Sorbents Based on Polyacrylonitrile Fiber for Complex Recovery of Artificial ¹³⁷Cs and Natural Radionuclides from Natural Media

Abstract

Sorbents based on polyacrylonitrile fiber, containing ferrocyanides of transition metals and manganese oxides (CoMn-PAN and FeMn-PAN) or iron(III) hydroxide (CoFe-PAN) in their structure were obtained, as confirmed by the results of X-ray diffraction and energy-dispersive analyses. The selectivity of the obtained sorbents was investigated, along with their ability to sorb Cs, Ba (as an analog of Ra), P, and Be from various natural media, including river water and seawater with varying salinity of 18.2 and 33.8 ‰. The data show that the sorbents are universal for the recovery of artificial ¹³⁷Cs and natural radionuclides from the natural environments, including complex salt composition (seawater). Researching the obtained sorbents during marine expeditions confirmed the efficiency of the obtained materials based on transition metal ferrocyanides and manganese oxides (CoMn-PAN and FeMn-PAN) for the sorption of ¹³⁷Cs, ⁷Be, ²¹⁰Pb, ²¹⁰Po, ²²⁶Ra, ²²⁸Ra, and ²³⁴Th. Additionally, the sorbent based on transition metal ferrocyanides and iron(III) hydroxide (CoFe-PAN) was effective for the sorption of ¹³⁷Cs, ⁷Be, ³²P, ³³P, ²¹⁰Pb, ²¹⁰Po, and ²³⁴Th. Based on the obtained results, methods for comprehensively determining artificial ¹³⁷Cs and natural radionuclides using these sorbents were developed.

1. Introduction

The concentration of artificial and natural radionuclides from natural media makes it possible to solve several important fundamental and practical problems in ecology [1,2], radioecology [3,4,5], and oceanology [6,7,8].

Artificial radionuclide levels allow us to determine the radioecological condition of natural media [9,10], while natural ones allow us to study the processes occurring in nature. The isotope ⁷Be is an indicator of fundamental processes associated with vertical transport in the ocean [11,12], while ³²P and ³³P are markers of phosphorus biodynamics, which allows us to obtain a more complete picture of the state of the studied water area [13,14]. Isotopes ²²⁶Ra and ²²⁸Ra indicate processes associated with the flow of freshwater into marine waters, particularly submarine groundwater discharge [15,16]; and ²¹⁰Pb, ²¹⁰Po with ²³⁴Th, reflects the sedimentation parameters and flows of suspended particulate matter and particulate organic carbon [17,18].

Sorption materials have become widely used for concentrating radionuclides from natural media [19]. However, the search for optimal materials and methods continues. Of particular interest are sorbents that enable the simultaneous concentration of a range of radionuclides. This simultaneous sorption of several radionuclides allows for the determination of them within a single sorbent sample. For example, measuring a sorbent sample using gamma spectrometry can reveal the activity of ⁷Be, ¹³⁷Cs, ²¹⁰Pb, ²²⁶Ra, ²²⁸Ra, and ²³⁴Th, which are identified by different photopeaks [20].

So far, many sorbents have been developed for the recovery of artificial radionuclides [21,22,23,24], some of which allow for the simultaneous recovery of several radionuclides [25,26,27,28,29,30,31,32,33,34,35,36]. Most of these materials are designed to recover ¹³⁷Cs and ⁹⁰Sr. Thus, of great interest is a composite material based on resorcinol–formaldehyde resin and barium silicate [25], intended for the combined sorption of ¹³⁷Cs and ⁹⁰Sr from seawater. The efficiency of this materials toward ⁹⁰Sr is increased with a higher barium silicate content, but it decreases for ¹³⁷Cs. This behavior is associated with the selectivity of barium silicate for ⁹⁰Sr: the higher its content in the sorbent, the better the ⁹⁰Sr removal efficiency. The developed sorption materials exhibit moderate purification coefficients at high flow rates in filter devices, allowing for the purification of large quantities of water contaminated with ¹³⁷Cs and ⁹⁰Sr radionuclides.

In the works [26,27], Grushicheva E.A. et al. proposed the use of clinoptilolite and tripoli for the sorption of ¹³⁷Cs and ⁹⁰Sr under static conditions. It has been demonstrated that these materials exhibit high sorption parameters. For clinoptilolite, the degree of recovery of ¹³⁷Cs and ⁹⁰Sr was more than 96 and 86%, respectively, and the distribution coefficients were more than (8 ± 2) × 10³ mL/g and (4 ± 1) × 10³ mL/g, respectively. For tripoli, the degree of recovery was more than 93 and 98%, respectively, and the distribution coefficients were more than (1 ± 0.2) × 10³ mL/g and (3 ± 1) × 10³ mL/g, respectively.

In the work [28], Nikiforov A.F. et al. showed that finely dispersed aluminosilicate sorbents Ekozol-401 and Miksorb effectively and quickly recover ¹³⁷Cs and ⁹⁰Sr from natural media. The maximum distribution coefficients of ¹³⁷Cs and ⁹⁰Sr for the Ecozol-401 were (2.9 ± 0.9) × 10⁴ mL/g and (1.4 ± 0.2) × 10³ mL/g, respectively, and for the Miksorb sorbent, they were (2.7 ± 0.7) × 10⁴ mL/g and (2.7 ± 0.5) × 10³ mL/g, respectively.

Additionally, for the sorption of radionuclides ¹³⁷Cs and ⁹⁰Sr, synthesized magnetic potassium zinc hexacyanoferrate(II) was used [29]. The maximum removal values for this sorbent are achieved within the first 5 to 15 min, with up to 97.2% of ¹³⁷Cs and 35.8% of ⁹⁰Sr being absorbed. This suggests a higher selectivity of the sorbent for ¹³⁷Cs. Additionally, a significant advantage of this composite sorbent is its capability to be separated from the solution using an external magnetic field.

A sorbent based on titanium(IV) oxohydroxophosphates can be used for the removal of ¹³⁴Cs, ¹³⁷Cs, and ⁹⁰Sr from high-salt solutions similar in composition to seawater [30]. However, the distribution coefficients of Cs and Sr are low: 530 and 840 mL/g, respectively.

In the work [31], Kuznetsov A. Yu. et al. presented porous, highly filled composite materials in the form of films. Ultrahigh-molecular-weight polyethylene was used as a matrix; solid paraffin (number of carbon atoms from 24 to 40) was used as a solvent; and Na-A zeolite, strontium ferrite, and nickel ferrocyanide served as fillers. High selectivity of the sorbent containing nickel ferrocyanide particles toward ¹³⁷Cs and ⁹⁰Sr was shown. Thus, for a sample containing 70% nickel ferrocyanide, the distribution coefficient of ⁹⁰Sr reached 3.2 × 10⁴ mL/g, and it reached 5.2 × 10⁴ mL/g for ¹³⁷Cs.

One of the natural sorbents for recovering various radionuclides is bentonite clays, the main feature of which is a high sorption capacity. Thus, in the work [32], a study was conducted on the sorption of ⁹⁰Sr and ¹³⁷Cs by natural and mono-cationic (Na, Mg, and Ca) forms of bentonite clay from the Taganskoye deposit (Kazakhstan). To obtain the studied sorbents of the Na, Mg, and Ca forms, bentonite was treated four times with chlorides, after which it was washed with distilled water to remove excess salt. Bentonite and its Na and Mg forms showed a high degree of recovery of ⁹⁰Sr and ¹³⁷Cs of more than 95 and 73%, respectively; the Ca form showed lower values of the degree of recovery of ⁹⁰Sr up to 84%, but it had higher ¹³⁷Cs—90%. The distribution coefficient of ⁹⁰Sr on bentonite and its Na and Mg forms was more than 2.1 × 10³ mL/g, and for ¹³⁷Cs, the highest result was shown by the Ca form—9.1 × 10² mL/g.

In addition, sorbents have been developed that allow for the simultaneous sorption of Co [33,34] or U and Pu [35] isotopes, along with Cs and Sr isotopes. Thus, in the work [33], Efimova et al. conducted a study on the sorption of radionuclides ⁶⁰Co, ¹³⁷Cs, and ⁸⁵Sr with iron oxides α-Fe₂O₃ and Fe₃O₄. The proposed materials recovered ⁶⁰Co from drinking water best of all (for α-Fe₂O₃, the degree of recovery reached 51.4%; for Fe₃O₄—71.6%) and slightly worse for ⁹⁰Sr (for α-Fe₂O₃ the degree of recovery was 19.8%, for Fe₃O₄—34.8%); and ¹³⁷Cs was practically not recovered (the degree of recovery for α-Fe₂O₃ and Fe₃O₄ did not exceed 8.7 and 18.7%, respectively). It was also shown that the sorption efficiency increases with an increasing pH value, with the highest degree of recovery established at pH 9. For ⁶⁰Co, ⁹⁰Sr, and ¹³⁷Cs it is 93, 73, and 26%, respectively (for Fe₃O₄), and 85, 41, and 18%, respectively (for α-Fe₂O₃). It should also be noted that at any pH value, the sorption efficiency of the studied radionuclides fits into the following series: ⁶⁰Co > ⁹⁰Sr > ¹³⁷Cs.

In the work [34], Kitikova et al. investigated phosphate sorbents FD-1, FD-2, and FD-3 based on natural dolomite, obtained by acidic and acid-free phosphating of the original sample using the phosphating agents H₃PO₄ (C = 20%), H₃PO₄ + NH₄OH (pH = 10), and NaH₂PO₄, respectively, while the FD-2 sample was pre-dissolved in HNO₃. Of the obtained samples, FD-2 showed the best sorption characteristics: the distribution coefficients of ⁶⁰Co and ⁸⁵Sr were 2.4 × 10⁴ mL/g and 2.7 × 10³ mL/g, respectively. In high-salt solutions (35 g/L), the FD-2 sorbent retains its selective sorption properties toward ⁶⁰Co, while the distribution coefficient of ⁸⁵Sr decreased to 0.2 × 10³ mL/g. The high selective sorption properties of this sorbent are also preserved when recovering ⁶⁰Co and ⁸⁵Sr radionuclides from a multi-component solution.

Avrorin et al. [35] proposed a composite inorganic sorbent in the form of a non-magnetic fraction of product of a metallurgical slag processing for the removal of ²³⁵U, ²³⁹Pu, ¹³⁷Cs, and ⁹⁰Sr from aqueous solutions in the pH range of 3–11 under dynamic conditions. The composition of the material was calcium silicate Ca₂SiO₄, lithium iron oxide Li_9.28Fe_21.34O₃₂, coesite SiO₂, ferrous grossular Ca3Al_1.332Fe_0.668Si₃O₁₂, ringwoodite Fe₂SiO₄, and sodium aluminosilicate Na_14.88Al_15.26Si_32.74O₉₆. The distribution coefficient for ²³⁵U was 3.1 × 10⁵ mL/g, while that of other radionuclides was several orders of magnitude less: ²³⁹Pu—5.6 × 10³ mL/g, ¹³⁷Cs—1.3 × 10² mL/g, and ⁹⁰Sr—1.4 × 10² mL/g.

It is also necessary to identify sorbents for the extraction of ¹⁵²Eu, ²⁴¹Am, and ⁸⁵Sr [36], which are silica gel KSK-1 or ion-exchange resin KU-2–8 modified with micro-amounts of fullerenes (C₆₀ – C₇₀) in the presence of crown ether for stabilization. The selectivity of sorption by the obtained samples decreased in the series: ²⁴¹Am > ¹⁵²Eu > ⁸⁵Sr.

Sorbents for the complex recovery of natural radionuclides are less commonly reported. Among the sorbents that allow for the comprehensive recovery of natural radionuclides, the most important are sorbents based on manganese dioxide. They demonstrate high sorption characteristics when recovering isotopes of radium [37,38,39], thorium [39,40,41], ²¹⁰Pb, ²¹⁰Po [42,43], and ⁷Be [43].

Several articles [44,45,46,47] discuss the use of these sorbents for the recovery of ⁹⁰Sr from natural media under static conditions. High distribution coefficients (exceeding 10³ mg/L) have been reported [46], indicating the possibility of utilizing these sorbents for the recovery of ⁹⁰Sr from radioactive waste containing seawater. However, under dynamic conditions, these sorbents have not been effectively applied for the sorption of ⁹⁰Sr from natural environments, because they actively sorb macro-components: calcium ions; and transition metals, including both dissolved and hydrolyzed forms. In addition, as indicated above, many radionuclides (⁷Be, isotopes of Th, Ra, Pb, and Po) are effectively sorbed.

Sorbents based on iron(III) hydroxide show high sorption efficiency of thorium [48,49], ²¹⁰Pb, and ²¹⁰Po [49,50] isotopes, along with cosmogenic ⁷Be [49,51,52], ³²P, and ³³P [49,53] isotopes.

In our previous review [54], we discussed the removal of radionuclides from seawater using various sorption materials, including those designed for the simultaneous extraction of multiple radionuclides. The paper also demonstrates the effectiveness of different sorbents under real conditions. However, there is a notable deficit of research focused on the production of sorbents for the complex extraction of both artificial and natural radionuclides from liquid media of natural origin.

Therefore, earlier [55], we developed new material based on sorbents made of polyacrylonitrile (PAN) fiber combined with manganese dioxide (PAN-MnO₂) [39] and iron(III) hydroxide (Fe-H) [49], which also contained transition metal ferrocyanides. We assessed the influence of the synthesis temperature of the sorbents, the concentration and pH of the potassium ferrocyanide solution, and the concentrations of the transition metal salts used in production. Optimal formulations and conditions for the synthesis of sorbents have been established: CoFe-PAN, Con-PAN, and FeMn-PAN. The structure of the obtained sorbents was studied using infrared spectroscopy, thermogravimetric analysis, and scanning electron microscopy. Research on sorbents was carried out for the recovery of microquantities of ¹³⁷Cs [55]. The physicochemical patterns of sorption and desorption of Cs by the obtained sorbents were also described [56].

This study is a logical continuation of the works [55,56], aimed at describing the structure of the obtained sorbents, assessing their selectivity toward radionuclides and the influence of the salt content of natural waters on their sorption efficiency, and researching the obtained sorbents on large volumes of natural waters to assess the possibility of recovering artificial ¹³⁷Cs and natural radionuclides.

2. Materials and Methods

2.1. Materials

CRMs (Certified Reference Materials of solutions of metal ions) [57] produced by Ormet LLC (Yekaterinburg, Russia) were used to prepare calibration solutions and solutions for studying the sorption selectivity: Li CRM 0138:2000, Na CRM 0143:2000, K CRM 0134:2000, Rb CRM 1058:2004, Cs CRM 0160:2004, Mg CRM 0139:2000, Ca CRM 0135:2000, Sr CRM 0148:2000, Ba CRM 0351:2002, Al CRM 0306:2002, Pb CRM 0525:2003, Bi CRM 0541:2003, Cr CRM 0293:2002, Mn CRM 0300:2002, Fe CRM 0518:2003, Co CRM 0305:2002, Ni CRM 0144:2000, Cu CRM 0523:2003, Zn CRM 0032:1998, and Cd CRM 0136:2000.

Cesium nitrate, strontium nitrate, potassium dihydrogen orthophosphate, and beryllium sulfate, chemically pure for analysis, produced by ReaKhim JSC LLC (Moscow, Russia), were used as additives for studying the sorption of elements from various natural solutions.

To obtain sorbents, we used sodium hydroxide, iron(III) chloride, ammonia, potassium ferrocyanide, cobalt(II) chloride (ReaKhim LLC, Moscow, Russia), and potassium permanganate (NevaReaktiv LLC, St. Petersburg, Russia) Pro Analysis qualifications, along with PAN fiber (Moscow Wool Spinning Factory, Moscow, Russia).

2.2. Methods for Preparing the Support and Obtaining Sorbents

The methods of preliminary preparation of the support are described in detail in [39,49], and the preparation of sorbents is described in [55].

Preliminary preparation of the initial PAN fiber was performed by two methods: by treatment with potassium permanganate (conversion of PAN to the form of PAN-MnO₂) [39] and by precipitation of iron(III) hydroxide with ammonia on pre-hydrolyzed PAN (conversion of PAN to the form of PAN-Fe(OH)₃) [49]. During the process of preliminary preparation of PAN fiber, additional ion-exchange groups (carboxyl, carbonyl, etc.) are formed, increasing the amount of ferrocyanide grafted to the carrier at the following stages of obtaining sorbents.

The sorbents were obtained by treating prepared PAN fiber in a weakly acidic solution of potassium ferrocyanide, followed by treatment in a solution of a transition metal salt, washing, and drying [55].

2.3. Study of the Sorbent’s Structure

X-ray diffraction analysis before and after annealing of sorbent samples at 800 °C was performed on a D8 ADVANCE diffractometer (Bruker, Billerica, MA, USA). X-ray patterns of the samples were recorded in the angle range of 3° < 2θ < 85°, with a step of 0.02° with a count at a point of 0.6 s. The phase composition was identified using the MATCH! Program version 4.1 and Crystallography Open Database (COD).

Energy-dispersive analysis was performed on a Carl Zeiss CrossBeam the XB 1540 (Zeiss Int., Oberkochen, Germany) with an attachment for energy-dispersive analysis.

2.4. Study of Sorbent Selectivity Under Static Conditions

The study of sorption selectivity was carried out by mixing 0.1 g of the sorbent with 10 mL of a solution with a concentration of each element (Li⁺, Na⁺, K⁺, Rb⁺, Cs⁺, Mg²⁺, Ca²⁺, Sr²⁺, Ba²⁺, Al³⁺, Pb²⁺, Bi³⁺, Cr³⁺, Mn²⁺, Fe³⁺, Co²⁺, Ni²⁺, Cu²⁺, Zn²⁺, and Cd²⁺) of 10 mg/L at pH ≈ 6. Each experiment was repeated at least three times.

The concentrations of elements in solutions were determined by inductively coupled plasma–mass spectrometry on an Agilent 8800 spectrometer (Agilent Technologies, Santa Clara, CA, USA) and by inductively coupled plasma–atomic emission spectrometry on an iCAP 7600 Duo spectrometer (Thermo Scientific, Waltham, MA, USA).

The distribution coefficient (K_d, mL/g) was calculated using the equation given in [58,59].

2.5. Study of Elements’ Sorption from Various Natural Solutions Under Static Conditions

To evaluate the efficiency of the obtained sorption materials to artificial and natural radionuclides, we conducted a series of laboratory tests using solutions of natural waters with the addition of a stable element (Cs, P, Be, and Ba as an analog of Ra). The approximate composition of the river water and seawater used in these tests is given in

Table 1. Approximate composition of natural waters.

Ion	Concentration, mg/L
	Black River [60]	Black Sea [61,62]	Pacific Ocean [63,64]
Cl–	5.1–6.2	8790–10,600	19,350–25,160
HCO3−	157–186	139–202	36–147
CO32−	0.2–20	73.2–88.2	10–96
SO42−	8.6–9.3	1290–2140	2700–5580
MoO42−	(0.01–0.97) × 10−3	(1–3.4) × 10−3	(1–4.8) × 10−3
K+	0.54–0.82	139–287	298–586
Na+	4–4.5	4990–6900	10,700–14,100
Ca2+	51–58.3	217–269	408–754
Mg2+	3.2–7.9	592–756	1290–3070
Sr2+	0.13–0.15	4.46–6.05	6–14
Co2+	(0.02–0.23) × 10−3	(0.2–1.6) × 10−3	(0.1–6.4) × 10−3
Cu2+	(0.55–1.14) × 10−3	(1–7) × 10−3	(0.6–18) × 10−3
Hg2+	(0–0.04) × 10−3	(0.3–0.4) × 10−3	(0.01–1.42) × 10−3
Fe2+ + Fe3+	(42.7–90) × 10−3	(5–40) × 10−3	(3–164) × 10−3
Mn2+	(4.7–170) × 10−3	(4–40) × 10−3	(0.1–111) × 10−3
Ni2+	(0.23–0.64) × 10−3	(0.5–3) × 10−3	(0.3–5.4) × 10−3
Average salinity, ‰	0.4	18.4	33.8

.

The sorption study under static conditions was carried out by mixing 0.1 g of the sorbent with 20 mL of the studied natural solution for 48 h. Afterward, the sorbent and solution were separated by filtration. Each experiment was repeated at least three times.

The concentration of Cs and Ba in solutions was determined on a KVANT-2 atomic absorption spectrophotometer (Kortek LLC, Moscow, Russia) in an air–acetylene flame using the method described in [65]: Cs—in emission mode at a wavelength of 852.1 nm; and Ba—in absorption mode at a wavelength of 553.6 nm.

The concentration of P was determined using a KFK-3-01 photocolorimeter (JSC ZOMZ, Sergiev Posad, Russia) at a wavelength of 880 nm, using the method described in [66]; and that of Be was determined using a KFK-3-01 photocolorimeter (JSC ZOMZ, Sergiev Posad, Russia) at a wavelength of 536 nm, using the method described in [67].

2.6. Evaluation of the Sorption Efficiency for Large Volumes of Natural Waters Under Dynamic Conditions

Surface samples of seawater were collected during coastal expeditions in the waters of the Heraclean Peninsula, as well as partially during the 126th cruise of the collective use center “RV Professor Vodyanitsky” (15 March–7 April 2023) and the 71st cruise of the collective use center “RV Akademik Oparin” (4 June–15 July 2024) using a UNIPUMP BAVLENETS BV 0.12-40-U5 vibration pump (Sabline Service LLC, Moscow, Russia) and a UNIPUMP BCP 3.5-0.5-65 borehole pump (Sabline Service LLC, Moscow, Russia).

The sorption efficiency of ¹³⁷Cs, ⁷Be, ²¹⁰Pb, ²¹⁰Po, ²²⁶Ra, ²²⁸Ra, and ²³⁴Th by the obtained sorbents was determined using a two-column method by passing 250 L of filtered seawater through two columns filled with 100 mL of sorbent each, at different speeds (1–7.5 C.V./min (column volumes per minute)), using a LongerPump WT600-2J peristaltic pump (Longer Precision Pump Co., Baoding, China). The sorption efficiency (E, %) of the studied radionuclides was determined using the equation for the two-column method given in [68].

The sorption efficiency of ³²P and ³³P by the obtained sorbents was determined by a single-column method by passing 250 L of filtered seawater with the addition of potassium dihydrogen phosphate to a concentration of 0.5 μmol/L as a tracer of sorption efficiency through one column filled with 100 mL of the sorbent at different speeds (1–7.5 C.V./min) using a LongerPump WT600-2J peristaltic pump (Longer Precision Pump Co., Baoding, China). The sorption efficiency (E, %) of ³²P and ³³P was determined using the equation for the single-column method given in [52,69].

2.7. Determination of Radionuclide Activity in Sorbent Samples

To determine the activity of ¹³⁷Cs, ⁷Be, ²²⁶Ra, ²²⁸Ra, and ²³⁴Th after passing seawater, the sorbents were ashed in an LOIP LF-15/11-G2 muffle furnace (JSC LOIP, St. Petersburg, Russia) to reduce the volume and, accordingly, the self-absorption coefficient when measured on a gamma spectrometer. The resulting ash was sealed in Petri dishes and kept for 3 weeks (to achieve an equilibrium of ²²⁶Ra with ²¹⁴Pb). For the CoFe-PAN sorbent, it is not necessary to keep it for 3 weeks, since it does not sorb radium isotopes. The activity of ¹³⁷Cs, ⁷Be, ²²⁶Ra, ²²⁸Ra, and ²³⁴Th was determined on a CANBERRA gamma spectrometer with a BE3825 high-purity germanium detector (Canberra Industries, Meriden, CT, USA). Detection efficiency was estimated using ISOCS/LabSOCS software version 4.0 (Canberra Industries, Meriden, CT, USA).

To further determine the activity of ²¹⁰Pb and ²¹⁰Po, radiochemical preparation was carried out, which consisted of dissolving the ash, followed by joint electrochemical deposition of ²¹⁰Bi and ²¹⁰Po. The activity of ²¹⁰Po was determined by alpha spectrometry using the UMF-Spectrum spectrometric complex (LLC NPP Doza, Zelenograd, Russia), and that of ²¹⁰Pb was determined by ²¹⁰Bi beta radiometry on the UMF-2000 alpha-beta radiometer (LLC NPP Doza, Zelenograd, Russia).

The determination of the activity of ³²P and ³³P was not performed in this work due to the low activity of the isotopes ³²P and ³³P in seawater—0.02–0.1 Bq/m³—as well as short half-lives of 14.3 and 24.3 days, respectively. Therefore, to determine the activity of ³²P and ³³P, the volume of seawater samples had to be at least 1000 L, which significantly increases the labor intensity of the work. In addition, the time between concentration and measurement of samples must be minimal, which is often difficult in marine expeditions.

The activity of radionuclides was determined using the equation given in [68].

3. Results and Discussion

3.1. X-Ray Diffraction Analysis

The results of X-ray diffraction analysis before and after annealing of sorbent samples at 800 °C are shown in Figure 1. The X-ray diffraction pattern of the initial composites before annealing contains a characteristic peak of 2Ɵ = 17° corresponding to the rhombic structure of PAN. Weak reflexes corresponding to amorphous transition metal oxides are also observed in the composites. Moreover, depending on the degree and type of heating of the finished fiber, the composition of the metal oxide component can vary from M_xO_y⋅MO(OH)⋅H₂O to M_xO_y. The characteristic signals corresponding to M_xO_y, Mn_xFe_y[Fe(CN)6]_z (Figure 1a, 2Ɵ = 17°, 29°), Mn_xCo_y[Fe(CN)6]_z (Figure 1a, 2Ɵ = 17°, 26°, 36°), and Co_xFe_y[Fe(CN)6]_z (Figure 1a, 2Ɵ = 17°, 29°, 36°) have low intensity, which may be associated with the overlap of this phases, the wide amorphous peak of PAN, and the formation of ultradispersed particles with coherent scattering domain sizes of ~2 nm.

The presence of oxides formed during the destruction of transition metal ferrocyanides in the structure of the annealed samples is visible. Also, the annealed samples contain manganese oxides and K_1.33Mn₈O₁₆ (CoMn-PAN and FeMn-PAN) or iron(III) oxide (CoFe-PAN). This finding indicates that these sorbents will extract not only ¹³⁷Cs, which was confirmed earlier [55,56], but also natural radionuclides, as confirmed by the results of the studies presented later in this article.

Based on the established structures, the following sorption mechanisms can be assumed for the obtained sorbents, which will be largely based on ion exchange:

−

Sorption of ¹³⁷Cs:

2nCs⁺ + Me^IIMe^II[Fe(CN)₆] = nMe²⁺ + Me^II_1–nCs_2nMe^II[Fe(CN)₆],

(1)

3nCs⁺ + Me^III_0.8Me^II_0.8[Fe(CN)₆] = nMe³⁺ + Me^III_0.8–nCs_3nMe^II[Fe(CN)₆],

(2)

−

Sorption of radium isotopes (for CoMn-PAN, FeMn-PAN sorbents):

nRa²⁺ + K_1.33Mn₈O₁₆ = 2nK⁺ + K_1.33–2nRa_nMn₈O₁₆,

(3)

−

Sorption of phosphorus isotopes (for CoFe-PAN sorbent):

nPO₄^3- + Fe(OH)₃ = 3nOH^- + Fe(OH)_3–3n nPO₄

(4)

Dissolved forms of thorium isotopes, ⁷Be, ²¹⁰Pb, and ²¹⁰Pb in seawater are represented mainly by neutral hydrolyzed forms, which are concentrated by the physical adsorption mechanism. A small portion of the positively charged forms of these isotopes is concentrated by the ion exchange mechanism.

3.2. Energy-Dispersive Analysis

Using energy-dispersive analysis (Figure 2), an assessment of the distribution of elements on the surface of the obtained composite materials was performed. The established results are presented in

Table 2. Elemental composition of the surface of the obtained sorbents.

Sorbent	Fe, %	Mn, %	Co, %	K, %	C, %	O, %	N, %
CoFe-PAN	26.75 ± 4.20	−	5.44 ± 0.88	3.54 ± 0.47	28.86 ± 5.13	17.24 ± 3.92	18.17 ± 3.05
CoMn-PAN	7.55 ± 1.21	15.23 ± 2.44	4.89 ± 0.65	4.01 ± 0.68	32.97 ± 5.34	15.32 ± 2.63	20.03 ± 2.91
FeMn-PAN	9.68 ± 1.17	33.37 ± 6.02	−	4.41 ± 0.47	32.38 ± 5.46	8.63 ± 1.37	11.53 ± 1.41

. It is shown that during the formation of the ferrocyanide layer, the distribution of all components of the mixture occurs uniformly, regardless of the synthesis conditions. In addition, the obtained data on the distribution of elements on the sorbents’ surface confirm the above assumption that these sorbents will extract not only ¹³⁷Cs but also natural radionuclides.

3.3. Sorption Selectivity

The results of the assessment of the sorption-selective properties of the adsorbents obtained, for solutions with pH6 and a concentration of each studied element of 10 mg/L, are presented in

Table 3. Distribution coefficient (K_d, mL/g) values for the obtained sorbents in the study of sorption selectivity.

Elements		CoFe-PAN	CoMn-PAN	FeMn-PAN
alkaline	Li	˂1	˂1	˂1
	Na	˂1	˂1	˂1
	K	˂1	˂1	˂1
	Rb	13.1 ± 2.7	3.11 ± 0.96	3.12 ± 0.87
	Cs	(1.8 ± 0.4) × 105	(1.3 ± 0.3) × 105	(9.5 ± 0.7) × 104
alkaline earth	Mg	˂1	˂1	˂1
	Ca	˂1	˂1	˂1
	Sr	˂1	163 ± 24	159 ± 18
	Ba	23.0 ± 4.5	592 ± 45	561 ± 62
p-elements	Al	˂1	18.2 ± 3.6	4.46 ±1.22
	Pb	(6.4 ± 0.9) × 103	(9.4 ± 0.3) × 103	(9.1 ± 0.4) × 103
	Bi	(7.9 ± 0.8) × 103	(9.2 ± 0.4) × 103	(9.3 ± 0.2) × 103
d-elements	Cr	16.6 ± 3.1	18.4 ± 0.9	25.6 ± 4.3
	Mn	˂1	˂1	˂1
	Fe	82.8 ± 7.9	189 ± 7	353 ± 38
	Co	˂1	˂1	19.6 ± 1.0
	Ni	2.50 ± 0.51	13.8 ± 1.5	26.0 ± 1.5
	Cu	76.7 ± 9.6	124 ± 26	210 ± 22
	Zn	4.09 ± 1.12	6.65 ± 0.88	6.07 ± 1.96
	Cd	2.65 ± 0.73	4.35 ± 1.02	6.48 ± 2.32

.

The data presented (Table 3) indicate that the obtained sorbents effectively remove Cs: the distribution coefficients’ values exceed 10⁵ mg/L. This is due to the correspondence between the sizes of the input windows of transition metal ferrocyanides and Cs⁺ [45]. Also, the sorbents CoMn-PAN and FeMn-PAN, which contain manganese oxides, exhibit high efficiency in Ba²⁺ recovery. Considering the patterns of change in the properties of elements, we can assume that the distribution coefficient of Ra²⁺ will be higher than that of Ba²⁺: Ca²⁺ < Sr²⁺ < Ba²⁺ < Ra²⁺.

It was shown that the obtained materials containing manganese and iron(III) oxides exhibit high selectivity for Pb²⁺, Bi³⁺, Fe³⁺, and Cu²⁺, which is comparable with the literature data for sorbents based on manganese oxides [70,71,72] and iron(III) oxides [71,73]. This suggests that the materials can also be used for the sorption of ²¹⁰Pb and ²¹⁰Po.

3.4. Study of Sorption from Various Natural Solutions

The results of the assessment of the influence of salt content on the distribution coefficients of the studied elements are presented in

Table 4. Dependence of the distribution coefficients (K_d, mL/g) of Cs, Be, P, and Ba on the salt content of natural waters.

Sorbent	Recovered Element	Natural Solutions (Salinity, ‰)
		Black River (0.4)	Black Sea (18.4)	Pacific Ocean (33.8)
CoFe-PAN	Cs	(2.4 ± 0.5) × 105	(5.6 ± 0.7) × 105	(2.8 ± 0.2) × 105
	Be	(1.0 ± 0.2) × 103	(1.2 ± 0.2) × 103	950 ± 79
	P	(5.3 ± 0.6) × 103	(1.1 ± 0.2) × 104	(8.1 ± 0.4) × 103
	Ba	31.0 ± 3.8	10.1 ± 2.4	˂1
CoMn-PAN	Cs	(1.8 ± 0.4) × 105	(1.8 ± 0.3) × 104	(1.5 ± 0.1) × 104
	Be	916 ± 98	(1.4 ± 0.1) × 103	(1.1 ± 0.2) × 103
	P	˂1	˂1	˂1
	Ba	745 ± 80	521 ± 57	479 ± 52
FeMn-PAN	Cs	(1.3 ± 0.2) × 105	(1.5 ± 0.2) × 105	(1.2 ± 0.2) × 105
	Be	853 ± 57	(1.3 ± 0.2) × 103	982 ± 48
	P	˂1	˂1	˂1
	Ba	684 ± 75	507 ± 64	452 ± 47

. The studies were conducted using natural solutions with additives of natural isotopes of Cs, Be, P, and Ba as an analog of Ra.

It was observed that with an increase in the salinity of the studied natural solution, the distribution coefficients of Cs decrease slightly, which is associated with the high selectivity of the ferrocyanide phase toward cesium ions [45].

The distribution coefficients for beryllium, phosphorus, and barium also decrease slightly with increasing salinity. Moreover, the distribution coefficients for phosphorus and beryllium in river water are lower than the value of these parameters in seawater, which is probably associated with some features of sorption process in seawater, the components of which can act as a salting-out agent, increasing the sorption efficiency of phosphorus and beryllium. In addition, due to the lower pH of river water, the protonation reaction can have an additional competing effect on the sorption of radionuclides [74].

It should be noted that the distribution coefficients for phosphorus, beryllium, barium, and strontium for the sorbents CoFe-PAN, CoMn-PAN, and FeMn-PAN are lower than those of the original sorbents based on polyacrylonitrile fiber and manganese dioxide (PAN-MnO2) and iron(III) hydroxide (Fe-H). This is due to the fact that during the synthesis process, the sorption-active phases of manganese and iron(III) oxides are partially replaced by the phases of transition metal ferrocyanides. As a result, the new sorbent effectively sorbs Cs, but the sorption efficiency for other elements is slightly reduced.

3.5. Testing of the Obtained Sorbents on Large Volumes of Natural Media

To assess the possibility of practical application of materials in expeditionary conditions, a study was conducted on the radionuclides’ sorption from large volumes of seawater in the Black Sea and the Pacific Ocean.

Studies using seawater from the Black Sea (average salinity 18.4 ‰) were conducted during coastal expeditions in the waters of the Heraclean Peninsula, and also partially during the 126th cruise of the collective use center “RV Professor Vodyanitsky” (15 March–7 April 2023) at different seawater flow rates. The results obtained are presented in Figure 3.

It is seen that the obtained sorbents CoMn-PAN and FeMn-PAN effectively recover ¹³⁷Cs, ²²⁶Ra, ²²⁸Ra, ²¹⁰Pb, ²¹⁰Po, ²³⁴Th, and the sorbent CoFe-PAN: ¹³⁷Cs, ²¹⁰Pb, ²¹⁰Po, ^234Th, and cosmogenic ³²P and ³³P at seawater flow rates of 1–4 C.V./min. The cosmogenic radionuclide 7Be is recovered somewhat worse by the obtained sorbents. However, this allows for a more accurate determination of its activity on the second adsorber when using the two-column method. In addition, it should be noted that ¹³⁷Cs is recovered with an efficiency exceeding 90% by the CoFe-PAN sorbent at a seawater flow rate of 1–5 C.V./min and by the FeMn-PAN sorbent at a seawater flow rate of 1–3 C.V./min.

Comparison of the sorption efficiency obtained during studies on the Black Sea water (average salinity of 18.4 ‰) and the sorption efficiency obtained during research in the Northwestern Pacific Ocean (average salinity of 33.8 ‰) during the 71st cruise of the collective use center “RV Akademik Oparin” (4 June–15 July 2024), at a seawater flow rate of 3 C.V./min, are presented in

Table 5. Sorption efficiency of radionuclide by the obtained sorbents (250 L of seawater, 100 mL of sorbent, and flow rate of 3 C.V./min).

Seawater (Salinity, ‰)	Sorbent	Sorption Efficiency, %
		137Cs	7Be	210Pb	210Po	226Ra	228Ra	234Th
Black Sea (18.4)	CoFe-PAN	94.0 ± 2.1	73.7 ± 1.6	89.7 ± 2.0	84.4 ± 1.9	–	–	98.5 ± 2.2
	CoMn-PAN	80.1 ± 1.8	63.6 ± 1.4	94.7 ± 2.1	90.9 ± 2.0	96.1 ± 2.1	98.1 ± 2.2	100
	FeMn-PAN	90.8 ± 2.0	62.7 ± 1.4	92.8 ± 2.0	87.7 ± 1.9	93.2 ± 2.1	95.3 ± 2.1	100
Pacific Ocean (33.8)	CoFe-PAN	93.5 ± 2.1	70.9 ± 1.6	85.3 ± 1.9	82.1 ± 1.8	–	–	95.5 ± 2.1
	CoMn-PAN	77.4 ± 1.7	58.4 ± 1.3	95.0 ± 2.1	88.9 ± 2.0	95.4 ± 2.1	97.1 ± 2.1	100
	FeMn-PAN	85.5 ± 1.9	63.2 ± 1.4	90.1 ± 2.0	84.2 ± 1.9	92.3 ± 2.0	94.2 ± 2.1	97.2 ± 2.1

. It is seen that an increase in the total salt content in seawater has little effect on the sorption efficiency of radionuclides from seawater, thus allowing us to recommend the adsorbent for large-scale use, regardless of the geography of seawater.

Based on the obtained results, methods for the comprehensive determination of artificial ¹³⁷Cs and natural radionuclides for the obtained sorbents were developed. Step-by-step schemes for each method are shown in Figure 4.

According to the proposed methods, the activity of ²¹⁰Pb can be determined by γ-spectrometry [75], but the probability of emitting γ-quanta during its decay is only 4.5%, and the energy of γ-quanta is also low (46.5 keV). As a result, for the γ-spectrometric determination of low activities of ²¹⁰Pb, special detectors are required to register low-energy γ-quanta and a long exposure time (up to several days) [76]. Therefore, the methods propose an alternative method of β-spectrometric determination by daughter ²¹⁰Bi [18] after radiochemical preparation.

Also, among the features of the proposed method for CoMn-PAN and FeMn-PAN sorbents, it is necessary to note the requirement for a hermetically sealed sample in a specified geometry (usually Petri dishes of a certain diameter and height for planar detectors or test tubes of a certain volume for detectors with a well) and a 3-week holding period to achieve equilibrium of ²²⁶Ra with ²¹⁴Pb. The sample volume for this method is from 200 L depending on the time interval between sampling and measurement of samples due to the short half-lives of ⁷Be (T_1/2 = 53.3 days [77]) and ²³⁴Th (T_1/2 = 24.1 days [77]).

Among the features of the proposed method for the CoFe-PAN sorbent, it should be noted that after gamma-spectrometric determination of ⁷Be, ¹³⁷Cs, ²¹⁰Pb, and ²³⁴Th, it is possible to determine either ²¹⁰Po and ²¹⁰Pb or ³²P and ³³P. Simultaneous determination of these radionuclides in one sorbent sample is not possible due to differences in sample preparation. In this case, when determining ³²P and ³³P, the sample volume should be more than 1000 L, which is associated not only with their short half-lives of 14.3 and 25.3 days, respectively [77], but also with their low activities in seawater—only 0.02–0.1 Bq/m³ [53].

The developed methods were studied during expeditionary research in the Black Sea and the Pacific Ocean. The sampling points are shown in Figure 5. The established values of radionuclide activity during the expeditionary research are presented in

Table 6. Parameters of stations and activities of radionuclides in selected samples.

Station Number		1	2	3	4	5	6
Black Sea (spring 2023)
Sorbent		FeMn-PAN	CoFe-PAN	CoMn-PAN	FeMn-PAN	CoFe-PAN	CoMn-PAN
Coordinates	North Latitude	44.64538	44.54168	44.33689	44.24846	44.22053	44.39808
	East Longitude	34.44425	34.53295	34.27128	33.99762	33.66563	33.67864
Activity, Bq/m3	137Cs	8.54 ± 0.80	9.34 ± 0.87	9.23 ± 0.91	9.20 ± 0.95	9.04 ± 0.85	8.68 ± 0.87
	7Be	– *	– *	4.81 ± 0.75	6.02 ± 0.83	5.64 ± 0.71	4.05 ± 0.64
	210Pb	1.22 ± 0.12	1.51 ± 0.15	1.18 ± 0.12	1.39 ± 0.14	1.54 ± 0.15	1.87 ± 0.19
	210Po	0.37 ± 0.04	0.62 ± 0.06	0.40 ± 0.04	0.58 ± 0.06	0.72 ± 0.07	0.87 ± 0.09
	226Ra	1.14 ± 0.11	–	0.75 ± 0.08	1.10 ± 0.11	–	0.88 ± 0.09
	228Ra	0.75 ± 0.08	–	0.52 ± 0.05	0.75 ± 0.08	–	0.67 ± 0.07
	234Th	– *	– *	12.4 ± 1.3	12.6 ± 1.1	13.2 ± 1.0	10.1 ± 1.1
Pacific Ocean (summer 2024)
Sorbent		CoMn-PAN	FeMn-PAN	CoFe-PAN	CoFe-PAN	CoMn-PAN	FeMn-PAN
Coordinates	North Latitude	145.51447	149.73365	149.77839	149.47866	148.79276	146.44370
	East Longitude	46.84548	34.50422	36.66245	41.49080	43.01880	44.32272
Activity, Bq/m3	137Cs	1.05 ± 0.14	1.61 ± 0.16	1.53 ± 0.16	1.13 ± 0.13	1.18 ± 0.15	1.40 ± 0.14
	7Be	– *	– *	– *	4.32 ± 0.68	3.85 ± 0.81	3.32 ± 0.74
	210Pb	0.72 ± 0.07	2.52 ± 0.25	2.21 ± 0.22	1.87 ± 0.19	1.66 ± 0.17	1.02 ± 0.10
	210Po	0.33 ± 0.03	1.16 ± 0.11	1.05 ± 0.08	0.84 ± 0.08	0.80 ± 0.08	0.45 ± 0.05
	226Ra	1.46 ± 0.15	1.82 ± 0.18	–	–	2.02 ± 0.20	2.14 ± 0.21
	228Ra	0.54 ± 0.05	0.24 ± 0.03	–	–	0.31 ± 0.03	0.68 ± 0.07
	234Th	– *	– *	– *	21.1 ± 3.2	20.2 ± 2.5	17.0 ± 1.8

Note: * Measurement of the activity of ⁷Be and ²³⁴Th at these points was not performed due to their short half-life and the large time interval between the time of sampling and the measurement of the sample.

.

The obtained values of activities coincide with the literature data for the Black Sea (¹³⁷Cs [10,69], ⁷Be [12,20], ²¹⁰Pb and ²¹⁰Po [49], ²²⁶Ra and ²²⁸Ra [39], and ²³⁴Th [39,78]) and the Pacific Ocean (¹³⁷Cs [79,80], ⁷Be [81], ²¹⁰Pb and ²¹⁰Po [82,83], ²²⁶Ra and ²²⁸Ra [80,84,85], and ²³⁴Th [86]), indicating the efficiency of the obtained sorption materials and the reliability of the results.

The activities of ²¹⁰Pb, ²¹⁰Po, ²²⁶Ra, and ²³⁴Th in the Pacific Ocean are higher than in the Black Sea, which is associated with a higher concentration of their parent isotope, ²³⁸U, in the Pacific Ocean, which directly depends on the salinity of seawater [78]. Isotopes of ⁷Be, ²¹⁰Pb, ²¹⁰Po, and ²³⁴Th are adsorbed on suspended matter. For the studied water areas, the activity of ⁷Be, ²¹⁰Pb, ²¹⁰Po, and ²³⁴Th increases with distance from the coast and residential areas due to a decrease in suspended matter [39].

The activity of ¹³⁷Cs in the Black Sea reached the values before the Chornobyl accident in 1986, but the activity of ¹³⁷Cs in the Black Sea is still higher compared to other waters of the World Ocean, due to the isolation of the water area and the presence of a secondary source of influx—the waters of the Dnieper River [87]. Although the studied area of the Pacific Ocean experienced unprecedented artificial impacts from the 2011 Fukushima nuclear power plant accident and ongoing radioactive water discharges, mixing and dilution reduced the activity of ¹³⁷Cs to pre-accident levels. Radionuclides from the coastal area adjacent to the nuclear power plant were drawn into the Kuroshio Current and then followed in the current flow in an easterly direction toward the open ocean [88].

4. Conclusions

Thus, according to the results of X-ray diffraction and energy-dispersive analyses, new sorbents based on transition metal ferrocyanides and manganese (CoMn-PAN and FeMn-PAN) or iron(III) (CoFe-PAN) oxides were synthesized. These sorbents have a high potential for application to the recovery of artificial ¹³⁷Cs and natural radionuclides from natural media, as confirmed by the results of a laboratory investigation of the sorbents’ selectivity, research on solutions with different salt contents, and expeditionary studies. The possible mechanisms of radionuclide sorption were described based on the established structures of the obtained sorbents.

The selectivity studying results indicate that the obtained sorbents are selective for Cs, as well as Pb and Bi. The distribution coefficients for Cs are exceed 10⁵ mg/L, while those for Pb and Bi exceed 10³ mg/L. Additionally, the sorbents CoMn-PAN and FeMn-PAN, which contain manganese oxides, are effective in Ba²⁺ (the analog of Ra) recovery. Therefore, considering the patterns of change in the properties of elements, we can assume that the distribution coefficient of Ra will be higher than that of Ba^.

It was shown that with an increase in the salinity of the studied natural solution, the distribution coefficients decrease slightly. Similar results are observed when comparing the sorption efficiency of radionuclides in large volumes of water from the Black Sea (average salinity, 18.4 ‰) and the northwestern Pacific Ocean (average salinity, 33.8 ‰).

Studies of the obtained sorbents during marine expeditions showed that the obtained materials based on transition metal ferrocyanides and manganese oxides CoMn-PAN and FeMn-PAN effectively recover ¹³⁷Cs, ²²⁶Ra, ²²⁸Ra, ²¹⁰Pb, ²¹⁰Po, and ²³⁴Th at a seawater flow rate of 1–4 C.V./min. The sorbent based on transition metal ferrocyanide and iron(III) oxide CoFe-PAN effectively recovers ¹³⁷Cs, ²¹⁰Pb, ²¹⁰Po, and ²³⁴Th, as well as ³²P and ³³P, at a seawater flow rate of 1–4 C.V./min.

Based on the obtained results, methods for the comprehensive determination of artificial ¹³⁷Cs and natural radionuclides for the developed sorbents were proposed. The results of researching the developed methods during expeditionary research in the Black Sea and the Pacific Ocean indicate the effectiveness of the obtained sorption materials and the possibility of their use in real expeditionary conditions.