He bubble evolution

A cross-sectional transmission electron microscopy (XTEM) image of irradiated sample S1 (5 × 10¹⁵ions/cm²) is displayed in Fig. 2(a). Some grain boundaries and dispersed TiC nanoparticles were observed. The SAED pattern in the damage band is shown in the inset of Fig. 2(a), indicating that the crystal structure of sample S1 remained unchanged during irradiation. Fig. 2(b) presents a magnified TEM image of the peak damage zone marked by the red box in Fig. 2(a). Many white defects were observed, which were identified as He bubbles via through-focus experiments. Figs. 2(c) and (d) display high-magnification underfocused and overfocused TEM images, respectively, from the other peak implantation zone. It can be seen that the He bubbles in Fig. 2(c) changed into black dots in the overfocused condition (Fig. 2(d)).

Fig. 2Full-screen View

a XTEM image of sample S1 irradiated by 5 × 10¹⁵ He ions/cm², the inset displays the SAED pattern of the peak damage zone. b The zoomed TEM image of the region marked by the red box. c and d the underfocus and overfocus TEM images of the peak damage zones

An XTEM image of sample S2 irradiated at 5 × 10¹⁶ ions/cm² is shown in Fig. 3(a). The overall He bubble band is shown in Fig. 3(b). Compared to sample S1, the density of He bubbles in the damage band of sample S2 increased significantly. However, the size of the bubbles decreased slightly. The average size of the He bubbles in the peak damage region appeared almost unchanged, as shown in Fig. 3(c).

Fig. 3Full-screen View

a XTEM image of sample S2 irradiated by 5 × 10¹⁶ He ions/cm², the inset displays the SAED pattern of the peak damage zone. b The TEM image of He bubble region marked by the red box. c Enlarged TEM image in peak damage zone

Fig. 4a shows an XTEM image of sample S3 irradiated at 1 × 10¹⁷ ions/cm². The enlarged TEM image of zone B in sample S3 is shown in Fig. 4(b). He bubbles formed at two grain boundaries. Fig. 4(c) displays the He bubble bands in the matrix, which are indicated by red box A. The corresponding high-magnification TEM image near the peak damage region is shown in Fig. 4(d). When the depth exceeded the peak damage zone, the density of bubbles decreased, and their size increased.

Fig. 4Full-screen View

XTEM image of the sample S3 irradiated by 1 × 10¹⁷ He ions/cm², the SAED pattern of the peak damage zone was displayed in the inset; b Amplified TEM image of zone B; c and d enlarged TEM image of zone A

The He bubbles that formed in the peak damage regions in the Ni matrix of the samples irradiated at different He ion doses are displayed in Fig. 5 for direct comparison. It can be clearly observed that the density and size of the He bubbles varied with the dose. When the ion fluence was increased from 5 × 10¹⁵ to 5 × 10¹⁶ ions/cm², the size decreased slightly; however, the bubble density increased significantly. When the fluence was further increased to 1 × 10¹⁷ ions/cm², larger bubbles formed, and the density showed a slight reduction.

Fig. 5Full-screen View

Bright field TEM images of He bubbles at the peak damage zones of the samples: a S1, b S2 and c S3

The number density and mean He bubble size for the three irradiated samples were calculated by the “Nano Measurer” software, as shown in Fig. 6(a). The number densities of bubbles in the samples (S1, S2, and S3) were approximately (3.56 ± 0.22) × 10²², (29.3 ± 0.62) × 10²², and (19.9 ± 0.51) × 10²² m^-3, respectively, and the corresponding mean diameters were approximately 2.85 ± 0.74, 2.44 ± 0.79, and 3.75 ± 1.22 nm, respectively. Clearly, the number density and size of He bubbles varied with the ion fluence. It is known that implanted He atoms can aggregate to form He bubbles. Furthermore, the implantation of energetic He ions results in the displacement of Ni atoms, which then collide with other Ni atoms to form displacement cascades. During this process, vacancies and interstitials are generated. In this study, when the dose was low, a low density of He bubbles formed. These bubbles were not equally distributed in the grains. They tended to be connected by a series of lines, which may be related to the pre-existing dislocation lines in the Ni–0.05TiC_NP composites. He atoms can become trapped at dislocations to form He bubbles ^[22]. At a low dose (5 × 10¹⁵ ions/cm²), the implanted He atoms would preferentially nucleate at dislocation lines and aggregate to form He bubbles. He atoms can also be trapped by displacement damage-induced vacancy clusters to form nucleation sites for He bubbles. When the He ion dose was increased to 5 × 10¹⁶ ions/cm², more He atoms and vacancies were generated, which could provide considerably more nucleation sites for He bubbles. Therefore, the number density of He bubbles in sample S2 increased significantly. Although the total number of He atoms increased, they were dispersed by He–vacancy complexes and the pre-existing dislocations, which caused the average size to decrease slightly from 2.85 ± 0.74 nm to 2.44 ± 0.79 nm. When the He ion dose was further increased to 1 × 10¹⁷ ions/cm², He bubbles became dense, and some bubbles gathered to form large bubbles through the “Migration and Coalescence” (MC) mechanism ^[9], causing the disappearance of some bubbles and a slight decrease in their number density.

Fig. 6Full-screen View

a The mean size and number density and b volume fraction of He bubbles in the peak damage regions of samples irradiated to different ion fluences

It is widely reported that He bubbles can cause swelling of Ni-based alloys ^{[10, 11, 13, 23, 24]}. For example, Lei et al. used atomic force microscopy (AFM) to investigate the swelling of a Ni–Mo–Cr alloy irradiated with 1.2 MeV He ions at 650 °C ^[10]. They found that a swelling of 2.67% was induced by He bubbles in the irradiated sample (3 × 10¹⁷ ions/cm², 6.18 dpa), which was larger than that of Ni irradiated to 13 dpa by Ni ions. The swelling can be calculated using the bubble size and number density ^[25]:

where $r$ is the equivalent radius, and $N_{\text{d}}$ is the number density of the bubbles. The corresponding volume expansion caused by He bubbles in the Ni grain were estimated to be 0.043%, 0.223%, and 0.549% for the samples irradiated at 5 × 10¹⁵, 5 × 10¹⁶, and 1 × 10¹⁷ ions/cm². Although the number density tended to saturate at a relatively high dose, the total swelling increased continuously with increasing He ion fluence, as shown in Fig. 6(b).

He bubbles formed at the grain boundaries in sample S3, as shown in Fig. 4(b). Furthermore, He bubbles were found at the grain boundaries in all the samples irradiated at different doses, as shown in Fig. 7. For sample S1 irradiated at 5 × 10¹⁵ ions/cm², the mean He bubble sizes in the grain boundary and matrix were similar (Fig. 7(a)). As shown in Figs. 7(b) and (c), the bubbles in the grain boundaries were noticeably larger than those in the grains for samples S2 and S3, indicating the preferred formation of He bubbles at the grain boundaries. These He bubbles on the grain boundaries could cause the so-called “He embrittlement” of the alloy, which would degrade the mechanical properties of the structural material in MSRs and should be considered in the design.

Fig. 7Full-screen View

He bubbles in the grain boundaries in samples a S1, b S2 and c and d S3

Fig. 7(d) shows an enlarged TEM image of He bubbles that formed in the Ni matrix and at the grain boundary in sample S3 irradiated at 1 × 10¹⁷ ions/cm². The MC and “Ostwald Ripening” (OR) mechanisms could explain the growth of He bubbles ^{[26, 27]}. When the concentration of He atoms is high, the MC mechanism is dominant, which requires the motion and contact of He bubbles. The He ion fluence in sample S3 was 1 × 10¹⁷ ions/cm², and the He concentration in the peak region reached 42750 ppm, which could provide enough He atoms to form a high density of bubbles and enhance the possibility of bubble migration and coalescence. He bubbles at the grain boundary could encounter and contact one another, as shown in Fig. 7(d). In particular, the two necked He bubbles indicated by an arrow were in the process of merging. These features indicate that the MC mechanism can explain the growth of He bubbles in this study. For the OR mechanism, He atoms and vacancies released from small dissociated bubbles could be absorbed by large bubbles, and thus larger bubbles could be formed ^{[26, 27]}. It has been reported that the OR mechanism is dominant at high temperatures or low concentrations of He atoms ^{[26, 27]}. The irradiation temperature (700 °C) was greater than 0.5 T_m of this alloy; therefore, the dissociation of He bubbles was possible. For sample S1 irradiated at a low dose, the OR mechanism may have been dominant. For samples S2 and S3, the MC and OR mechanisms could both contribute to the growth of He bubbles. Some large ellipsoidal He bubbles elongated along grain boundaries, which resulted from the compressional deformation of bubbles by both sides of the grains during their growth process, were also observed.

It was also noted that the He bubbles in the three irradiated samples had different shapes, which varied with the bubble size. The He bubbles in samples S1 and S2 were relatively small and spherical, as shown in Figs. 5(a) and (b). Both spherical and polygonal He bubbles were observed in sample S3, as shown in Figs. 5(c) and 7(d). The shape of a He bubble is related to its pressure ^[28]. The internal pressure in a He bubbles is inversely proportional to the bubble size according to the equilibrium bubble pressure equation ^[28]:

where γ and $r$ are the surface energy of the material and radius of the He bubble, respectively. When the bubble is small, the pressure is relatively large, and spherical He bubbles form easily. When the bubble radius increases, the pressure decreases, resulting in the transformation of He bubbles from stable spherical to polygonal shapes ^[24].

In the last several years, many types of oxide dispersion-strengthened (ODS) alloys, such as ferritic–martensitic ODS steels ^[29] and ODS-W alloys ^[30], have been developed to enhance the irradiation resistance of structural materials in fission and fusion reactors. The dispersed oxide nanoparticles were found to act as He bubble-trapping sites ^{[31, 32]}. Similarly, it has been reported that dispersed SiC nanoparticles can inhibit bubble growth and thus reduce the swelling of Ni-based alloys ^[11], which is attributed to the preferential diffusion of He atoms to the SiC–Ni interface, as revealed by density functional theory (DFT) calculations ^[11]. However, no He bubbles were observed at the SiC–Ni interfaces in these studies by TEM. A possible reason is that the size of the He bubbles that formed at these interfaces was too small to be visible. In this study, there were numerous interfaces between the TiC nanoparticles and Ni matrix. In particular, He bubbles were observed at the interfaces between the TiC nanoparticles and Ni matrix, as shown in Fig. 8. Furthermore, it can be seen from Figs. 8(a) and (b) that He bubbles also formed in the interior of the TiC nanoparticles, some of which were even larger than the He bubbles in the Ni grains (Fig. 8(b)). It is proposed that the TiC particles could delay the nucleation and growth of He bubbles in the grains and grain boundaries by confining He atoms to the TiC–Ni interfaces and the interior of the TiC nanoparticles.

Fig. 8Full-screen View

He bubbles trapped in the TiC_NP and the interfaces between TiC_NP and Ni matrix in samples: a S1, b and c S3

Nanoindentation tests

Nanohardness was measured using nanoindentation to study the influence of He irradiation on the hardness of the samples. It is noted that the high-temperature annealing during irradiation could influence the hardness of the irradiated sample. Thus, to eliminate this interference, the nanohardness of both the unirradiated and irradiated zones of the samples was measured for direct comparison. Figs. 9(a)-(c) display the variation in the average nanohardness with depth of the unirradiated and irradiated regions in samples S1, S2, and S3, respectively. The corresponding standard deviations are also included. The hardness data for indentation depths less than 100 nm are not shown in Fig. 9 because of inaccuracies in the data due to high uncertainty. The curves demonstrate that the hardness decreases with increasing depth because of the indentation size effect (ISE) ^{[12, 13, 24, 33, 34]}. Surprisingly, for each sample, the nanohardness of the irradiated region was lower than that of the unirradiated region, exhibiting an uncommon irradiation-induced softening phenomenon, as shown in Figs. 9(a)-(c). This is quite different from the He irradiation-induced hardening in Ni-based alloys, such as GH3535 ^[10], Hastelloy N ^[12], pure Ni, and SiC nanoparticle-reinforced Ni alloys ^[24]. He ion irradiation-induced hardening of Ni-based alloys was reported in all these studies, and the nanohardness increased with increasing irradiation ion fluence. This type of hardening was attributed to He bubbles and/or dislocation loops. However, although He bubbles formed in this study, the nanohardness of the irradiated samples did not increase but decreased.

Fig. 9Full-screen View

The average nanoindentation hardness with the depth of unirradiated and irradiated sections of samples a S1, b S2 and c S3

To investigate the possible reason for the He irradiation-induced softening, we compared the dislocation microstructure of the unirradiated and irradiated regions of the He-irradiated Ni–TiC_NP composite. Fig. 10(a) displays typical bright-field TEM images of the interface between the bubble region and bubble-free region in sample S2 taken under the two-beam condition (Z = ［011］ and g = $11\overline{1}$). In the magnified TEM image of the bubble-free region (Fig. 10(b)), numerous dislocations can be observed. The evolution of these pre-existing dislocation lines can affect the hardness of the material. In the He bubble region (Fig. 10(c)), the dislocation density decreased, and it appears that some dislocation lines were eliminated by He bubbles. Similarly, molecular dynamics (MD) simulations indicated that the intrinsic dislocations in single-crystal α-Fe would move under continuous He ion irradiation ^[35]. The disappearance of intrinsic dislocations in Ni-based alloys induced by Xe ion irradiation was also reported by Chen et al. ^[36]. We reasonably conclude that the disappearance of pre-existing dislocation lines in the sample resulted from He ion irradiation; however, direct evidence is still missing. In situ TEM observations using He ion irradiation performed within a TEM, which can directly track the growth of He bubbles and the interaction between dislocations and irradiation-induced defects, would be helpful to clarify the interaction between He bubbles and pre-existing dislocation lines in this material ^[37-39]. When the irradiation dose was low (5 × 10¹⁵ ions/cm²), He bubbles were preferentially distributed along a line in the Ni matrix (Figs. 2 and 5), which may have resulted from the interaction between He bubbles and the intrinsic dislocation lines. It was reported that He atoms can be more easily trapped and segregated on dislocations lines ^[35]. He bubbles formed preferentially along the dislocation lines. During this process, some of the pre-existing dislocation lines disappeared. Thus, the decrease in hardness may be attributed to the disappearance of the intrinsic dislocations. On the one hand, He bubbles increase the hardness of irradiated samples. On the other hand, the disappearance of the intrinsic dislocations induced by He ion irradiation would decrease the hardness, which exceeds the increment of hardness contributed by He bubbles. Overall, the He ion irradiation induced a softening phenomenon in the Ni–TiC_NP composite.

Fig. 10Full-screen View

Bright field TEM images taken under two-beam condition (g=$11\overline{1}$ and Z = [011]). a Low-magnification TEM image of the interface between He bubble-free zone and bubble zone, b the zoomed TEM image of the bubble-free region and c the zoomed TEM image of the bubble region