Impact of Iron-Bearing Fillers on the Mechanical Strength and Chemical Stability of Magnesium Potassium Phosphate Matrices Incorporating Rhenium

Abstract

We report on the study of the immobilization process of non-radioactive rhenium (Re), a chemical analogue of technetium-99 (⁹⁹Tc), in compounds based on magnesium potassium phosphate (MKP), as well as the possibility of enhancing their properties with iron-bearing additives/fillers. Powdered Re₂O₇ was used as the initial Re-containing source. Because of the solubility and high leachability of Tc (VII), which is also volatile at high temperatures, its immobilization for long-term storage and disposal poses a serious challenge to researchers. Taking this into account, low-temperature stabilization technology based on MKP, a cementitious material, is currently considered promising. We prepared experimental specimens based on Re-incorporated MKP matrices and analyzed their microstructure in detail using analytical methods of X-ray diffraction (XRD), Raman spectroscopy, scanning electron microscopy (SEM), and transmission electron microscopy (TEM). Considering that iron-bearing substances can reduce Tc (VII) to the lower-valence form Tc (IV), which is more stable, attention was also paid to evaluate the effect of fillers (Fe₂O₃, Fe₃O₄, Fe, FeS and blast furnace slag (BFS)) on strength, oxidation state, and water resistance (expressed as leaching cumulative concentration). The addition of fillers ensures the formation of denser compounds based on MKP after 28 days of curing under ambient conditions and increases their mechanical strength. The oxidation state of Re and the reduction from Re (VII) to Re (IV) was estimated using X-ray-absorption near-edge structure (XANES) analysis. Considering the Re leaching concentrations from tests using the ANS-16.1 standard in water, enhanced leachability indices (LI) for Re from MKP matrices were determined with the addition of iron-bearing fillers. Overall, the average LI values were greater than the minimum limit, indicating their acceptance for disposal recommended by the U.S. Nuclear Regulatory Commission.

1. Introduction

Technetium (Tc) and the most abundant Tc isotope, ⁹⁹Tc from liquid radioactive waste stored in various tanks (such as at the Savannah River site and Hanford site) and decommissioned wastes, are hazardous radionuclides that pose a serious threat to the environment. The wide variation and complex chemistries of liquid nuclear waste streams containing ⁹⁹Tc create specific challenges on mechanical strength and chemical stability when immobilizing the waste in materials/matrices suitable for long-term storage and disposal.

The high environmental risk associated with Tc is due to its long half-life (2.1 × 10⁵ years) and the high mobility of the oxidized anionic species Tc(VII)O₄⁻ in which technetium can exist once introduced into the environment. Under reducing conditions, Tc (VII) is reduced to Tc (IV) as hydrous Tc oxide (Tc(IV)O₂·nH₂O), which is relatively insoluble and immobile [1,2]. Therefore, if Tc (VII) is reduced to more stable Tc (IV) by natural or synthetic reducing agents, immobilizing ⁹⁹Tc becomes more feasible [3]. Recent work [4,5,6,7,8,9,10,11] has revealed that absorbers of Fe and Fe-bearing oxides are effective matrices using reduction of TcO₄⁻ from Tc (VII) to Tc (IV) via reductive adsorption or precipitation.

For decades, various waste forms, including glass, ceramics, and cementitious materials, have been tested and used for solidification, immobilization, and retention of ⁹⁹Tc, as reviewed in [12,13]. In several studies, rhenium (Re) has been used as a preferred non-radioactive surrogate for ⁹⁹Tc due to its similar speciation, ionic size, and hydration energy [14,15,16,17].

Among various sequestration methods, since the volatilization and boiling points of ⁹⁹Tc are 311 °C and 900 °C [18], respectively, low-temperature immobilization processes (cementing) are of the greatest interest. However, cementation is complicated by the fact that technetium in a cement compound in the oxidized form Tc (VII) is easily leached [19]. One approach to improve technetium retention is to use a reducing grout waste form, for example, when blast furnace slag and fly ash are added [20].

In the last few decades, Portland cement has been the primary cementitious material in various fields of application. However, its production involves significant CO₂ emissions [21]. In light of the foregoing facts, there would appear to be little doubt that alternative binders less aggressive to the environment must (at least partially) replace Portland cement [22].

Magnesium phosphate cements represent a class of inorganic “low-CO₂” cements that have garnered considerable interest in recent years for their superior performance in environmental protection and energy conservation compared to traditional Portland cement and have diverse applications in the construction and engineering sectors [23]. One of their most interesting applications is in the immobilization of radioactive waste [24]. Magnesium phosphate cements based on the (MgO–PO₄³⁻–H₂O) system, are mainly formulated from calcined magnesium oxide, phosphate, and retarder and other mineral admixtures in certain proportions, representing a cementitious system different from ordinary cement [25].

Magnesium potassium phosphate (MKP) cement, one of the environmentally friendly binders with broad prospects, is a clinker-free acid–base cement in which mechanical strength development is a direct result of the rapid formation of cementitious hydrogel-type products [26,27]. The chemical reaction resulting in the formation of MKP cement binder (Equation (1)) is based on the dissolution of MgO and KH₂PO₄ reacting in solution to form magnesium potassium phosphate, which is isostructural to struvite (NH₄MgPO₄·6H₂O) [28] and is naturally cementitious [29].

MgO + KH₂PO₄ + 5H₂O → MgKPO₄·6H₂O + Q

(1)

The reaction product is MKP, struvite K (magnesium potassium phosphate hexahydrate MgKPO₄·6H₂O), which sets at room temperature to form a crystalline material [28,30,31]. The MKP cement-based waste form can be used to treat solids, liquids, and sludges by chemical immobilization, microencapsulation, and/or macroencapsulation [30]. When compared to conventional Portland cements, MKP cement is a potential magnesia-based cementitious material that has advantageous properties, including near-neutral pH, low water demand, low drying shrinkage, and high early compressive strength [32]. In references [29,33,34], it was demonstrated that fillers/aggregates of iron/iron oxides can improve the compressive strength and water resistance of a solid MKP-based matrix. MKP cement, with its strong and dense matrix, has excellent pollutant binding capacity, making it a candidate for macroencapsulation.

An MKP-based compound called Ceramicrete, described as a chemically bonded phosphate ceramic (CBPC), is a relatively new engineering material developed at Argonne National Laboratory to stabilize radioactive and hazardous waste streams [31,35,36]. MKP cements have been widely proposed and studied to treat various real nuclear wastes in the United States and Russia, including waste containing technetium radioactive elements [30,37,38,39]. In [40], Ceramicrete, was used to solidify liquid waste containing ⁹⁹Tc. SnCl₂ (2–3 wt.% of binder mix) as a reducing reagent for the reduction of Tc (VII) to Tc (IV). To improve the structural integrity of the final waste form, fly ash (38 wt.% of the total content of the binder powder mixture) was added. These waste forms demonstrated compressive strength of approximately 30 MPa and were highly resistant to aqueous media. In [41], specimens of Ceramicrete composition containing goethite (α-FeO(OH)) were proposed and tested for the ability to stabilize residual waste containing ⁹⁹Tc from the C-202 tank at Hanford. Leaching tests in water showed a positive effect: the addition of goethite to the stabilization matrix based on Ceramicrete reduced the release of technetium from 29% to 5.8%–6.3% of the total. Reference [42] is devoted to the selection of modifiers of various types to increase the stability of technetium in MKP during the immobilization of technetium-containing waste, in which various fillers were considered, including iron-containing ones.

In view of the above, as well as having experience in conducting joint research by partners from the USA, Ukraine, and the Czech Republic on studying magnesium potassium phosphate materials [43,44,45], we demonstrate the process of low-temperature stabilization of Re, a Tc surrogate, in MKP-based matrices. Our research was aimed at exploiting the synergy between an MKP-based matrix containing Re and additives (Fe₂O₃, Fe₃O₄, Fe, FeS, and BFS), which can improve its physical properties and chemical stability.

2. Results and Discussion

2.1. Characterization of MKP Matrices Incorporating Rhenium

The phase composition of all investigated MKP–Re specimens was studied using XRD analysis. XRD patterns of MKP-based specimens without and with the addition of Re₂O₇ are shown in Figure 1.

As can be seen in Figure 1a and based on the calculated data on the quantitative content of phases in accordance with the standard Rietveld method, the MKP compound consists of two phases—K-struvite (MgKPO₄·6H₂O) 77% and periclase (MgO) 23%—which indicates the presence of unreacted particles of MgO powder. The lattice parameters of K-struvite are presented in

Table 1. Unit cell parameters of the MKP specimens.

Specimen	Unit Cell Parameters
	a/nm	b/nm	c/nm	V/nm3
MKP	0.68751 (2)	0.61631 (2)	1.10942 (3)	0.470082
MKP–Re	0.68795 (1)	0.616512 (1)	1.10979 (1)	0.470695

.

The addition of 5% Re₂O₇ leads to changes in the phase structure of MKP: in addition to K-struvite (77%) and periclase (16%), a new phase appears—potassium perrhenate or potassium rhenium oxide (KReO₄) with a content of 7% (Figure 1b). In this case, the lattice parameters of K-struvite slightly increased with a corresponding increase in the unit cell volume (ΔV = 0.13%) compared to pure MKP. This fact may indicate partial inclusion of rhenium in the crystal lattice of K-struvite, given its ability to undergo numerous substitutions [46]. Quantitative analysis using the Rietveld method revealed a decrease in the amount of unreacted MgO from 23% to 16%, which is associated with an increase in the acidity of the medium due to the dissolution of Re₂O₇. However, the amount of K-struvite did not change significantly.

As is known, in the process of obtaining the MKP compound according to Equation (2), KH₂PO₄ is mixed with water and its further dissolution occurs, in accordance with Equation (3).

KH₂PO₄ → K⁺ + H₂PO₄⁻

(2)

In turn, the added Re₂O₇ powder also dissolves in the aqueous medium, forming ion ReO₄⁻. Further, because of the reaction (Equation (3)), KReO₄ is formed.

K⁺ + ReO₄⁻ → KReO₄

(3)

The Raman spectra of pure MKP, MKP with the addition of 5% Re₂O₇ (MKP–Re), and initial rhenium oxide powder are shown in Figure 2. The Raman spectra of MKP and MKP–Re specimens clearly shows peaks at 228, 426, 462, 562, and 943 cm⁻¹, characteristic of K-struvite. Also, Raman bands are observed in the 3000-wavenumber range, corresponding to stretching vibrations of water [47]. The peak near 295 cm⁻¹ corresponds to MgO [48]. These data confirm the results of XRD analysis on the presence of the main phase, K-struvite, in the MKP matrix, as well as the presence of MgO (Figure 1).

The remaining peaks at 50, 333, 347, 892, 919, and 959 cm⁻¹ in the Raman spectrum of the MKP–Re specimen belong to potassium perrhenate (KReO₄), the presence of which is also indicated by XRD data (Figure 1b). In [49], the following characteristic peaks for KReO₄ are given: 49, 335, 893, 910, and 969 cm⁻¹. As can be seen in

Table 2. Raman spectroscopy data for K-struvite and KReO₄.

Raman Modes of Struvite (cm−1)		Raman Modes of KReO4 (cm−1)
Observed	Ref. [47]	Observed	Ref. [49]
		50	49
228	228.5
		333	335
		347	-
426	427.5
462	462.8
562	564.1
		892	893
		919	910
943	941.8
		959	969

, the Raman peaks of KReO₄, found in the MKP–Re composition, differ from the values given in this work by the presence of a peak at 347 cm⁻¹ and a shift of ~10 cm⁻¹ of the peaks at 919 and 959 cm⁻¹.

The detected peaks in the Raman spectrum for Re₂O₇ powder—332, 793, 826, 988, and 1000 cm⁻¹—are close to the peaks for crystalline Re₂O₇ determined by the authors of reference [50]. The absence of the peaks characteristic of Re₂O₇ in the Raman spectrum for the MKP–Re compound indicates that Re₂O₇ has completely dissolved and reacted to form KReO₄.

Fractured sections of freshly broken pieces of MKP–Re after 28 days of curing were analyzed using SEM. Many light particles surrounded by a dark matrix were detected (Figure 3a).

EDS analysis of the light particles (spectrum 1, Figure 3b) shows the presence of stronger peaks of Re, K, and O, indicating that these particles are composed of KReO₄. Small, light KReO₄ particles (no more than 5 μm) are characterized by a broad size distribution, are unevenly distributed in the matrix, and are prone to aggregation, which is consistent with reference [51]. Spectrum 2 (Figure 3c) of the matrix material shows intense peaks of Mg, K, P, and O, indicating the presence of magnesium potassium phosphate (K-struvite MgKPO₄·6H₂O).

Based on the elemental maps (Figure 4), uniform distribution of K, Mg, and P is observed throughout the matrix, consistent with the formation of crystalline K-struvite as the main binding phase of the system. The presence of areas (or clusters of Mg-containing particles) with elevated Mg content is associated with the presence of embedded unreacted MgO particles, which is consistent with the XRD data in Figure 1b.

The solid MKP–Re specimen was ground into particles and analyzed using a transmission electron microscope (Figure 5).

In area 1, a dark, spherical particle ~150 nm in size consisting of KReO₄ is clearly visible. This is indicated by the presence of stronger peaks of K, Re, and O in the EDS spectrum of region 1. Peaks of Au and C are associated with target materials containing specimens for TEM analysis. As can be seen, the KReO₄ particle is associated with particles of the MKP matrix, which are combined into agglomerates. Large agglomerates ~500 nm in size include small particles of the MKP matrix, 10–30 nm in size. In area 2, EDS data indicate the presence of the major components of K-struvite, namely Mg, K, P, and O.

Thus, using XRD, Raman spectroscopy, SEM/EDS, and TEM/EDS analysis of the MKP-based compound, it was established that in the case of incorporation of powdered Re₂O₇ oxide (5 wt.%), the resulting matrix contains, in addition to K-struvite and unreacted MgO, potassium perrhenate KReO₄. SEM and TEM observations showed that KReO₄ was present in the form of particles of various sizes, embedded in the MKP matrix.

2.2. Properties of MKP-Based Matrices Incorporating Re Obtained Using Iron-Bearing Additives

2.2.1. Compression Strength

Mechanical strength tests of MKP-based specimens with different Fe-containing fillers were carried out after 1, 7, and 28 days of curing, and the results are presented in Figure 6. For all specimen proportions, there were three replications of each measurement and their average value was taken as the value of strength. The difference between the maximum deviation of the measured strength value and the average strength is expressed on the graph as the standard deviation.

The mechanical strength of the pure MKP compound without additives was 37.2 MPa. The addition of rhenium oxide led to a decrease in strength due to the formation of KReO₄ (according to XRD analysis, Figure 1b), which is shown in Figure 6 for the MKP–Re specimen (34 MPa). The compressive strength of the MKP-based specimens incorporating 5% Re₂O₇ increased with the addition of fillers for all curing periods. The maximum compressive strength after 28 days reached 49.6 MPa for the specimen with BFS added. The addition of 10 wt.% BFS increased strength by 46% compared to the initial MKP–Re specimen.

Different countries have set their own values for waste acceptance criteria (WAC), namely the minimum compressive strength required for the acceptance of cementitious waste forms at a radioactive waste disposal facility. In the USA, the WAC compressive strength is 500 psi or 3.45 MPa [52], almost the same as the 3.44 MPa in South Korea [53], and in European research projects to develop novel cementitious materials, a value of 5 MPa is currently used [54].

Our findings that BFS additives contribute to the highest mechanical strength of the MKP matrix have been confirmed by other researchers. Gardner et al. [29] conducted a study of the structure and characteristics of MKP-based materials with the addition of 50 wt.% BFS, and a significant increase in the compressive strength after 28 curing period from 22 MPa to 34 MPa was shown. In [55], it was shown that a compressive strength of 21.6 MPa was achieved for an MKP compound with BFS added (28.33 wt.%). According to reference [33], the addition of BFS enhanced the early mechanical performance of MKP mortar due to both the physical filler effect and chemical reaction.

One common explanation for the increased compressive strength of MKP with the addition of BFS is related to the chemical composition of this additive. BFS is rich in reactive forms of SiO₂ and Al₂O₃, which are in an amorphous or slightly crystalline phase, making them readily available for dissolution in the alkaline-salt environment of MKP. As a result of the reaction of silicon and aluminum, additional aluminosilicate phases are formed [29,33]. The secondary products forming in MKP blended with BFS are highly enriched in Al and Si, which could potentially lead to the formation of a potassium aluminosilicate phases that may enhance overall densification and potentially increase the durability of MKP binders. In [56], this reaction made the bridging structure different from the struvite or K-struvite structure formed in the system, which caused more connections between the magnesia particles, and the overall structure was more compact.

The compressive strength values of the MKP specimens presented in Figure 6 exceed the WAC requirements for cement matrices. Unlike the BFS additive, the addition of Fe₂O₃, Fe₃O₄, Fe, and FeS slightly increases the compressive strength of MKP. The authors of this article previously showed [8] that the addition of Fe₂O₃ and Fe₃O₄ to the MKP system does not cause the formation of new crystalline phases and that these oxides do not enter into chemical interaction, but act as inert fillers. Apparently, Fe and FeS also have only a physical effect as fillers on the strength properties of the MKP matrix.

2.2.2. Re Oxidation State

Because leaching of Tc is strongly dependent on its oxidation state [41], it was important to estimate the redox conditions of the Re as Tc surrogate in the MKP matrix for reducing Re from the (VII+) to the (IV+) valence insoluble oxidation state in the presence of the reducing agent, such as Fe₂O₃, Fe₃O₄, Fe, FeS, and BFS. The characteristic Re L₃ edge XANES spectra are illustrated in Figure 7, and the edge positions are presented in

Table 3. Position of Re L₃ edge peaks in XANES spectra.

Specimen	Edge Energy (E0) *
MKP–Re–Fe2O3	10,543.9
MKP–Re–Fe3O4	10,544.0
MKP–Re–Fe	10,544.7
MKP–Re/BFS	10,545.0
MKP–Re–FeS	10,545.3

* Edge energy obtained from the position of the zero crossing of the second derivative.

. In all cases, edge energy is obtained from the position of the zero crossing of the second derivative. The spectra look like high-oxidation-state Re (VII) with some reduced Re. The edge position for Re in the MKP–Re–Fe₂O₃ specimen is 10,543.87 eV, while for MKP–Re–FeS it is 10,545.27 eV, a difference of 1.4 eV. As is known [57], the L₃ edge gives an estimation of the oxidation state. In a general sense, the height of the white line is defined by electron localization, and as the oxidation state changes (partially) to lower oxidation states, the white line becomes less pronounced and the edge position shifts. Thus, the data can be interpreted as Re (VII), and for the specimens with some reduction, then the data show a mix of oxidation states—possibly (VI) and/or (IV) mixed with (VII). The general trend in reduction state is: FeS (most reduced), BFS, Fe, Fe₃O₄, Fe₂O₃ (least reduced).

2.2.3. Leaching Behavior

The data from leaching tests of MKP specimens in accordance with the ANSI/AN16.1 standard [58] are presented as dependencies of the leaching cumulative concentrations (C_c) of Re (Figure 8). Comparing the C_c of Re from pure MKP compound and with fillers, it is clearly evident that the addition of fillers leads to a decrease in its value. This can be explained by the presence of locally reducing conditions for Re (VII) arising from the presence of iron ions in the fillers, as confirmed by XANES data (Figure 7, Table 3). In the case of iron-bearing fillers, Re is leached to the greatest extent from the MKP matrix with the addition of Fe₂O₃. Leaching decreases with the addition of Fe₃O₄, Fe, FeS, and BFS, with the lowest cumulative concentration.

The effective diffusion coefficients (D_e) and the corresponding leachability indices (LI) for Re were calculated using Equations (4)–(6) for the MKP-based specimens and are given in

Table 4. Calculated effective diffusion coefficients (D_e) and leaching indices (LI) for Re.

Specimen	MKP–Re	MKP–Re–Fe2O3	MKP–Re–Fe3O4	MKP–Re–Fe	MKP–Re–FeS	MKP–Re–BFS
De (cm2/s)	1.14 × 10−8	4.36 × 10−9	2.64 × 10−9	1.45 × 10−9	1.10 × 10−9	8.94 × 10−10
LI	8.20	8.62	8.87	9.13	9.26	9.31

.

The obtained data showed that D_e values ranged from 1.14 × 10⁻⁸ cm²/s for MKP without filler to 8.94 × 10⁻¹⁰ cm²/s with the addition of BFS. The XANES results above generally support the hypothesis that reduction reduces leachability. The values of all obtained LIs are above the WAC for radioactive waste forms to be disposed of at controlled sites, which should be at least 6.0 [59]. The higher the LI and the lower the D_e, the worse the leaching of the pollutant from the matrix. It should be noted that higher targets have been set for technetium (⁹⁹Tc) than the 6.0 value, namely LI > 9 ([37], page iv), based on early waste-disposal risk and performance assessment analyses.

Filler additions improve resistance to Re leaching, as evidenced by increased LI values. The addition of hematite (Fe₂O₃) increases LI values from 8.20 to 8.62, while magnetite (Fe₃O₄) increases them to 8.87, although they remain below the target level (LI > 9). LI values > 9 were observed for MKP with additions of metallic iron (Fe), iron sulfide (FeS), and BFS. The presented results clearly demonstrate improved rhenium retention with the addition of BFS (LI = 9.31). This means that Re (and therefore Tc also) is better retained in the MKP matrix with BFS filler than in the other MKP specimens considered. As is known, BFS contains both ferrous ion and sulfides, which are strong reductants for ⁹⁹Tc to enhance the immobilization of ⁹⁹Tc inside waste forms [60]. Both Fe (II) and S (II) become reactive reductants once they dissolve, and they can effectively reduce ⁹⁹Tc that may be present in the pore water to form more stable reduced Tc (IV). The superior reducing properties of both FeS and BFS in our study were confirmed by XANES data (Figure 7, Table 3). It was determined that the FeS additive provides the highest reduction of Re (VII) to Re (IV) compared to other additives, while BFS ranks second in reducing capacity. Furthermore, the addition of BFS provides the highest compressive strength for the MKP specimens (Figure 6), indicating the formation of a denser structure compared to specimens modified with Fe₂O₃, Fe₃O₄, Fe, and FeS.

3. Materials and Methods

3.1. Raw Materials

Magnesium oxide (MgO) and potassium dihydrogen phosphate KH₂PO₄, both ≥98% purity, manufactured by Carl Roth^®, Karlsruhe, Germany, were used to synthesis MKP-based matrices according to Equation (1). The MgO was calcined at 1200 °C for 2 h in air to reduce reactivity. Boric acid (H₃BO₃) with a purity of >99% (provided by Penta Chemicals Unlimited Co., Prague, Czech Republic) was added as a setting retarder. Rhenium oxide powder (Re₂O₇) with a purity of 99.9% (provided by Sigma-Aldrich^®, St. Louis, MO, USA) was used as a source of rhenium.

The modifying Fe-bearing fillers were used during preparation of MKP-based specimens. The following powders were used as fillers: hematite (Fe₂O₃) provided by Sigma-Aldrich^® s.r.o., Prague, Czech Republic; magnetite (Fe₃O₄), provided by Thermo Fisher GmbH, Kandel, Germany; metallic iron (Fe), provided by Sigma-Aldrich^®, Stockholm, Sweden; iron sulfide (FeS), provided by Lachema N. P. Brno, Czech Republic; and BFS, in which the chemical composition is presented as: CaO (25–40) wt.%, SiO₂ (30–40) wt.%, Al₂O₃ (10–20) wt.%, MgO (1–5) wt.%, FeO (1–2) wt.% (provided by Moravia Steel, Trinec, Czech Republic).

The choice of fillers was determined by the desire to improve the properties of the matrices, but also by the fact that iron ions, also present in BFS, are capable of effectively stabilizing ⁹⁹Tc in nuclear waste, reducing it from oxidation state (VII) to (IV), as described in the Introduction.

3.2. Specimen Preparation

In the present study, to obtain MKP-based matrices in accordance with Equation (1), a magnesium oxide (MgO)-to phosphate-(KH₂PO₄) molar ratio (M/P) of 2.25 and a water-to-cement mass ratio (W/C) of 0.5 were used, which were previously proposed for the immobilization of minor actinides [45]. Re₂O₇ (5 wt.%) was introduced into a dry mixture of MgO with KH₂PO₄. The laboratory process for obtaining MKP specimens is shown in Figure 9.

Dry raw components (MgO, KH₂PO₄, Re₂O₇) were initially weighed and then mixed in an Alpine Augsburg (Germany) lab grinding mill machine for 10 min according to the mixing proportions. Then, H₃BO₃ (5% of the MgO and KH₂PO₄ mass) was dissolved in deionized water. The dry component mixture was added to this solution, stirred mechanically for 3 min, and poured into (3 × 3 × 3 cm) plastic molds. These were tightly closed to avoid desiccation and air-cured for 1, 7, and 28 days at room temperature (20–25 °C). Fillers were added at the stage of preparing the dry powder mixture, and the studied basis proportions are presented in

Table 5. Specimen powder proportions.

Batch	Dry Powder Mix Proportion (wt.%)
	MgO + KH2PO4	Re2O7	Fe2O3	Fe3O4	Fe	FeS	BFS
MKP	100	0	0	0	0	0	0
MKP–Re	95	5	0	0	0	0	0
MKP–Re–Fe2O3	85	5	10	0	0	0	0
MKP–Re–Fe3O4	85	5	0	10	0	0	0
MKP–Re–Fe	85	5	0	0	10	0	0
MKP–Re–FeS	85	5	0	0	0	10	0
MKP–Re/BFS	85	5	0	0	0	0	10

.

Each filler (10 wt.%) was added based on references [61,62], in which the authors showed that the addition of various additives at 10 wt.% had a positive effect on increasing the strength of the magnesium phosphate-based compound.

3.3. Experimental Methods

3.3.1. Strength Test

The mechanical strength of MKP-based specimens (3 × 3 × 3 cm) after various curing periods was estimated by measuring the compressive strength with the use of a MEGA II-600 DMI-S (FORM+TEST Seidner & Co. GmbH, Riedlingen, Germany) universal testing machine at a loading rate of 0.25 MPa/s. The average compressive strength was determined by averaging the strength values of three MKP-based specimens.

3.3.2. Leaching Test

To assess the stability of the MKP matrix against leaching, the obtained MKP specimens were subjected to leaching tests according to ANSI/ANS-16.1 in deionized water for five leaching intervals at 1, 4, 14, 28, and 43 days (total 90 days) in a plastic vial at a room temperature. The volume of deionized water was 540 mL. The average effective diffusion coefficient D_e, often expressed in cm²/s, was calculated from the ICP-MS concentration values measured for each leaching interval t_n − t_n−1 using this common expression:

$$D_e=\pi {\left[{\displaystyle \frac{\raisebox{1ex}{$a_n$}\!\left/ \!\raisebox{-1ex}{$A_0$}\right.}{{\Delta t}_n}}\right]}^2{\left(V/S\right)}^{2}T$$

(4)

where V is the volume of the specimen (cm³), S is the geometric surface area of the solid specimen (cm²), a_n is the quantity of the element released from the specimen for the n’th leaching interval, A₀ is the initial quantity of the element in the specimen at the beginning of the first leaching interval, Δt_n = t_n − t_n−1 is the duration of the n’th leaching interval, and T is the leaching time representing the mean time of the leaching interval (s) as follows:

$$T ={\left[{\displaystyle \frac{1}{2}}\left(\sqrt{t_n}-\sqrt{t_{n-1}}\right)\right]}^2$$

(5)

The leachability index (LI) of an element of concern in a material was defined as:

$$LI={\displaystyle \frac{1}{5}}\sum _1^5\left[\log \left({\displaystyle \frac{\beta }{D_e}}\right)\right]$$

(6)

where β is a constant (~1 cm²/s).

The leachability index (LI) of an element is a parameter that suggests the effectiveness of the process for immobilization, and is used to regulate waste conditioning. According to the U.S. Nuclear Regulatory Commission (NRC) standard [59], the WAC for LI that is evaluated based on the ANSI/AN16.1 method should be greater than 6.0 to meet the requirement during the encapsulation process for radioactive waste conditioning.

3.3.3. Microstructural Analyses

The phase composition of MKP-based materials was determined using conventional Bragg–Brentano X-ray powder diffraction (XRD) measurements. Diffraction patterns were collected with a PANalytical X’Pert PRO diffractometer equipped with a conventional X-ray tube (CuK_α radiation, 40 kV, 30 mA, line focus) and a linear position-sensitive detector PIXCel with an anti-scatter shield. Qualitative analysis was performed with the HighScorePlus software package (Malvern PANalytical, version 5.3.1) together with the PDF-5+ database [63]. Quantification of the experimental data was performed with the Rietveld method [64]. The Profex 4.3.6/BGMN 4.2.22 code was used for all calculations [65,66].

X-ray absorption spectroscopy was employed to determine the average oxidation state of Re in the specimens and to probe the local atomic environment around the Re absorber atoms. XANES spectra of the Re series were collected at the 10-ID-B beamline of the advanced photon source [67]. For measurements, each specimen was mixed with BN powder using a mortar and pestle to achieve a homogeneous mixture, which was then loaded into a 19 mm nylon washer and sealed with Kapton tape on both sides. The Re L₃ edge (10,534 eV) XANES spectra were collected in transmission mode, with each spectrum averaged from triplicate measurements. A Re metal foil was simultaneously measured as a reference. The collected spectra were processed using Athena from the Demeter software package (version 0.9.26) [68], running on the IFEFFIT algorithm, to perform reference foil measurement energy alignment and normalization. The absorption edge position (E₀) for the Re L₃ edge was defined as the zero-crossing point of the second derivative of each absorption curve.

The surface morphology and structure of specimens were analyzed by SEM with X-ray energy-dispersive spectroscopy (EDS) using a JSM-6510 microscope with an X-act-10 mm² SDD (JEOL Ltd., Tokyo, Japan) and a Talos F200X (Thermo Fisher Scientific Inc., Waltham, MA, USA) for TEM.

Raman spectroscopy was performed on a DXR Raman spectrometer (Thermo Fisher Scientific Inc., Waltham, MA, USA). The spectral range used to characterize MKP materials was between 0 and 3600 cm⁻¹.

4. Conclusions

In this study, the effects of iron-bearing additives (Fe₂O₃, Fe₃O₄, Fe, FeS and BFS) on the mechanical properties and water resistance of MKP-based specimens containing non-radioactive Re as a chemical analogue of long-lived radionuclide ⁹⁹Tc were investigated. Combining the results of XRD, Raman spectroscopy, SEM, TEM, XANES, and mechanical and leaching tests, the essential findings were as follows.

(1)

The phase composition of the MKP-based matrix without iron-bearing additives changes upon the addition of 5 wt.% powdered Re₂O₇: in addition to K-struvite (77%) and periclase (16%), a new phase appears—potassium perrhenate or potassium rhenium oxide (KReO₄) (7%). In this case, a slight increase in the lattice parameters of K-struvite was observed, which led to an increase in the unit cell volume (ΔV = 0.13%) compared to pure MKP, which may indicate a partial inclusion of rhenium in the crystal lattice of K-struvite, given its ability for numerous substitutions. It was found that KReO₄ particles are characterized by a wide size distribution (up to 5 μm as a maximum) and are embedded in the MKP matrix. Therefore, the K-struvite matrix surrounds the KReO₄ particles and hinders the contact of Re (accordingly for Tc also) with the external environment, ensuring its immobilization in stable/durable waste form.

(2)

The use of iron-bearing additives in the initial powder mixture (MKP + 5 wt.% Re₂O₇) had a positive effect on the mechanical properties of the resulting solid specimens. The measured compressive strengths of the matrices prepared with the addition of 10 wt.% Fe₂O₃, Fe₃O₄, Fe, FeS, and BFS after 28 days of curing were 37.8, 39.4, 37.1, 35.9, and 49.6 MPa, respectively, indicating an improvement over the initial strength (34 MPa) without additives.

(3)

The oxidation state of Re in the presence of the reducing agent, such as Fe₂O₃, Fe₃O₄, Fe, FeS, and BFS, was preliminary revealed. The results of measurements using XANES can be interpreted as Re (VII), as a predominant oxidation state. At the same time, we can talk about a possible reduction process, which was confirmed by the detection of a mix of oxidation states—possibly Re (VI) and/or Re (IV) mixed with Re (VII). The general trend in reduction state is this: FeS (highest), BFS, Fe, Fe₃O₄, Fe₂O₃ (lowest).

(4)

Addition of iron-bearing fillers to the MKP-based matrix improves its water resistance, as evidenced by the results of leaching tests according to the ANSI/AN16.1 standard. The leachability index (LI) for Re from the MKP matrix prepared using the BFS additive was a maximum of 9.31, which meets the target LI > 9. This confirms the effectiveness of the waste form based on MKP with the addition of BFS for the Tc encapsulation process. Further research is ongoing to determine the optimal amount of BFS additive that will provide the greatest positive effect in preventing Re leaching from the MKP matrix.