J Am Ceram Soc. 2020;103:3425–3435. wileyonlinelibrary.com/journal/jace | 3425© 2020 The American Ceramic Society

Received: 4 October 2019 | Revised: 10 December 2019 | Accepted: 7 January 2020

DOI: 10.1111/jace.17030

O R I G I N A L A R T I C L E

He irradiation-induced lattice distortion and surface blistering of Gd2Zr2O7 defect-fluorite ceramics

Zhangyi Huang1,2,3,4 | Mao Zhou1 | Zhangkai Cao5 | Zhe Tang1,6 | Yutong Zhang1,6 | Junjing Duan1,6 | Jianqi Qi1,2,6 | Xiaofeng Guo3,7 | Tiecheng Lu1,2,6 | Di Wu3,4,7,8

1College of Physics, Sichuan University, Chengdu, China2Key Laboratory of Radiation Physics and Technology of Ministry of Education, Sichuan University, Chengdu, China3Alexandra Navrotsky Institute for Experimental Thermodynamics, Washington State University, Pullman, WA, USA4The Gene and Linda Voiland, School of Chemical Engineering and Bioengineering, Washington State University, Pullman, WA, USA5Department of Physics, Beijing Normal University, Beijing, China6Key Laboratory of High Energy Density Physics of Ministry of Education, Sichuan University, Chengdu, China7Department of Chemistry, Washington State University, Pullman, WA, USA8Materials Science and Engineering, Washington State University, Pullman, WA, USA

CorrespondenceJianqi Qi and Tiecheng Lu, College of Physics, Sichuan University, Chengdu 610064, China.Email: lutiecheng@scu.edu.cn (T. L.); qijianqi@scu.edu.cn (J. Q.)

Di Wu, Alexandra Navrotsky Institute for Experimental Thermodynamics, Washington State University, Pullman, Washington 99164, United States.Email: d.wu@wsu.edu

Funding informationDepartment of Science and Technology of Sichuan Province, Grant/Award Number: 15CXTD0025; National Natural Science Foundation of China, Grant/Award Number: 11775152 and U1830136

AbstractGd2Zr2O7 ceramics with different grain sizes ranging from nanoscale to submicron scale (91, 204, and 634 nm) were irradiated at room temperature using 190 keV He ions with doses ranging from 5 × 1016 to 5 × 1017  ions/cm2. We fully character-ized the pre- and post-irradiation samples using grazing-incidence X-ray diffrac-tion (GIXRD), scanning electron microscope (SEM), and atomic force microscope (AFM) as the grain size and degree of irradiation vary. The results suggested that all three Gd2Zr2O7 samples demonstrate outstanding radiation tolerance to displace-ment damage by retaining their crystallinity after irradiation at 5 × 1017 ions/cm2. which is equal to 16 displacement per atom (dpa) at peak positions. Although lattice expansion was observed at a He irradiation at 5 × 1016  ions/cm2 and beyond, the lattice remained stable for the nanograin ceramic, while the degree of distortion for the sample with the largest grain size (634 nm) continuously increased. Moreover, a delayed He bubble evolution process was seen for the nanograin ceramic, which did not appear for the submicron-grained sample. Interestingly, the grain size-dependent surface blistering was also found to be a function of ion fluence. After He irradia-tion at 5 × 1017 ions/cm2 the AFM root-mean-square(RMS) roughness variation for Gd2Zr2O7 ceramics of 91, 204, and 634 nm were 4.8, 7.0, and 11.1 nm, respectively.

K E Y W O R D S

nanostructures, nuclear waste, pyrochlore

3426 | HUANG et Al.

1 | INTRODUCTION

Synthetic rock (SYNROC) composed of mutually compatible ceramic phases, such as pyrochlore, perovskite, and rutile, was proposed as an alternative waste form to borosilicate glass.1‒3 These ceramic waste forms with intricate crystal structures enable the immobilization of nuclear waste elements at their specific crystallographic lattice sites, leading to enhanced chemical stability compared with borosilicate glass in deep geological repository environments.2,4,5 Pyrochlore phase is one promising candidate component of SYNROC owing to its highly compositionally flexible structure for immobiliza-tion of radioactive actinides.6

The crystal structure of A2B2O7-type material is deter-mined by the cation radius ratio (rA/rB). When the rA/rB is less than 1.46, fluorite structure is favored, while the pyrochlore phase is preferred when rA/rB is larger than 1.46. Interestingly, the rA/rB ratio for Gd2Zr2O7 is exactly 1.46. As a result, competing formation of pyrochlore and defect-fluorite phases is expected.6 Pyrochlore oxides have a general formula of A2B2O7 (Fd3m, No. 227), which are typically considered as the derivatively ordered superstruc-tures of the isometric fluorite structure (AO2, Fm3m, No. 225). Specifically, in a pyrochlore structure, there are two distinct cation sites, (a) the A-site, with trivalent cations (8-coordinate) occupying the 16d site, and (b) the B-site with tetravalent metal cations (6-coordinate) locating at the 16c site. Moreover, pyrochlore has three anion sites, the 48f site with six oxygens, 8b site having one oxygen, and 8a with one oxygen vacancy. On the other hand, in a defect-fluorite structure with eleven atoms per cell, four cations are randomly distributed at the 4a site, while seven oxygen anions and one oxygen vacancy are randomly po-sitioned at the 8c site. Figure 1 illustrates the differences between the pyrochlore and defect-fluorite structures.

So far, studies on the A2B2O7-type materials and their applications in the nuclear industry mainly focus on the fol-lowing topics: the interrelation between radiation stability and cation radius ratio,6,7 the role of electronic energy loss, the synergy between nuclear and electronic energy losses in the ion-irradiation processes,8‒10 and the critical ion dose re-quired to induce amorphization or phase transition.11‒13 These studies are critical for a fundamental understanding of the ir-radiation effects and further design of radiation-resistant ce-ramics used in extreme radioactive environments. According to these studies, Gd2Zr2O7 is one of the most promising ce-ramic waste forms owing to its (a) extremely high cross-sec-tion of Gd for neutron absorption which enables criticality safety, (b) high-capacity immobilization of actinides at the A- and B- sites, and (c) high resistance to amorphization by energetic particles up to 100 dpa.3,6,11,14

Recently, nanostructured materials with large interfacial areas and/or grain boundaries were found to have enhanced

radiation resistance to displacement damage because of ef-ficient annihilation of point defects generated in the irra-diation process at the interfaces and/or grain boundaries. Such phenomena were observed for heavy ion irradiation of MgGa2O4, YSZ, Gd2(Ti0.65Zr0.35)2O7, etc.15‒17 Moreover, the He atoms undergo a continuous accumulation, forming He bubbles, and lead to surface swelling and blistering once the He bubble nucleation concentration is reached.18 In addition, He accumulation also causes mechanical stress that results in cracks decreasing the chemical durability of the waste forms.19,20 The irradiation-introduced He atoms are insolu-ble within the host ceramic matrices and are considered to be impurities. Such irradiation damage is intrinsically differ-ent compared with structural damages. Very recently, stud-ies have shown that interfaces play a positive role in helium accumulation inhibition.21,22 For example, interfaces can re-duce the He bubble density during irradiation of solid-state materials.21,22 Grain boundaries are found to play a similar role as that of interfaces. However, although Gd2Zr2O7 is one of the most promising ceramic materials for immobilization of radioactive actinides, studies on its radiation resistance behaviors, especially, systematic investigations into the grain size—radiation resistance relationships on a nanoscale, are far from complete.

In this work, Gd2Zr2O7 ceramics with different grain sizes ranging from nanometer to submicron meter were synthe-sized to study the effects of helium irradiation. The ion flu-ence covered from 5 × 1016 to 5 × 1017  ions/cm2, with the corresponding peak He concentrations evolving from 2.39 to 23.9 at%. The samples were thoroughly characterized using GIXRD, SEM, and AFM to determine the post-irradiation lattice distortions, helium bubble formation and surface blis-tering, respectively.

2 | EXPERIMENTAL METHODS

2.1 | Material preparation

Gd2Zr2O7 nanocrystalline powders with an average crystal-lite size of 6.3 ± 0.3 nm were synthesized by a solvothermal method followed by powder calcination at 800°C for 4 hours. The detailed preparation procedures are described in our pre-vious publication.23 The calcined powders were uniaxially dry-pressed into Φ10 mm pellets under 8 MPa. Subsequently, these green pellets were cold isostatically pressed under 200  MPa. Finally, Gd2Zr2O7 ceramics with different grain sizes were obtained after pellet sintering in ambient con-ditions at elevated temperatures. Several heating profiles were used: (Condition 1) 1380°C for 1  hours followed by 1210°C for 20 hours, (Condition 2) 1500°C for 2 hours, and (Condition 3) 1600°C for 2 hours. The as-obtained samples were mirror-polished on both sides using fine metallographic

| 3427HUANG et Al.

abrasive paper and diamond paste until the surface roughness was less than 10 nm.

2.2 | Irradiation treatment

Gd2Zr2O7 ceramics with different grain sizes were irradi-ated with He ions (190 keV) of various fluences: 5 × 1016, 1 × 1017, and 5 × 1017 ions/cm2 at room temperature using an ion implanter (200 keV, LC22-100-01, Beijing Zhongkexin Electronics Equipment Co., Ltd) at the Centre for Ion Beam Application, Wuhan University. The corresponding ion range, displacement per atom (dpa), and He concentration at an implanted ion fluence of 5 × 1017 ions/cm2 as a function of incident ion penetration depth were computed using full cascade calculations in the Monte Carlo computer code—Stopping Range of Ions in Matter (SRIM, version 2008). The displacement energies of Gd, Zr, and O atoms used for cal-culation were 38.8, 41.4, and 18.6  eV, respectively.24 The density was set to be 6.3 g/cm3.

Figure 2 shows the SRIM calculated damage (dpa) and He concentration profiles for Gd2Zr2O7 sample irradiated with 190 keV He ions to a fluence of 5 × 1017 ions/cm2. The simulation indicates that the irradiated region extends to a depth of 652 ± 152 nm underneath the surface and the max-imum irradiation damage is approximately 16 dpa at a depth of ~650 nm. He concentration reaches a maximum value of ~23.9 at.% at a depth of ~730 nm.

2.3 | Sample characterizations

The surface morphologies and roughness of the samples before irradiation were checked by an optical microscope (BX53M

System Microscope, Olympus) and a surface profiler (KLA Tencor, model Alpha-Step IQ). The scan length of each pol-ished sample on a random area using the surface profiler is 500 µm, and the step size for recording data is 0.04 µm. The phases of pre-irradiation samples were examined using a pow-der X-ray diffractometer (XRD, DX-2700, Dandong Fangyuan, Dandong) using Ni-filtered Cu-Ka (λ = 1.5406 Å, 0.03° per step, 2θ from 10 to 100°). Grazing-incidence X-ray diffrac-tion (GIXRD) was performed on a Bruker AXS D8 Advances X-ray diffractometer with CuKa radiation (λ = 1.5406 Å). The GIXRD data were collected by step-scanning method from 20° to 70° with a step interval of 0.02° and a count time of 2 sec-onds per step. The phase transition and lattice parameter evolu-tion for post-irradiation Gd2Zr2O7 ceramics were determined by the whole pattern profile fitting of the diffraction data using

F I G U R E 1 Schematic images comparing the arrangements of cations and anions in the unit cells of pyrochlore and defect fluorite compounds (both with a general formula of A2B2O7)

F I G U R E 2 Depth profiles of radiation damage obtained from SRIM simulation with dpa (blue) and helium concentration (at.%, red) as the unit

3428 | HUANG et Al.

MDI Jade 6.0. Before the GIXRD experiments, the X-ray pen-etration depth in Gd2Zr2O7 was calculated on the basis of total external reflection theory using a density value of 6.3 g/cm3 (about 1.65  ×  1024  electrons/cm3).25,26 Figure 3 presents the X-ray penetration depth profile as a function of grazing angle. At an incident angle of 4.2°, the X-ray penetration depth reaches ~814 nm, slightly higher than the sum of ion average range and straggling. Hence, we chose 4.2° for the GIXRD tests. The frac-ture cross-section images of pre- and post-irradiation Gd2Zr2O7 ceramics were collected using a scanning electron microscope (SEM, FEI, Inspect-F50). The average grain sizes of the sin-tered ceramics were determined by selecting and measuring the dimensions of more than 100 identifiable grains in the SEM images using Nanomeasure software. The bulk densities of all samples were measured by the Archimedes method using water as the media. The morphological surface modifications induced by implanted He ions were monitored using an atomic force microscope (AFM, DI-EnviroScope, Veeco Instruments).

3 | RESULTS AND DISCUSSIONS

Figure 4 shows the XRD patterns of the sintered ceramics under various conditions. The XRD peaks of the sample sintered in Condition 1 is in excellent agreement with the diffraction peaks of defect fluorite (DF) phase (PDF: 80-0471). In contrast, when sintered at 1500 and 1600°C for 2  hours, respectively (Condition 2 and 3), the pyrochlore (P) phase (PDF: 80-0470) can be observed (see the inset of Figure 4). One earlier study demonstrates that Gd2Zr2O7 ceramic prepared using nanocrys-talline powders synthesized by the wet-chemical method ap-pears to have DF to P phase transition at about 1250°C. The P phase Gd2Zr2O7 is preferred at sintering temperature between 1300 and 1500°C, within which the degree of lattice orderness

increases as the sintering temperature increases.27 In our study, Gd2Zr2O7 sintered in Condition 1 presents a DF phase. We at-tribute this phenomenon to the insufficient time (1 hour) for the formation of the P phase. Similarly, the relatively low intensity of P phase superlattice diffraction peaks for the sample sintered in Condition 2 may be also due to the insufficient time (2 hour) for the complete phase transition to highly crystalline P phase. Our sample sintered at 1600°C (Condition 3) also exhibits a P phase. But the intensities of the characteristic diffraction peaks of the P phase in the sample prepared at 1600°C was much lower than that prepared at 1500°C. Typically, the high-temper-ature phase transition of Gd2Zr2O7 from P to DF phase occurs primarily at a temperature higher than 1530°C.28 Thus, the P to DF phase transition was not complete due to the insufficient time (2 hour). These results show that the thermally induced phase transition of Gd2Zr2O7 is a slow process. Interestingly, peak broadening suggesting small crystallite size was observed for the sample sintered in Condition 1, a strong indication that the nanograin ceramic is synthesized successfully.

Figure 5 presents the fracture surfaces of the samples sintered in different conditions. All the samples exhibit a homogeneous microstructure. As the sintering temperature increases, the grain grows accordingly. The statistical aver-age sample grain sizes from more than 100 measurements are 91 ± 14, 204 ± 18, and 634 ± 182 nm for samples sin-tered in Condition 1, 2, and 3, respectively. The microstruc-ture images in Figure 6 taken by an optical metallographic microscope show that the polished samples have smooth surfaces without obvious scratches. The root mean square roughnesses (Rq) for samples sintered in Condition 1, 2, and 3 are about 4.0, 4.4, and 3.2 nm, respectively, as shown in Figure 6. Other than grain boundaries and interphase

F I G U R E 3 X-ray penetration of Gd2Zr2O7 ceramic with a density of ~6.30 g/cm3 as a function of incident angle. At 4.2°, the X-ray penetration (~810 nm) is very close to the maximum ion range. The average ion range is 653 nm with a straggling of 144 nm

F I G U R E 4 XRD patterns of Gd2Zr2O7 ceramics sintered at different temperatures. The inset highlights the data collected from 36° to 46°

| 3429HUANG et Al.

boundaries,15,29 free surfaces also play critical roles to serve as sinks for radiation-induced defects and/or to accommo-date impurities enhancing the radiation resistance of ma-terials.30 In the present study, ceramics with porosity (free surface) were successfully obtained, reflected by the den-sities of the Gd2Zr2O7 ceramics investigated. Specifically, the relative densities measured by the Archimedes method are 88, 89, and 92% for samples prepared in Condition 1, 2, and 3, respectively. A theoretical density of 6.9 g/cm3 was used for Gd2Zr2O7 for calculation.31

The _ENREF_15GIXRD data for all pre- and post-irra-diation Gd2Zr2O7 ceramics with various grain sizes are pre-sented in Figure 7. The 190 keV He irradiation leads to two

major radiation effects: (1) displacement damage and (2) accumulation of the implanted He ions. Numerous heavy-ion irradiation experiments on ceramic materials have shown that accumulation of displacement damage leads to structural degradation such as disordering or amorphiza-tion, evidenced by intensity reductions and broadening of diffraction peaks.16 We found that the degree of intensity reduction and peak broadening increases as the ion dose increases (see Figure 7). Nevertheless, full amorphization was not observed for all three Gd2Zr2O7 samples even after an ion dose of as high as 5 × 1017 ions/cm2, which is com-parable to the peak displacement damage of 16 dpa. This highlights the outstanding radiation tolerance of Gd2Zr2O7

F I G U R E 5 SEM images of fracture surfaces for Gd2Zr2O7 samples sintered under different treatments: (A) at 1380°C for 1 h followed by heating at 1210°C for 20 h, (B) at 1500°C for 2 h, and (C) at 1600°C for 2 h

(A) (B) (C)

F I G U R E 6 The microstructure images of Gd2Zr2O7 ceramics with different average grain sizes (A) 91 nm, (B) 204 nm, and (C) 634 nm before He irradiation. (D), (E), and (F) are corresponding surface roughness data taken on a surface profiler

(A) (B) (C)

(D) (E) (F)

3430 | HUANG et Al.

to displacement damage. Notably, the characteristic dif-fraction peaks of P phase still exist at an ion fluence of 5 × 1016 ions/cm2 (equal to peak displacement damage of ~1.6 dpa) but disappear at an ion fluence of 1 × 1017 ions/cm2 (peak damage: ~3.2 dpa) in the samples which were fabricated at 1500°C and 1600°C (Condition 2 and 3) as shown in the insets of Figure 7B,C. Previous studies have demonstrated the P-to-DF phase transition in Gd2Zr2O7 at low temperature is at a dose as low as 0.4  dpa.32 In this work, the higher threshold dose for the P-to-DF phase transition is attributed to that the area with displacement damage less than 0.4 dpa could be also detected by grazing incident x-rays even at a dose with peak damage of 1.6 dpa. Furthermore, a clear peak shift to lower 2θ angles, sug-gesting lattice expansion, were observed for all post-irra-diation Gd2Zr2O7 ceramics. However, the degree of peak shift cannot be distinguished among samples with different grain sizes. For the nanograin sample (91 nm), a peak shift to lower 2θ was observed at 5 × 1016 ions/cm2. Then, the peak position tends to level off as the ion fluence increases (see Figure 7D). Figure 7E shows the peak position of the sample sintered at 1500°C for 2 hours (Condition 2), which increases until the ion fluence reaches 1 × 1017 ions/cm2. A slight increase in peak shift was observed as the ion fluence increase to 5 × 1017 ions/cm2. However, the peak position of Gd2Zr2O7 ceramic with the largest grain sizes continues its shift to lower 2θ until the fluence reaches 5 × 1017 ions/cm2 (see Figure 7F). To further elucidate the underlying

mechanisms for GIXRD peak shifts, the peak positions were converted to lattice parameters. We plotted the lat-tice parameters as a function of ion fluence (see Figure 8). Generally, for each sample, the trend of peak shift ap-pears to be identical. The lattice parameter increases as the ion fluence increases. This trend levels at a fluence of 5 × 1016 ions/cm2.

Figure 9 shows the cross-sectional images of three Gd2Zr2O7 ceramics exposed to He irradiation of various flu-ences. 5 × 1016 ions/cm2 He irradiation does not result in de-tectable irradiated regions was for all three samples. As the ion fluence increases from 5 × 1016 to 5 × 1017 ions/cm2, a damaged layer is clearly observed at a subsurface depth of 640 nm for samples with the grain sizes of 204 and 634 nm. Additionally, as the grain size increases, the damaged region appears to be better resolved. There is a horizontal white line highlighting the interface with peak He concentration (see Figure 7H,I). The size (width) of the damaged region is in good agreement with the ranges predicted by SRIM simu-lation. However, the damaged region was not seen for the post-irradiation nanograin ceramic, even at 5  ×  1017  ions/cm2. This strongly suggests a delayed evolution process for He bubble formation. Interestingly, compared with the unir-radiated region, the irradiated region of nanograin ceramic is found to have a lower degree of porosity after irradiation at 5 × 1016  ions/cm2. Since irradiation-induced interstitial de-fects normally have a higher migration rate than that of va-cancy defects, the porosity within the irradiated region may

F I G U R E 7 GIXRD patterns of Gd2Zr2O7 ceramics with different average grain sizes (A) 91 nm, (B) 204 nm, and (C) 634 nm before and after He irradiation at various fluences. (D), (E) and (F) highlights the GIXRD peaks with the highest intensity reflection, in (A), (B) and (C), respectively. The insets in (B) and (C) highlight the data collected from 35° to 40°

| 3431HUANG et Al.

serve as favorable sites to accommodate and regulate the in-terstitial defects and implanted He ions,33,34 leading to pore volume reduction.

Figure 10 shows the AFM images of the Gd2Zr2O7 ceramics surfaces before and after He ions irradiation. Figure 11 presents the corresponding RMS roughness as a function of ion fluence. Prior to irradiation, the surfaces of all samples appear to be relatively smooth with RMS roughness less than 4.1 nm. Upon 5 × 1016 ions/cm2 irra-diation. Several tapered blistering emerge (less than three in a 20 μm × 20 μm box). As the ion fluence increases, the number of the tapered blistering and the RMS roughness both increase. On the other hand, at the same ion fluence as the average grain size increases the number of the tapered

blistering and the RMS roughness both increase. When the ion fluence increases to 5 × 1017  ions/cm2, a distinct dif-ference in surface morphology can be observed. The RMS roughness changes for Gd2Zr2O7 ceramics with different grain sizes, 91, 204, and 634 nm, are 4.8, 7.0, and 11.1 nm, respectively. Grain size reduction was tested to be an ef-fective strategy to decrease the size and number of irradia-tion-induced surface blistering.

Generally, He irradiation can induce structural dam-ages, including interstitial-type, and vacancy-type defects. Interstitial-type defects have a higher migration rate, which prefers to migrate to the material surface, while vacancies slowly agglomerate to form larger vacancy clusters.34 Vacancy clusters can be stabilized by further absorption of He atoms

F I G U R E 8 The lattice parameter of Gd2Zr2O7 ceramics as a function of the magnitude of fluence for an average sample grain size of (A) 91 nm, (B) 204 nm, and (C) 634 nm

F I G U R E 9 Cross-sectional area SEM images of Gd2Zr2O7 ceramic with different average grain size (A, B and C) 91 nm, (D, E and F) 204 nm, and (G, H and I) 634 nm, before and after He irradiation at various fluences

(A) (B) (C)

(D) (E) (F)

(G) (H) (I)

3432 | HUANG et Al.

forming larger He-vacancy clusters, which have higher sta-bility.35 These clusters tend to continue their growth through the mechanism of helium and vacancy absorption until reaching the critical size, at which no longer further growth is observed.36,37 However, we found that the nucleation and growth of He bubbles at the interfaces, such as the grain and/or interphase boundaries, present a new morphological trans-formation mechanism. Specifically, first, the formation of a relatively stable state with platelet-shaped He-filled cavities was observed at certain interfaces. Subsequently, these cavi-ties grow and lead to the formation of He bubbles.38 Typically, the shape of He bubble is determined by the interplays and competitions between elastic free energy and surface energy.

Faceted and spherical bubbles are typically observed in surface energy dominated processes, while precipitation of oblate or ribbon-like bubbles is seen in elastic-free energy governed processes.39

Talking YSZ for an example, as the He ion fluence in-creases helium bubble formation in YSZ ceramics exhibits four distinct stages. They are Stage 1, generation of isolated bubbles along the grain boundaries instead of within the grains at a peak fluence concentration of 3.2 at.%; Stage 2 formation of coalesced bubbles and interconnected channels along the grain boundary at a peak fluence concentration of 5.1 at.%; Stage 3, the formation of isolated spherical and ribbon-like bubbles inside the grains, which are parallel with the sample surface, with a peak fluence concentration of ~12.8 at.%; Stage 4, the coalescence of ribbon-like bubbles and the formation of large cracks at a peak fluence concentration of 27.7 at.%.39

For our Gd2Zr2O7 ceramics—He system, according to the width, morphology and locations, the cracks were seen in post-He-irradiated YSZ (width ~27 nm) appear to be very similar as the cracks observed in Figure 9I, shown as the horizontal white lines in the SEM images with a width of ~32 nm. Interestingly, at a fluence of 5 × 1017 ions/cm2 bub-ble formation in Stage 4 was seen for the submicron sample (634 nm), while for the nanograined Gd2Zr2O7 ceramic, such phenomenon was not observed (see Figure 9C). This prob-ably suggests a delayed He bubble formation for nanograin Gd2Zr2O7 ceramic, which has been proved in our previous study.40 Surface blisters as large as 60  nm are imaged for the submicron Gd2Zr2O7 ceramic (see the AFM images in Figure 10). This indicates that the grain boundaries, instead of the inner grains, are the favorable nucleation sites for the formation of He bubbles, which prefer to align and distribute themselves along grain boundaries. Figure 12 gives a brief illustration of the surface blistering induced by the formation

F I G U R E 1 0 AFM images of Gd2Zr2O7 ceramics with different average grain sizes before and after He irradiation at various fluences

F I G U R E 1 1 RMS roughness evolutions of Gd2Zr2O7 ceramics as a function of He irradiation fluence for the sample with different average grain sizes

| 3433HUANG et Al.

of large He bubbles in coarse-grained ceramic. Furthermore, faster accumulation of the irradiation-induced interstitial-type defects at the surface in coarse-grained ceramic rather than in nanograin ceramic also gives rise to the surface blistering, as clearly shown in the figure in the Ref [42].

Cumulative structural defects result in lattice distortion, which is reflected by variation in lattice parameters. It is found

that vacancy and vacancy clusters induce the stress field, which leads to lattice expansion.41 For the post-irradiation nanoceram-ics, we speculate the cause for the slight variation for the lattice parameter—grain size dependence shown in Figure 8A. It is pos-sible that the enhanced annihilation of interstitial and vacancy defects via interstitial emission34 enables an equilibrium state between annihilation and vacancy generation upon continuous

F I G U R E 1 2 A brief illustration of the surface blistering induced by the formation of the helium bubble

F I G U R E 1 3 A brief illustration of the interactions between the irradiation-induced defects (Gd/Zr/O interstitial and vacancy defects/defects clusters) and surface/interface (grain boundaries, surface and pores) in nanograin ceramic and coarse-grain ceramic

3434 | HUANG et Al.

radiation damage. In contrast, the annihilation of interstitial and vacancy defects in the submicron sample is not able to coun-terbalance the continuous generation of the irradiation-induced vacancies due to insufficient grain boundaries. Therefore, as the ion fluence increases, the lattice expands. Notably, the unit cell swelling at 5 × 1016  ions/cm2 in the sample with the smallest grain size (91 nm) is the largest compared to the other two sam-ples. Figure 13 is a brief illustration of the interactions between the irradiation-induced defects (Gd/Zr/O interstitial and vacancy defects/defects clusters) and surface/interface (grain boundaries, surface and pores) in nanograin ceramic and coarse-grain ce-ramic, showing what may cause this phenomenon. When ener-getic He+ (190 keV) was implanted into the Gd2Zr2O7 ceramics, Gd/Zr/O atoms will be displaced out of their lattices, forming interstitial and vacancy defects. Grain boundaries and the ma-terial surface were proved to be the preferred dwelling site for interstitials.34,42 We have seen in SEM images of nanograin ce-ramics that the pores within the irradiated region may also serve as favorable sites to accommodate and regulate the interstitial defects. However, in coarse-grained ceramics, only the ceramic surface fewer pores can arrange interstitial defects, and the dif-fusion distance for interstitial defects to the material surface is longer than the distance to the grain boundaries or the pores inside the material. Thus, more interstitial defects in nanograin ceramic were trapped by grain boundaries, pores and material surfaces, while more mobile interstitial defects remained in coarse-grained ceramics. As the interstitial defects normally have a higher migration rate than that of vacancy defects, the interstitial vacancy recombination should be dominated by the concentration of mobile interstitial defects. Therefore, when the ion dose is not high, there should be more vacancies annihilated by the recombination with interstitials in coarse grain ceramics, leaving behind fewer residual vacancy defects and causing less increase in lattice parameter. When the concentration of intersti-tial defects on the grain boundary reaches saturation, enhanced annihilation between interstitial and vacancy defects via inter-stitial emission should occur, leading to an enhanced radiation resistance to structural defects accumulation in nanograin ce-ramics. However, this work confirms that pores can absorb in-terstitial defects, but it is still uncertain whether it can promote the annihilation effect between interstitial-vacancy defects at a relatively high ion dose. We will provide a greater understanding of these results in our future work.

4 | CONCLUSIONS

In summary, the lattice distortion, helium behavior evolution and surface blistering of Gd2Zr2O7 ceramics were thoroughly investigated as the ceramic grain size and a dose of He irra-diation vary. Using GIXRD, SEM, and AFM as fundamental experimental tools, we reveal the following general phenom-ena. All Gd2Zr2O7 ceramic retains defect fluorite structures

after He irradiation with peak dpa up to 16, and lattice expan-sion is effectively inhibited and avoided by grain size reduction to nanoscale even at a dose of 5 × 1016 ions/cm2. Nanograin Gd2Zr2O7 ceramic has delayed He bubble evolution process. This is because its large grain boundary areas facilitate favora-ble nucleation and distribution sites for the He bubbles. Pores can serve as favorable sites to accommodate and regulate the in-terstitial defects. Moreover, size-dependent surface blistering as a function of the degree of ion fluence is revealed. Specifically, the RMS roughness variations for irradiated Gd2Zr2O7 ceramics (5 × 1017 ions/cm2) are 4.8, 7.0, and 11.1 nm for samples with grain sizes of 91, 204, and 634  nm, respectively. Nanograin Gd2Zr2O7 ceramics demonstrate promising performance to in-hibit lattice distortion (mechanical stress to the lattice), helium bubble formation and surface blistering. The fundamental in-sights obtained from this research may contribute to the nuclear material society by offering general guidance for tuning the grain size of A2B2O7 ceramics for radiation-resistant materials applied in extremely radioactive environments.

ACKNOWLEDGMENTSThis work was supported by the National Natural Science Foundation of the People's Republic of China under Grant Nos. 11775152 and U1830136, and the Science and Technology Innovation Team of Sichuan province under Grant NO.2015TD0003. Di Wu acknowledges the fund of Alexandra Navrotsky Institute for Experimental Thermodynamics, and the institutional funds from the Gene and Linda Voiland School of Chemical Engineering and Bioengineering at Washington State University.

ORCIDJianqi Qi  https://orcid.org/0000-0002-6884-1258 Di Wu  https://orcid.org/0000-0001-6879-321X

REFERENCES 1. Lumpkin GR. Ceramic waste forms for actinides. Elements.

2006;2(6):365–72. 2. Ringwood A, Kesson S, Ware N, Hibberson W, Major A.

Immobilisation of high level nuclear reactor wastes in SYNROC. Nature. 1979;278(15):219–23.

3. Ewing RC, Weber WJ, Lian J. Nuclear waste disposal—pyrochlore (A2B2O7): nuclear waste form for the immobilization of plutonium and “minor” actinides. J Appl Phys. 2004;95(11):5949–71.

4. Weber WJ, Ewing RC, Catlow CRA, de la Rubia TD, Hobbs LW, Kinoshita C, et al. Radiation effects in crystalline ceramics for the immobilization of high-level nuclear waste and plutonium. J Mater Res. 1998;13(6):1434–84.

5. Boccaccini AR, Berthier T, Seglem S. Encapsulated gadolinium zirconate pyrochlore particles in soda borosilicate glass as novel radioactive waste form. Ceram Int. 2007;33(7):1231–5.

6. Sickafus KE. Radiation tolerance of complex oxides. Science. 2000;289(5480):748–51.

7. Lian J, Helean KB, Kennedy BJ, Wang LM, Navrotsky A, Ewing RC. Effect of structure and thermodynamic stability on the response

| 3435HUANG et Al.

of lanthanide stannate-pyrochlores to ion beam irradiation. J Phys Chem B. 2006;110(5):2343–50.

8. Weber WJ, Duffy DM, Thomé L, Zhang Y. The role of electronic energy loss in ion beam modification of materials. Curr Opin Solid State Mater. 2015;19(1):1–11.

9. Weber WJ, Zarkadoula E, Pakarinen OH, Sachan R, Chisholm MF, Liu P, et al. Synergy of elastic and inelastic energy loss on ion track formation in SrTiO3. Sci Rep. 2015;5:7726.

10. Zhang Y, Aidhy DS, Varga T, Moll S, Edmondson PD, Namavar F, et al. The effect of electronic energy loss on irradiation-induced grain growth in nanocrystalline oxides. Phys Chem Chem Phys. 2014;16(17):8051–9.

11. Lian J, Zu XT, Kutty KV, Chen J, Wang LM, Ewing RC. Ion-irradiation-induced amorphization of La2Zr2O7 pyrochlore. Phys Rev B. 2002;66(5):054108–12.

12. Lian J, Wang L, Chen J, Sun K, Ewing RC, Matt Farmer J, et al. The order–disorder transition in ion-irradiated pyrochlore. Acta Mater. 2003;51(5):1493–502.

13. Lian J, Chen J, Wang LM, Ewing RC, Farmer JM, Boatner LA, et al. Radiation-induced amorphization of rare-earth titanate py-rochlores. Phys Rev B. 2003;68(13):134107–15.

14. Weber WJ, Ewing RC. Plutonium immobilization and radiation ef-fects. Science. 2000;289(5487):2051–2.

15. Shen TD, Feng S, Tang M, Valdez JA, Wang Y, Sickafus KE. Enhanced radiation tolerance in nanocrystalline MgGa2O4. Appl Phys Lett. 2007;90(26):263115–8.

16. Dey S, Drazin JW, Wang Y, Valdez JA, Holesinger TG, Uberuaga BP, et al. Radiation tolerance of nanocrystalline ceramics: insights from Yttria Stabilized Zirconia. Sci Rep. 2015;5:7746.

17. Zhang J, Lian J, Fuentes AF, Zhang F, Lang M, Lu F, et al. Enhanced radiation resistance of nanocrystalline pyrochlore Gd2(Ti0.65Zr0.35)2O7. Appl Phys Lett. 2009;94(24):243110–3.

18. Kuri G, Döbeli M, Gavillet D. Damage evolution and surface mor-phology of helium implanted yttria-stabilized zirconia single crys-tal. Nucl Instrum Meth B. 2006;245(2):445–54.

19. Taylor CA, Patel MK, Aguiar JA, Zhang Y, Crespillo ML, Wen J, et al. Bubble formation and lattice parameter changes result-ing from He irradiation of defect-fluorite Gd2Zr2O7. Acta Mater. 2016;115:115–22.

20. Taylor CA, Patel MK, Aguiar JA, Zhang Y, Crespillo ML, Wen J, et al. Combined effects of radiation damage and He accumulation on bubble nucleation in Gd2Ti2O7. J Nucl Mater. 2016;479:542–7.

21. Si S, Li W, Zhao X, Han M, Yue Y, Wu W, et al. Significant ra-diation tolerance and moderate reduction in thermal transport of a tungsten nanofilm by inserting monolayer graphene. Adv Mater. 2017;29(3):1604623–9.

22. Yu KY, Liu Y, Sun C, Wang H, Shao L, Fu EG, et al. Radiation damage in helium ion irradiated nanocrystalline Fe. J Nucl Mater. 2012;425(13):140–6.

23. Huang Z, Qi J, Zhou M, Gong Y, Shi Q, Yang M, et al. A facile sol-vothermal method for high-quality Gd2Zr2O7 nanopowder prepara-tion. Ceram Int. 2018;44(2):1334–42.

24. Wang XJ, Xiao HY, Zu XT, Zhang Y, Weber WJ. Ab initio molecu-lar dynamics simulations of ion–solid interactions in Gd2Zr2O7 and Gd2Ti2O7. J Mater Chem C. 2013;1(8):1665–73.

25. Renaud G. Oxide surfaces and metal/oxide interfaces stud-ied by grazing incidence X-ray scattering. Surf Sci Rep. 1998;32(1–2):5–90.

26. Toney MF, Huang TC, Brennan S, Rek Z. X-ray depth profiling of iron oxide thin films. J Mater Res. 2011;3(02):351–6.

27. Zhou LI, Huang Z, Qi J, Feng Z, Wu D, Zhang W, et al. Thermal-driven fluorite–pyrochlore–fluorite phase transitions of Gd2Zr2O7 ceramics probed in large range of sintering temperature. Metall Mater Trans A. 2015;47(1):623–30.

28. Mandal BP, Tyagi AK. Preparation and high temperature-XRD studies on a pyrochlore series with the general composition Gd2−

xNdxZr2O7. J Alloy Compd. 2007;437(1–2):260–3. 29. Han W, Demkowicz MJ, Mara NA, Fu E, Sinha S, Rollett AD, et

al. Design of radiation tolerant materials via interface engineering. Adv Mater. 2013;25(48):6975–9.

30. Bringa EM, Monk JD, Caro A, Misra A, Zepeda-Ruiz L, Duchaineau M, et al. Are nanoporous materials radiation resistant? Nano lett. 2012;12(7):3351–5.

31. Zhao Y, Grivel JC, Napari M, Pavlopoulos D, Bednarčík J, Zimmermann M. Highly textured Gd2Zr2O7 films grown on tex-tured Ni-5at.%W substrates by solution deposition route: growth, texture evolution, and microstructure dependency. Thin Solid Films. 2012;520(6):1965–72.

32. Wang SX, Begg BD, Wang LM, Ewing RC, Weber WJ, Kutty KV. Radiation stability of gadolinium zirconate: a waste form for pluto-nium disposition. J Mater Res. 1999;14(12):4470–3.

33. Dey S, Mardinly J, Wang Y, Valdez JA, Holesinger TG, Uberuaga BP, et al. Irradiation-induced grain growth and defect evolution in nanocrystalline zirconia with doped grain boundaries. Phys Chem Chem Phys. 2016;18(25):16921–9.

34. Bai XM, Voter A, Hoagland R, Nastasi M, Uberuaga BP. Efficient annealing of radiation damage near grain boundaries via interstitial emission. Science. 2010;327(26):1631–4.

35. Xiao HY, Zhang Y, Weber WJ. Ab initio molecular dynamics sim-ulations of low-energy recoil events in ThO2, CeO2, and ZrO2. Phys Rev B. 2012;86(5):054109.

36. Greenwood G, Foreman A, Rimmer D. The role of vacancies and dislocations in the nucleation and growth of gas bubbles in irradi-ated fissile material. J Nucl Mater. 1959;1(4):305–24.

37. Russell K. The theory of void nucleation in metals. Acta Metall. 1978;26(10):1615–30.

38. Beyerlein IJ, Caro A, Demkowicz MJ, Mara NA, Misra A, Uberuaga BP. Radiation damage tolerant nanomaterials. Mater Today. 2013;16(1):443–9.

39. Yang T, Taylor CA, Wang C, Zhang Y, Weber WJ, Xiao J, et al. Effects of He irradiation on yttria-stabilized zirconia ceramics. J Am Ceram Soc. 2015;98(4):1314–22.

40. Huang ZY, Ma NN, Qi JQ, Guo XF, Yang M, Tang Z, et al. Defect-fluorite Gd2Zr2O7 ceramics under helium irradiation: amorphiza-tion, cell volume expansion, and multi-stage bubble formation. J Am Ceram Soc. 2019;102(8):4911–8.

41. Qin W, Chen Z, Huang P, Zhuang Y. Crystal lattice expansion of nanocrystalline materials. J Alloy Compd. 1999;292(1–2):230–2.

42. Ackland G. Controlling radiation damage. Science. 2010;327(5973):1587–8.

How to cite this article: Huang Z, Zhou M, Cao Z, et al. He irradiation-induced lattice distortion and surface blistering of Gd2Zr2O7 defect-fluorite ceramics. J Am Ceram Soc. 2020;103:3425–3435. https ://doi.org/10.1111/jace.17030