Introduction

A large amount of high-level radioactive wastes (HLWs), which contain high radioactive nuclides, corrosion products, and fuel clad material, have been produced with the development of nuclear power and the decommissioning of military equipment [1]. The radionuclides in HLW include various fission products (FPs) and actinides (Ans) with multiple oxidation states, which possess long-term radiotoxicity and high corrosivity. The safe disposal and isolation of HLW to prevent the leakage and migration of radioactive nuclides is one of the key issues related to the sustainable development of nuclear power worldwide. The long-term treatment of HLW requires the selection of suitable host matrices which should have low leaching rate and high thermal, chemical, radiation, and mechanical stability under repository conditions. Ceramics, as nuclear waste solidification matrices, have been investigated for many years due to their superior properties compared with glass waste forms in terms of the chemical, thermal, and radiation resistance [2]. At present, some potential ceramics matrices, such as sodium zirconium phosphate (NaZr2(PO4)3) [3,4,5], monazite [6,7], pyrochlore [8], zirconolite [9], and apatite [10], have been extensively concerned during recent years. However, most ceramics matrices are only suitable for accommodating a certain FP or An radionuclide due to the finiteness of their lattice substitution, which limits the application of these ceramics matrices in immobilizing HLW containing various FP and An radionuclides. As to this problem, it may be a feasible approach to utilize the multiphase composite ceramics to concurrently immobilize FP and An with diverse valences and ionic radius.

In the last few decades, the research on the multiphase composite ceramics for the HLW immobilization mostly focuses on titanate-based ceramic waste forms. Among these composite ceramics, the multiphase titanate-based ceramics targeting hollandite, zirconolite/pyrochlore, and perovskite phases have been extensively investigated on the preparation [11], irradiation stability [12], and chemical durability [13,14]. It was reported that Cr addition can facilitate the formation and stability of a Cs-containing hollandite phase in the kind of multiphase ceramics prepared by melt process [11]. Clark et al. revealed the chemical durability of multiphase ceramic designer waste forms could be improved by adjusting the phase proportion. It was found that the fractional Cs release decreased as the amount of hollandite phase increased; however, the zirconolite and pyrochlore phases did not significantly contribute to the elemental release from the hollandite phase [14]. In addition, hollandite–perovskite composite ceramics could be also considered a customized host matrix for immobilization of the separated Cs and Sr from HLW streams, exhibiting excellent chemical durability [15]. Besides, zirconolite–sphene composite ceramics were explored for immobilizing tetravalent actinide U, which was incorporated in the Zr site of zirconolite and in the Ca site of sphene [16]. Similarly, tetravalent actinide Ce could be simultaneously immobilized in Zr site of zirconia phase and zircon phase in 0.2Zr1−xCexO2/Zr1−yCeySiO4 ceramics [17]. The SiC–MgAl2O4 composite ceramic was intended to immobilize 14C and other high-level radioactive nuclides, which was reported by Teng YC [18,19]. It was found that MgAl2O4 phase, as a matrix for immobilizing long-lived nuclides, could improve the sintering of SiC ceramics, and the SiC–MgAl2O4 composite ceramics presented good physical and chemical stability. Thus, the multiphase composite ceramics can provide the flexibility, selectivity, and high loading for radionuclides by combining with the structural advantages of their respective crystalline phases. Considering the better practical application and performance of composite ceramics than those of single-phase ceramics, it is essential to develop new composite ceramics to immobilize complex high-level radionuclides with diverse valences and ionic radius [20].

NZP family compounds, whose crystalline structure is composed of three-dimensional hexagonal framework of PO4 tetrahedra sharing corners with ZrO6 octahedra [21], are well known for its ionic conductivity, low thermal expansion, and flexibility on ionic substitution [22,23]. As a potential host, the NZP-type ceramics are widely studied for immobilizing FP radionuclides (90Sr/ 137Cs) due to its abundant ionic substitution. For instance, Hashimoto et al. investigated the immobilization process of Cs and Sr into HZr2(PO4)3 by using an autoclave [24]. Pet’kov et al. reported the thermophysical properties and hydrolytic stability of Sr0.5Zr2(PO4)3 for immobilizing 90Sr [25]. Meanwhile, as the host for FP radionuclide (90Sr/137Cs) immobilization, NZP family presented high chemical stability, as well as radiation resistance properties [3,26]. Monazite, an anhydrous monoclinic rare-earth orthophosphate mineral (REPO4) consisting of distorted PO4 tetrahedra and REO9 polyhedra, is a high-profile matrix for the disposal of trivalent and tetravalent An radionuclides due to the flexibility of REO9 polyhedral structure. This kind ceramic waste forms present the excellent mechanical, thermal and chemical properties, and high radionuclide loading capacity as well as superior radiation resistance [27]. Besides, it was also previously reported that the monazite-zirconium phosphate–type composite ceramics worked well as a matrix to immobilize HLW, especially to decrease the leaching rate of alkali and alkaline ions [28]. Similarly, the presence of monazite did not deteriorate the aqueous stabilities of NZP ceramics [29]. So NZP-monazite-type composite ceramics would be an applicable matrix for immobilizing complex HLW containing FP and An with multiple valence, on which there are few reports so far [20,30]. In particular, the proportion of NZP and monazite phases in NZP-monazite-type composite ceramics can be adjusted for the requirement of simultaneously immobilizing FP and An radionuclides of indefinite quantity. Thus, by comparison with other ceramic waste forms previously reported, the novel NZP-monazite-type composite ceramics proposed in this work may have potential application advantages in immobilizing HLW.

In this work, Sr and Ce from the raw materials of Sr(NO3)2 and CeO2 were introduced as the surrogates for the fission nuclide 90Sr and variable valence actinide nuclide (trivalent valence and tetravalent valence actinide), respectively. The chemical formula of NZP-monazite-type composite ceramics was designed as (1 − x)Sr0.5Zr2(PO4)3 − xCePO4 (x = 0, 0.2, 0.4, 0.6, 0.8, and 1.0). The series of Sr0.5Zr2(PO4)3–CePO4 composite samples were in situ prepared by one-step microwave sintering technique in this work. The synthesis of crystalline phases and the sintering of composite ceramics were concurrently achieved in microwave processing of the in situ preparation, which was different from the conventional preparation process of composite ceramics [31,32]. The evolution of phase composition, microstructure, physical properties, and chemical stability of the as-prepared samples were systematically investigated. It was aimed to discuss the feasibility of (1 − x)Sr0.5Zr2(PO4)3–xCePO4 composite ceramic waste forms for simultaneously immobilizing FP and An. In addition, the valence state of variable valence Ce in the as-prepared composite ceramics was also analyzed.

Experimental

Preparation of samples

(1 − x)Sr0.5Zr2(PO4)3–xCePO4 composite ceramics (x = 0, 0.2, 0.4, 0.6, 0.8, and 1.0) were prepared by using Sr(NO3)2 (99.0%) in appropriate amounts and CeO2 (99.99%), ZrO2 (99.0%), and NH4H2PO4 (99.0%) in high purity as the starting materials and by employing one-step microwave sintering technique. To be specific, the starting materials were weighted by stoichiometric ratios and then thoroughly ball-mill mixed and ground. The mixture powders were preheated at 600 °C holding for 8 h in a muffle furnace to decompose Sr(NO3)2 and NH4H2PO4 for emission of nitrogen dioxide, ammonia, and water vapors [33]. Subsequently, the preheated powders were again ball-milled together with adding the sintering aid of 1.0 wt% ZnO powders. After being dried and passed through a 200-mesh sieve, the resultant powders were pressed into Ø12 × 2 mm tablets by cold isostatic pressing under 200 MPa. Finally, the green compacts were microwave sintered at different temperatures (1050 ~ 1200 °C) for 2 h with a heating rate of 5 °C/min by a multi-mode of 2.45-GHz, 4-kW commercial microwave workstation (MobileLab Workstation, Tangshan Nayuan Microwave Thermal Instrument Manufacturing Co. Ltd., China). Thus, the series of composite ceramics waste forms for immobilizing radionuclides Sr and Ce were obtained.

Characterization

Thermogravimetry–differential scanning calorimetry (TG-DSC, SDT Q600, TA, USA) analysis of the preheated powders was performed under nitrogen atmosphere with a heating rate of 10 °C/min. The phase evolution with varying components was examined by powder X-ray diffraction (XRD, DMAX1400, Rigaku Inc., Japan) using Cu-Kα radiation. The microstructures and chemical compositions of the samples were acquired by scanning electron microscopy (SEM) attached with an energy-dispersive spectrometer (EDS) (SEM–EDS, Hitachi TM-4000, Japan). The binding energy and oxidation state of the composite ceramics (x = 0.2–0.8) were analyzed by X-ray photoelectron spectroscopy (XPS, ESCALAB 250Xi, Thermo Scientific, USA) using Al-Kα radiation with a step size of 0.05 eV. The ion-electronic charge compensation system was used to neutralize the charge of the sample in the experiments. All peaks were calibrated against the C1s peak at 284.8 eV. The XPS spectra were analyzed using the software XPSPEAK41. The bulk density (ρ) of the as-prepared samples was obtained via the Archimedes method. The relative density was calculated using the formula of ρ/ρ0 × 100%, in which ρ0 denotes theoretical density. The specific calculation process was carefully introduced in our previous work [34]. The Vickers hardness (HV) of the samples was measured by the indentation fracture technique utilizing a Vickers 136°-diamond indenter (HVS-1000Z, Shanghai Wanheng Precision-instrument Co., Ltd., China) with 1 kgf load for 10 s. The HV results were averaged from three samples and each sample was tested five times, whose value was dimensionsless unit and HV1 refers to the HV test with 1kgf load.

Chemical stability

The chemical stability of typical (1 − x)Sr0.5Zr2(PO4)3 − xCePO4 (x = 0, 0.4, 0.6, and 1.0) samples was investigated by the Product Consistency Test (PCT) which is a standard static leaching method from ASTM C1285-14 [35]. The sintered samples were smashed and screened out the powders between 100 and 200 mesh, and then washed by absolute ethanol and ultrapure water. A certain proportion of the as-obtained powders and deionized water was put into a well-sealed hydrothermal reaction kettle (304 stainless steel shell and teflon container), and held at 90 °C for 7 days. The concentration of Sr, Ce, Zr, and P in the leaching solution was measured by inductively coupled plasma optical emission spectroscopy (ICP-OES, Thermo iCAP6500, Thermo Fisher, USA). The normalized elemental leach rates (LRi) were calculated according to the following formula [36,37]:

$${LR}_{i}=\frac{{C}_{i}\bullet V}{{f}_{i}\bullet {A}_{s}\bullet \Delta t}$$
(1)

where Ci is the concentration of element i in the leaching solution (g/m3), V is the volume of the leaching solution (m3), fi is the mass fraction of element i in the sample (wt%), As is the geometric surface area of the sample (m2), and Δt is the leaching time (day).

Results and discussion

TG-DSC analysis

To determine the suitable sintering temperature of the composite ceramics, the TG-DSC analysis of the typical preheated powders corresponding to 0.6Sr0.5Zr2(PO4)3–0.4CePO4 component is performed, whose curves are shown in Fig. 1. As seen in Fig. 1, the total weight loss of the preheated powders is about 1.8% and 5.1% in the ranges of room temperature to 350 °C and 350 to 400 °C, respectively, which is associated with the dehydration of free water and bound water of the powders. Meanwhile, the endothermic peak at 346.25 °C on the DSC curve is caused by the dehydration of bound water. The weight loss of approximately 0.8% at the temperature range of 400–650 °C may be attributed to incomplete decomposition of raw material powders after preheating treatment. Clearly, three exothermic peaks at 743.7, 850.4, and 1029.2 °C are observed on the DSC curve, but no obvious weight loss at the corresponding temperatures is found from TG curve. It is inferred that CePO4 phase and Sr0.5Zr2(PO4)3 phase are synthesized at 743.7 °C and 850.4 °C, respectively. The sintering densification of the composite ceramics may be proceeded at 1029.2 °C, which also corresponds to the work temperature of ZnO sintering aid [[[24]]]. Thus, the sintering temperature ranging from 1050 to 1200 °C was preliminarily chosen on the basis of the TG-DSC analysis.

Fig. 1
figure 1

TG-DSC curves of the preheated powders of 0.6Sr0.5Zr2(PO4)3–0.4CePO4 composition

Full size image

Evolution of the phase composition

The powder XRD patterns of 0.6Sr0.5Zr2(PO4)3–0.4CePO4 sample after microwave sintering at 1050–1200 °C for 2 h are shown in Fig. 2a. From Fig. 2a, it can be seen that the 0.6Sr0.5Zr2(PO4)3–0.4CePO4 sample sintered at different temperatures all shows the characteristic diffraction peaks of crystalline Sr0.5Zr2(PO4)3 phase (JCPDS PDF#33–1360) and CePO4 phase (JCPDS PDF#32–0199). The crystallization peak strength and peak width slightly change as the sintering temperature varies from 1050 to 1200 °C. It is indicated that the two crystalline phases are relatively stable and the composite ceramics can be in situ prepared in a wide temperature range. Considering the energy consumption during sintering process, the (1 − x)Sr0.5Zr2(PO4)3–xCePO4 samples (x = 0–1.0) for subsequent analyses were prepared at the lowest sintering temperature (1050 °C) in this work. The powder XRD patterns of the samples sintered at 1050 °C for 2 h are presented in Fig. 2b. As seen in Fig. 2b, the (1 − x)Sr0.5Zr2(PO4)3–xCePO4 samples show the characteristic diffraction peaks of pure Sr0.5Zr2(PO4)3 phase and pure CePO4 phase for x value are 0 and 1.0, respectively. Besides, the phase composition of the composite ceramic samples (x = 0.2–0.8) only contains Sr0.5Zr2(PO4)3 phase and CePO4 phase and no other crystalline phases are found. As expected, the diffraction peak intensity of Sr0.5Zr2(PO4)3 phase just decreases regularly as the x value raises, whereas that of CePO4 phase gradually increases. Based on the XRD results, it is confirmed that there is no noticeable chemical reactions between Sr0.5Zr2(PO4)3 and CePO4 phases, which reflects the stability and compatibility of the two crystalline phases.

Fig. 2
figure 2

XRD patterns of (1 − x)Sr0.5Zr2(PO4)3–xCePO4 composite ceramics: a x = 0.4 sample after microwave sintering at different temperatures for 2 h; b x = 0–1.0 samples after microwave sintering at 1050 °C for 2 h

Full size image

Analysis of valence state of Ce

Ce is a good surrogate for actinides radionuclide Pu with multiple oxidation states. In order to ascertain the valence states of Ce element in the prepared composite ceramics, XPS test was performed on (1 − x)Sr0.5Zr2(PO4)3–xCePO4 composite ceramic samples (x = 0.2–0.8), whose high resolution Ce 3d XPS spectra are shown in Fig. 3. It is well known that the electrons of Ce atom have spin–orbit interaction, causing the energy split. Thus, the XPS spectra consist of two pairs of spin–orbit split peaks and associated shake-up satellites (denoted as u0/u0′ and u1/u1′, respectively). The components u0 and u1 refer to 3d5/2, while u0′ and u1′ stand for 3d3/2, which all indicates the presence of Ce3+. As shown in Fig. 3, the peaks at 882.0, 885.5, 900.1, and 903.8 eV are clearly seen, corresponding to u1, u0, u1′, and u0′, respectively. However, the peak around 916.3 eV, which is considered a marker for the presence of Ce4+ [38], is almost unobserved. Then, it is verified that the valence state of Ce in the as-prepared composite ceramics mainly exists in trivalent state, which is in accord with the valence state of Ce in CePO4 phase.

Fig. 3
figure 3

High-resolution Ce 3d XPS spectra of (1 − x)Sr0.5Zr2(PO4)3–xCePO4 composite ceramics

Full size image

Micromorphology analysis

Figure 4 shows the backscattered electron SEM images of the fracture surfaces of (1 − x)Sr0.5Zr2(PO4)3–xCePO4 samples. It is clearly found that all samples present a well-densified microstructure. By comparison, the pure Sr0.5Zr2(PO4)3 sample (x = 0) presents transgranular fracture (Fig. 4a), while the pure CePO4 sample (x = 1) shows the characteristic of intercrystalline fracture (Fig. 4f). It implies that the sintering densification of Sr0.5Zr2(PO4)3 sample is superior than that of CePO4 sample. Nonetheless, the CePO4 sample still possesses a dense microstructure with closely packed crystal grains. As shown in Fig. 4b–e, the composite ceramics (x = 0.2–0.8) also exhibit the characteristic of transgranular fracture and the crystalline grains of the samples show different brightness and sizes, which can be clearly differentiated by the contrast of their morphologies. According to the above XRD results and the imaging features of SEM, the bigger and darker grains might be related to Sr0.5Zr2(PO4)3 phase, and the smaller and brighter grains might be correlated with CePO4 phase. Obviously, the two phases are evenly distributed in the composite ceramics. Additionally, the crystalline sizes of the composite ceramic samples are all obviously smaller than that of pure Sr0.5Zr2(PO4)3 sample and CePO4 sample, though the CePO4 grain size gradually grows with the increasing x value. That is to say, the existence of CePO4 phase brought about the grain refinement and could facilitate the densification of the composite ceramics.

Fig. 4
figure 4

SEM images of (1 − x)Sr0.5Zr2(PO4)3–xCePO4 composite ceramics: a x = 0; b x = 0.2; c x = 0.4; d x = 0.6; e x = 0.8; f x = 1.0

Full size image

Figure 5 displays the SEM–EDS elemental distribution mapping of the representative 0.6Sr0.5Zr2(PO4)3–0.4CePO4 sample. To be specific, Fig. 5a represents the polished surface SEM image of the sample and Fig. 5b–f show the results of EDS elemental mapping, in which the chemical compositions of the sample are composed of Ce, Sr, Zr, P, and O. It is clear that Ce elemental distribution mapping (Fig. 5b) is just correlated with the brighter CePO4 phase in Fig. 5a, while Sr elemental distribution mapping (Fig. 5c) is associated with the darker Sr0.5Zr2(PO4)3 phase. Besides, on the basis of the elemental atomic ratios listed in the table (embedded in Fig. 5g), the calculated atomic ratios of P/O and Sr/Ce/Zr are 14.62/42.87 and 4.85/10.26/27.41, respectively, which are very close to the theoretical stoichiometric atomic ratios of [PO4] (1:4) and 0.6Sr0.5Zr2(PO4)3–0.4CePO4 formula (Sr/Ce/Zr is about 3:4:12). The SEM–EDS results further proved that Sr and Ce radionuclides were independently incorporated into the crystalline lattices of Sr0.5Zr2(PO4)3 and CePO4 phases in the composite ceramics waste forms.

Fig. 5
figure 5

SEM–EDS elemental mapping of 0.6Sr0.5Zr2(PO4)3–0.4CePO4 composite ceramics

Full size image

Density and Vickers hardness analyses

To further investigate the physical properties of the as-prepared composite ceramics, the density and Vickers hardness of (1 − x)Sr0.5Zr2(PO4)3–xCePO4 samples are tested and their results are given in Fig. 6a and b, respectively. It is found in Fig. 6a that the bulk density of (1 − x)Sr0.5Zr2(PO4)3–xCePO4 sample (x = 0–1.0) increases gradually as x value rises, from 3.27 g/cm3 of pure Sr0.5Zr2(PO4)3 sample (x = 0) to 4.96 g/cm3 of pure CePO4 sample (x = 1.0), which well conforms to the compound effect regulation. For Sr0.5Zr2(PO4)3–CePO4 composite samples, the calculated relative densities are all greater than 95%, even up to 97.7%. By comparison, the relative densities of all composite samples are almost higher than that of pure CePO4 sample and also comparable with that of pure Sr0.5Zr2(PO4)3 sample. The densification analyses are in good agreement with SEM results. Just as expected from the above SEM results, the dense NZP-monazite-type composite ceramics waste forms can be achieved by on-site preparation in this work. Figure 6b shows that the Vickers hardness of pure Sr0.5Zr2(PO4)3 sample (x = 0) is 624 HV1, and the Vickers hardness of the (1 − x)Sr0.5Zr2(PO4)3–xCePO4 composite sample gradually increases as the x value goes up from 0.2 to 0.8. When the x value is 0.8, the 0.2Sr0.5Zr2(PO4)3–0.8CePO4 sample accomplishes the highest hardness of 774 HV1. However, pure CePO4 sample (x = 1.0) only presents the Vickers hardness of 723 HV1. Notably, the introduction of CePO4 phase not only promotes the densification of the Sr0.5Zr2(PO4)3–CePO4 composite ceramics but also improves their Vickers hardness. According to the above analyses, it is inferred that the introduction of CePO4 phase could refine Sr0.5Zr2(PO4)3 grains and then strengthen grain interfaces of the composite ceramics without decreasing their densification. Thus, as the CePO4 phase increases, the grain refinement effect and the good densification would jointly improve the hardness of Sr0.5Zr2(PO4)3–CePO4 composite ceramics, which is beneficial for the shock resistance of Sr0.5Zr2(PO4)3–CePO4 composites as nuclear waste form.

Fig. 6
figure 6

Density and Vickers-hardness of (1 − x)Sr0.5Zr2(PO4)3–xCePO4 composite ceramics

Full size image

Chemical stability of composite ceramics

The typical (1–x)Sr0.5Zr2(PO4)3–xCePO4 (x = 0, 0.4, 0.6, and 1.0) samples were conducted by PCT leaching test at 90 °C for 7 days to evaluate their chemical stability. The normalized leaching rates LRi (i = Sr, Ce, Zr, and P) of the samples are presented in Table 1. It is evident that the normalized elemental leaching rates of all samples are quite low, which may be related to the stable crystalline structures of NZP family and monazite as well as the high densification of the samples. By contrast, the values of LRCe and LRZr are significantly lower than those of LRSr and LRP, which is mainly due to the more stable and stronger Ce–O bond of [CeO9] polyhedron in CePO4 crystalline structure and Zr–O bond of [ZrO6] octahedron in Sr0.5Zr2(PO4)3 crystalline structure. In addition, the LRSr (< 4.142 × 10−4 g·m−2·day−1), LRZr (< 2.191 × 10−7 g·m−2·day−1), and LRP (< 1.014 × 10−4 g·m−2·day−1) of the composite ceramics (x = 0.4, 0.6) are all lower 1–2 orders of magnitude than that of the monophase ceramics, especially for Sr0.5Zr2(PO4)3 ceramics. The above LRi results of composite ceramics indicated that the introduction of CePO4 phase could improve the chemical stability of Sr0.5Zr2(PO4)3 phase in the composite ceramics. It may be attribute to the uniform and grain-refined microstructure as well as the excellent densification of the Sr0.5Zr2(PO4)3–CePO4 composite ceramics with the introduction of CePO4 phase. Thus, compared with the monophase ceramics, the composite ceramics present a better chemical stability, although the LRCe (< 4.676 × 10−7 g·m−2·day−1) of the composite samples is slightly higher than the LRCe (< 7.038 × 10−8 g·m−2·day−1) of pure CePO4 sample. According to the PCT leaching test results, it is suggested that the in situ prepared (1 − x)Sr0.5Zr2(PO4)3–xCePO4 composite ceramics waste forms possess the superior chemical stability and the NZP-monazite-type composite ceramics could be qualified as a potential host for immobilizing HLW.

Table 1 Normalized elemental leaching rates of (1 − x)Sr0.5Zr2(PO4)3–xCePO4 (x = 0, 0.4, 0.6, and 1.0) composite ceramics
Full size table

Conclusions

The (1 − x)Sr0.5Zr2(PO4)3–xCePO4 (x = 0–1.0) composite ceramics for simultaneously immobilizing Sr and Ce radionuclides were in situ prepared via one-step microwave sintering technique. The composite ceramics were composed of Sr0.5Zr2(PO4)3 and CePO4 phases that were stable and compatible well. As the surrogates for fission product and variable trivalent/tetravalent actinides, the simulated radionuclides of Sr and Ce were independently immobilized into Sr0.5Zr2(PO4)3 phase and CePO4 phase, respectively. The valence state of Ce in the as-prepared composite ceramics existed in trivalent state. The Sr0.5Zr2(PO4)3 phase and CePO4 phase were evenly distributed in the composite ceramics and the existence of CePO4 phase brought about the grain refinement. Moreover, the Sr0.5Zr2(PO4)3–CePO4 composite ceramics possessed the excellent densification with the relative density all above 95% and the high Vickers hardness up to 774 HV1. The PCT leaching test results showed that the Sr0.5Zr2(PO4)3–CePO4 composite ceramics exhibited the better chemical stability than the Sr0.5Zr2(PO4)3 or CePO4 monophase ceramics. It is suggested that NZP-monazite-type composite ceramics could be a potential host for the co-immobilization of fission products and actinide nuclides with diverse valences and ionic radius.