁻¹ in batch mode when an artificial solution containing 100 mg dm⁻³ rhenium is used. Among the nongranulated adsorbents, the adsorption properties of the RHp−850VA and SCV-35A samples are very similar because they were produced using nearly identical methods. The difference is that compared with SCV-35A (650 °C), RHp−850VA was produced at a significantly lower carbonisation temperature (450 °C). The literature88 indicates that the optimal carbonisation temperature for RH is 600 °C. The results of the present study, however, show that 450 °C is sufficient for the carbonisation of RH. RHp−850VA even has a slightly higher iodine adsorption value (71.8%) than that of SCV-35A (67.5%). Additionally, RHp−850VA achieved high rhenium extraction (97%) from realistic production solutions in column experiments, surpassing that of the granular deashed sorbent (70% rhenium extraction). A comparison of the results obtained in the present study with published data confirms the importance of ensuring a high carbon content in the RH–based adsorbent by removing silicon dioxide present in the large amount in the raw material88. The advantage of the RH–based adsorbent proposed in this study lies in the significant reduction in the temperature required for its production while maintaining high adsorption performance for rhenium.

Conclusions

A selective adsorbent composition has been developed for the purification of a rhenium-containing solution from organic contaminants, followed by the extraction of rhenium. The development of the composition of adsorbents is part of a technological scheme for the processing of carbon- and silica-containing waste from the coke and rice industries. This technology involves the closure of material flows and aligns with the trends of green chemistry, circular economy, and resource conservation, ensuring the production of sustainable products.

A new adsorbent composition was tested for rhenium adsorption processes using both model and realistic production solutions. The experimental results demonstrate that special coke fines selectively extract organic impurities and, when combined with a RH–based adsorbent, provide excellent performance indicators for a cascade system at a specific load of 4 h–1, which simulates the industrial sorption-based rhenium extraction process. The full dynamic exchange capacity of special coke fines for organic impurities is 22.5 mg g⁻¹. The full dynamic exchange capacity of the RHp−850VA adsorbent for rhenium is 120 mg g⁻¹. RH–based adsorbents remove 90% of rhenium ions from aqueous solution (pH of 2), with a distribution coefficient of 1005 cm3 g−1. After their first elution with an 8% ammonia solution, the RHp−850VA sample has almost the same adsorption capacity for rhenium ions. The desorption efficiency is 90%.

The rhenium adsorption process follows both the Langmuir and Freundlich isotherm models well, but it is best described by the Langmuir (R2 = 0.998) equation. The calculated parameters revealed that rhenium adsorption onto the RHp−850VA surface proceeds favourably and can be described by a pseudo-second-order kinetic model (R2 = 0.998). According to the IR spectroscopy data, rhenium is adsorbed in the Re—O form. This is consistent with of the co-phase distribution of rhenium and oxygen on the surface of the RHp−850VA adsorbent, which was established by SEM with EDS and WDS elemental mapping.

Thus, this study establishes the feasibility of using carbon adsorbents produced from secondary raw materials to purify rhenium-containing solutions by removing organic impurities and achieving high recovery of the target metal. It also shows the possibility of using special coke fines in this process without any prior treatment and of reducing the temperature of RH carbonisation to obtain a carbon adsorbent with the required adsorption characteristics. Although the objective of the study has been achieved, the results of the laboratory studies must be verified on a larger and semi-industrial scale. Thermodynamic calculations of the adsorption process for rhenium ions are needed to confirm the nature of the interaction between rhenium and the RH–based carbon adsorbent. Particular attention will be given to determining the required number of columns in a cascade system to ensure process profitability. Additional studies will also be conducted to refine the adsorbent regeneration conditions, reduce adsorbent loss, and determine the maximum possible number of reuse cycles.