3.1 Analysis of Gd2 − xNdxZr2O7 (0.0 ≤ x ≤ 2.0) nanocrystalline powders
The XRD patterns of calcined Gd2 − xNdxZr2O7 (0.0 ≤ x ≤ 2.0) nanocrystalline powders formed by the solvothermal method are shown in Fig. 2. Interestingly, the diffraction peaks of all powders present basically the same parameters as the JCPDS cards (PDF No. 80–0471 and 78-1289), which indicates that they belong to defective fluorite structure. According to Subramanian theory , Gd2 − xNdxZr2O7 (0.5 ≤ x ≤ 2.0) should be a pyrochlore structure because rA/rB values are between 1.46 and 1.54. The cation radius ratio at the A and B sites is based on the following formula:
Where r is the radius ratio of the cations at the A and B sites; rA is the average cation radius at the A site, consisting of Gd3+ and Nd3+; is 1.05 Å, is 1.11 Å and rB is equal to (0.72 Å); x is the matching value in Gd2−xNdxZr2O7 (0.0 ≤ x ≤ 2.0). The above phenomenon occurs due to the influence of surface/interface energy on the Gibbs free energy, so nanocrystalline powders always exist in the form of a high-temperature phase (defective fluorite structure) . In addition, it can be discovered that the diffraction peaks gradually move to a smaller angle as the value of x increases. This occurrence is because the radius of Nd3+ is larger than that of Gd3+, which causes the unit cell to expand. The XRD consequences also point out that it is feasible to introduce the simulated radionuclide Nd into the crystal structure using the solvothermal method.
The powders (at x = 0.0 and 2.0) were characterized by TEM to study the doping amount influence on the micro-morphology. Figure 3(a) and (b) are the micro-morphology of Gd2Zr2O7 nanocrystalline powder (x = 0.0), while (c) and (d) correspond to Nd2Zr2O7 (x = 2.0). The crystallite shape of both powders is approximately spherical and the average grain size of the two powders are 6.5 nm and 7.0 nm. According to the fitting calculation of their XRD data by Jade software, the grain size of Gd2Zr2O7 and Nd2Zr2O7 nanocrystalline powders are 6.4 and 6.9 nm. This conclusion is basically in consistence with the TEM results, which reveals that Nd doping has no significant effect on the micro-morphology of nanopowders prepared by solvothermal method.
3.2 XRD and Raman analysis of Gd2 − xNdxZr2O7 (0.0 ≤ x ≤ 2.0) ceramics
Fig. 4(a) presents the XRD patterns of Gd2-xNdxZr2O7 (0.0 ≤ x ≤ 2.0) ceramics. The original matrix Gd2Zr2O7 indicates a defective fluorite structure. This result can be attributed to the high sintering temperature of 1642 °C as Gd2Zr2O7 exhibits defective fluorite structure when the sintering temperature is higher than 1550 °C . Interestingly, the structure of Gd2-xNdxZr2O7 ceramics starts to change from pyrochlore structure to defective fluorite structure when x value goes beyond 1.0. With the doping of Nd, the cation radius ratio at the rA/rB is greater than 1.46 and its structure should theoretically be a pyrochlore structure. However, since the grain size of prepared ceramics is in sub-micron scale and the samples have more grain boundaries and crystal planes, which also affects the Gibbs free energy of the system. So it is still a defective fluorite structure . The cation radius ratio and the grain size gradually increases with the Nd content when x value goes beyond 1.0, which makes the ceramics form pyrochlore structure. Therefore, the phenomenon of structural change from defective fluorite to pyrochlore appears. In addition, it can be found that the diffraction peaks gradually shift to a small angle with the Nd content increases. This outcome is in accordance with the XRD results of nanocrystalline powders, which can be attributed to lattice expansion caused by Nd doping. The phenomena of this study are somewhat different from previous report . Lu et al. have used SPS (1700 °C sintering temperature) to prepare ceramics with a grain size of about 9 μm. Transition from defective fluorite structure to pyrochlore structure appears when x ≥ 0.8. The reason for this difference is that the sub-micron ceramic samples have a greater influence on the Gibbs free energy of the system, and there is also a certain difference in the sintering temperature between the two systems. Thus, the critical point of transition from defective fluorite structure to pyrochlore structure is different between the two studies.
As a supplement and verification of XRD results, the Raman spectroscopy analysis was also conducted to further understand the ordered structure of Gd2 − xNdxZr2O7 ceramics as presented in Fig. 4(b). Firstly, the original matrix Gd2Zr2O7 ceramic shows three Raman active modes of the defective fluorite structure, including Eg, F2g and A1g . When x increases from 0 to 1.0, Gd2 − xNdxZr2O7 (0.5 ≤ x ≤ 1.0) ceramics also display the same three Raman active modes as above. When x = 1.5 and 2.0, Gd0.5Nd1.5Zr2O7 and Nd2Zr2O7 demonstrate four kinds of pyrochlore structure Raman active modes, which are Eg + A1g + 2F2g . In these Raman active modes, A1g corresponds to O-Zr-O bending vibration type, Eg corresponds to Zr-O bending vibration type and F2g corresponds to Zr-O and Gd (Nd)-O stretching vibration type. This result is in line with XRD analysis that the pyrochlore structure appears at x = 1.5 and 2.0. Secondly, the Raman peaks shift to lower wavenumbers as x increases from 0 to 2.0 in Fig. 4(b). It may be ascribed to the change of force constant caused by Nd doping in the A site. The average cation radius ratio (rA/rB) increases from 1.46 to 1.54 with the continuous doping of Nd at the A site. This condition makes the bond force constant of Gd (Nd)-O and Zr-O variations, which leads to the change of ZrO6 coordination octahedron vibration modes in Gd2 − xNdxZr2O7 ceramics . Hence, the Raman vibration peaks generated by these groups shift to lower frequencies as the average rA/rB ratio increases.
3.3 Micromorphology and density analysis of Gd2 − xNdxZr2O7 (0.0 ≤ x ≤ 2.0) ceramics
Fig. 5 demonstrates the micromorphology of Gd2-xNdxZr2O7 (0.0 ≤ x ≤ 2.0) ceramics sintered by SCF/QP technology. It can be discovered that the grain boundary of ceramics after SCF sintering is clear. The grain shape is approximately spherical and all display a dense microscopic state. The grain size distribution was counted by Nano measure software and the statistical results are depicted in the upper right corner of Fig. 5(a)-(e). The average grain size of the original matrix Gd2Zr2O7 is 79 nm, which is in consistent with our previous work . When the x values are 0.5, 1.0, 1.5, and 2.0, the average grain size is 116 nm, 164 nm, 205 nm, and 265 nm. Fig. 5(f) shows the fitting result of XRD datas with Jade software to obtain the grain size. The results indicate that the calculated grain size is in accordance with the SEM grain statistics. The above results indicate that the average grain size of the ceramic samples also gradually enlarges with the Nd doping amount increases. This phenomenon is due to the fact that the atomic radius of Nd is larger than Gd, which leads to the grain size gradually enlarges after sintering.
The density measured by the Archimedes drainage method is displayed in Fig. 6. At the same time, the results are compared with previous research . It can be seen that the sample density prepared by the three sintering methods decreases with the increase of Nd content. Density is directly proportional to mass and inversely proportional to volume. As the radius of Nd atom is larger than Gd atom while Nd atom is lighter than Gd atom, the sample density gradually decreases in this system. In addition, the sample density after SCF/QP sintering is between SPS sintering and conventional sintering. However, the grain size of the samples sintered by SPS (1700 °C, 3 min) and conventional sintering (1500 °C, 72 h) is in micron scale. This consequence is due to SPS sintering temperature as high as 1700 °C and conventional sintering time as long as 72 hours. SCF sintering can produce dense samples with sub-micron grain sizes in just a few minutes. The density of Gd2Zr2O7 after SCF sintering is 6.17 g·cm− 3. Compared with the previous result of 5.53 g·cm− 3, the improved preparation process did improve the density . Nd2Zr2O7 with the largest amount of Nd doping has a density of 5.87 g·cm− 3 and relative density of 90.3%.
3.4 Aqueous durability analysis of Nd2Zr2O7 ceramic
It is well known that nuclear waste forms will eventually be disposed in deep geological repositories. The waste forms will inevitably contact with groundwater and aqueous durability is a very important property to evaluate the safety of nuclear waste forms [39, 40]. Standard MCC-1 measurements were performed at 90 °C for 1–42 days to estimate the leaching behavior of Nd2Zr2O7 sample. The normalized elemantal leaching rates (LRi) are depicted in Fig. 7. As the immersion time was extended from 1 day to 42 days, the degree of LRNd value reduction during the first 28 days is extremely large, which exceeds two orders of magnitude. But the value gradually stabilizes from the 28th to the 42nd day. The LRi value for Zr shows a different trend. The LRZr value decreases significantly in the first seven days. Then, there is a slight fluctuation between the 7th and 42nd days. For short-term leaching tests, fluctuations in LRi values on the same order of magnitude is a common phenomenon . The leaching values of Nd and Zr on the 42nd day are 1.1 × 10− 6 g•m− 2•d− 1 and 2.5 × 10− 7 g•m− 2•d− 1, respectively. Their values are lower than the leaching data of ordinary SYNROC (at the order of 10− 4 to 10− 3 g•m− 2•d− 1) . This result demonstrates that Gd2Zr2O7 indicates great potential as immobilizaiton matrix of radioactive wastes.
The influence of MMC-1 leaching experiment on the samples phase structure at different depths was further studied by the analysis of GIXRD, which has been used to evaluate the phase structure leaching stability of samples. The GIXRD patterns of Nd2Zr2O7 after the MCC-1 leaching experiment are presented in Fig. 8. GIXRD analysis with different incident angles was carried out to achieve the purpose of detecting different depths. The incident angle γ was selected to be 4°, 2°, 1° and 0.5°, and the corresponding detection depths are 1440 nm, 720 nm, 360 nm and 180 nm. These depth values are calculated by the critical angle model, which has been described in the previous work . The Nd2Zr2O7 maintains the pyrochlore structure when the incident angle decreases from 4° to 1.0°. However, some characteristic peaks of pyrochlore structure begin to disappear (2θ = 60.3° and 70.9°) when the incident angle drops to 0.5°. Compared with the original Nd2Zr2O7 sample, the characteristic peaks of pyrochlore structure do indeed weaken after leaching, and even some weaker characteristic peaks disappear. This phenomenon demonstrates that the leaching experiment mainly caused slight damage to the crystal structure in the depth range of 180 nm. There is basically no effect in the depth range of more than 360 nm. This conclusion proves that Gd2Zr2O7 nanocrystalline ceramics possess excellent aqueous stability in immobilizaiton radioactive wastes.