Materials

The detailed development process for SIMP steel can be found in a reference [21]. The as-received steel samples were normalized for 0.5 h at (1040 ± 10) ℃ and tempered for 1.5 h at (760 ± 10) ℃ [21]. Their chemical composition (wt%, with Fe making up the balance) included Cr (10.24), Si (1.22), C (0.22), Mn (0.52), W (1.45), Ta (0.12), V (0.18), S (0.004), and P (0.004). The SIMP samples had a cross-section of approximately 7 × 15 mm². They were ground and polished to form a suitable surface for irradiation, and then annealed for 1 h at 600 ℃ in a vacuum at better than 1 × 10^-4 Pa to remove the work hardening surface layer and prevent it from influencing the subsequent nanoindentation test.

Irradiation experiments

Irradiation experiments were carried out at the 320 kV multi-discipline research platform for highly charged ions at the Institute of Modern Physics (IMP) in Lanzhou, China. Two irradiation modes were adopted for comparison, one was the sequential irradiation of He and H (He+H), and the other was the separate irradiation of He and H. Based on SRIM 2013 calculations in the “quick Kinchin-Pease calculation” mode [22, 23], 100-250 keV He⁺ and 60-130 keV H⁺ were chosen to produce an ion deposition plateau (PR) at 300-650 nm (Fig. 1). In Fig. 1, SR represents the surface region, which was easily influenced by surface treatment. TR refers to the track region [24]. In the SRIM calculations, the density of the target material was 7.87 g/cm³. The displacement threshold energy of key elements (Fe and Cr) in the steel was 40 eV [23]. For the sequential irradiation of He and H, the damage dose for H irradiation was equal to that for He irradiation, with values of 0.005, 0.025, and 0.125 dpa, and the concentration ratio of H to He was fixed at approximately 10 in the PR. For the separate irradiation of He and H, the damage dose for both H and He irradiation was 0.125 dpa in the PR. The detailed irradiation parameters are given in Table 1.

Fig. 1Full-screen View

(Color online) Depth profiles of (a) damage level and (b) H/He concentration in SIMP steel irradiated by He and/or H ions with multiple energy levels, according to SRIM 2013 calculations. An ion deposition plateau was produced at approximately 300–650 nm.

Spherical nanoindentation

All of the irradiated samples were tested using a Nano Indenter G200 in the continuous stiffness measurement (CSM) mode at room temperature. A spherical indenter with a radius of 5 µm was used in the tests. The oscillation frequency and amplitude were 45 Hz and 2 nm, respectively. A 2 × 5 indentation array with 70 µm spacing between adjacent indents was created on each specimen to eliminate the influence of the grain orientation. The maximum penetrating depth of the indenter was 2 µm, covering the whole irradiated region (see Fig. 1).

The protocol used in the spherical nanoindentation data analysis (including an effective zero-point correction) was based on Hertz's model [25]. For the frictionless contact between two linear isotropic elastic solids, the effective modulus of the indenter and specimen system, E_eff, can be described as follows: $\frac{1}{{\text{E}}_{\text{eff}}}=\frac{1-{\nu }_s^2}{{\text{E}}_{\text{s}}}+\frac{1-{\nu }_i^2}{E_i}.$ (1) Here, ν and E refer to Poisson's ratio and Young’s modulus, respectively, while subscripts s and i represent the specimen and indenter, respectively. In this test, Poisson’s ratios for the specimen (υ_s) and diamond indenter (υ_i) were 0.29 and 0.07, respectively. Young’s modulus for the indenter (E_i) was 1140 GPa.

The indentation stress, σ_ind, and strain, ε_ind, can be defined as follows [13, 16, 26]: ${\sigma }_{\text{ind}}=\frac{P}{\pi a^2},$ (2) ${\epsilon }_{\text{ind}}=\frac{h}{2.4a}.$ (3) Here, P and h denote the load and displacement, respectively. It is important to note that both P and h have already been zero-point corrected by the method reported in a reference [16]. The contact radius, a, is determined from the CSM stiffness, S, and effective modulus, E_eff, in Eq. (4) [16]. $\text{a}=\frac{S}{2E_{eff}}$ (4)

Combined with the above equations, the indentation stress–strain curve can be obtained by converting from the load–displacement curve.

Microstructure characterization

Indentation stress–strain response

The spherical indentation stress–strain curves were obtained according to Eqs. (3) and (4), as shown in Figs. 2a and b. The spherical indentation stress–strain relationship of an irradiated sample with a higher damage dose (such as sample #3) could be divided into three segments, including the initial hardening, “strain softening (where the indentation stress decreased with the indentation strain),” and working hardening [28, 29]. In the initial hardening segment, the slope of the linear elastic stage represents the indentation modulus (E_ind), which can be calculated to be 183 GPa using Eq. (2), where Young's modulus of the sample (E_s) is approximately 200 GPa, as obtained by the indentation experiments. The yield stress (σ_y) refers to the indentation stress at the beginning of plastic deformation. In the “strain softening” segment, σ_s and σ_e represent the indentation stress at the start and end of the “strain softening,” respectively. After σ_e, the indentation stress gradually increases with the strain as a result of the effect of working hardening.

Fig. 2Full-screen View

(Color online) Spherical indentation stress–strain curves of H/He irradiated samples. (a) Dose effect of He and H sequentially irradiated samples. (b) Comparison of the He and H sequentially and separately irradiated samples. The curves can be divided into three segments: (Ⅰ) the initial hardening segment, (Ⅱ) “strain softening” segment, and (Ⅲ)working hardening segment.

Figure 2a shows the spherical indentation stress–strain curves of the samples sequentially irradiated with He and H at different irradiation doses. The indentation stress values (σ_y, σ_s, and σ_e) for all the samples are given in Table 2. Compared with the unirradiated sample, the yield stress (σ_y) values of samples #1 and #2 hardly changed, but that of sample #3 increased by 1 GPa. Meanwhile, “strain softening” was observed in sample #3. Figure 2b shows a comparison of the spherical indentation stress–strain responses of the samples subjected to sequential and separate irradiation with He and H, when the irradiation doses of both H and He were 0.125 dpa in the PR. Combined with the stress values, σ_s and σ_e, given in Table 2, it shows that the yield stress (σ_y) was greater and the “strain softening” was more remarkable after the sequential He and H irradiation, compared with those for single H/He irradiation.

Irradiation-induced micro-defects

3.2.1

Vacancy-type defects

In DBS measurements, the concentration and size of vacancy-type defects can be characterized by the S parameter because of its sensitivity to low-momentum electrons (that is, valence electrons) around vacancies [30]. In this study, we used the relative change in the S parameter after irradiation (△S/S) [27] to determine the vacancy-type defects caused by irradiation.

Figure 3a shows the △S/S values versus the depth after the sequential irradiation of He and H at different irradiation doses. When the damage level induced by H/He in the PR increased from 0.005 (sample #1) to 0.025 dpa (sample #2), △S/S showed little change, but when it increased to 0.125 dpa (sample #3), △S/S increased significantly. This revealed that when the damage level was high enough, the formation of vacancy-type defects was more obvious.

Fig. 3Full-screen View

(Color online) (a) Depth profiles of △S/S after sequential irradiation of He and H at different damage levels induced by H/He (0.005 dpa for sample #1, 0.025 dpa for sample #2, and 0.125 dpa for sample #3) in plateau region. (b) Comparison of △S/S values after He and H sequential irradiation and separate irradiation when the damage level induced by H/He was 0.125 dpa in the plateau region.

Figure 3b shows a comparison of the △S/S values after the sequential (sample #3) and separate irradiation of He and H (samples 4# and #5). In the SR, there was no difference in △S/S. However, in the TR and PR, the △S/S value after He and H sequential irradiation (sample #3) was larger than that for single He irradiation (sample #5), but lower than that for single H irradiation (sample #4). This shows that the S parameter was influenced by gas atoms (He/H). It has been reported that both He and H prefer to occupy a region with low electronic density, such as a vacancy [31]. Once occupied by He or H, the vacancy will become less effective in the trapping of positrons, resulting in a decrease in the S parameter [32-34]. Furthermore, the effect of He on the S parameter is more significant than that of H [24, 33, 35]. The preferential occupation of a vacancy by He followed by H (this could suggest the formation of He-V-H defect clusters) led to a decrease in the S parameter after He and H sequential irradiation.

3.2.2

Bubbles

Figure 4 shows the TEM morphology of the bubbles distributed in the PR for H/He irradiated samples obtained using the under-focus condition (approximately 300 nm). It should be noted that visible bubbles were only formed in sample #3, showing the promoted formation of bubbles by the sequential irradiation of He and H.

Fig. 4Full-screen View

TEM morphology of bubbles distributed in PR for H (sample #4), He (sample #5), and He+H (sample #3) samples. Photographs were taken in the bright-field under-focus (approximately 300 nm) mode.

3.2.3

Dislocation loops

Fig. 5 shows that a higher density of black-dots existed in sample #3, compared with samples #4 and #5. Two kinds of dislocation loops are generally formed in Fe-based alloys during irradiation, with Bragger vectors b = 1/2<111> and b = <100> [36, 37]. The ratio of 1/2<111> to <100> loops depends largely on the irradiation temperature and dose [38, 39]. At a low temperature and low dose, irradiation-induced dislocation loops are predominantly small 1/2<111> loops formed by a coalescence of self-interstitial atoms (SIAs) [40], appearing as “black-dots” under TEM observation.

Fig. 5Full-screen View

TEM morphology of small interstitial dislocation loops (black dots) distributed in PR for H (sample #4), He (sample #5), and He+H (sample #3) samples. Photographs were taken under two-beam conditions with g = (0–11).

Spherical nanoindentation stress–strain response

An important phenomenon for materials that have been irradiated with a higher dose (such as sample #3) is an increase in the indentation yield stress, which can be considered to be a manifestation of irradiation hardening [28, 41]. Irradiation hardening in metals or alloys can be described by the Orowan hardening model [42]. According to this model, the irradiation hardening is ascribed to dislocation pinning by irradiation-induced defect clusters. Meanwhile, the change in the uniaxial yield stress (∆σ_uy) induced by irradiation can be quantified by the following equation: $\Delta {\sigma }_{uy}=M\sigma \mu b\sqrt{Nd},$ (5) where M is the Taylor factor (approximately 0.36); σ represents the strength factor of the specific barrier for dislocation slipping (0.2 for black dots, 0.1 for voids or bubbles, 0.2 for dislocation lines) [42]; u is the shear modulus of the material (80 GPa for bcc-Fe); b is the Burgers vector (0.248 nm for 1/2<111> dislocation loops in bcc-Fe) [43]; and N and d refer to the density and mean diameter of the specific barrier, respectively.

Based on TEM observations, the densities and mean diameters of the bubbles and black dots in the H/He irradiated samples are given in Table 3. Correspondingly, the increase in the uniaxial yield strength (σ_uy) can be calculated by Eq. (6). It has been shown that the indentation yield strength is approximately double the uniaxial yield strength (σ_uy) [20, 44, 45]. Accordingly, the increase in the indentation yield strength can be converted to that of the uniaxial yield strength. A comparison of the uniaxial yield strength from the Orowan hardening model with that from the spherical indentation showed that the former was lower than the latter. This discrepancy was attributed to the indentation size effect (ISE). For spherical indenters, the ISE is more significantly dependent on the indenter size rather than the indentation depth [46, 47]. In particular, as the radius of the sphere decreases, the hardness increases. For indenter radii of 100 μm and less, the indentation stress-strain curves shift upward with the decreasing radii and deviate measurably from the macroscopic stress-strain behavior [46]. In this study, the radius of the spherical indenter was 5 μm (less than 100 μm). Therefore, the yield stress from indentation was higher than the uniaxial yield strength from the uniaxial tensile test or Orowan model.

Another phenomenon is the appearance of “strain-softening” after the yield point. This “strain-softening” is not related to the material deformation or hardening behavior, but is largely attributable to the softening effect induced by the substrate [17, 28, 29]. The effect of a softer substrate is related to the size of the plasticity affected region (PAR). The volume of the PAR can be calculated as follows [48]: $V=2/3\pi {\left(R_{\text{PAR}}\right)}^3,$ (6) where R_PAR refers to the radius of the plastic zone, which can be approximated by R_PAR ≈ fa, where f is the proportional coefficient.

The indentation stress increased with the strain before reaching the indentation stress at the onset of “strain-softening” (σ_s). In the meantime, the PAR did not exceed the damage layer. As the PAR extended toward the softer substrate beneath the damage layer, the hardness appeared to decrease. Figure 2 shows that the indentation strain at the onset of “strain-softening” was 0.04, and the corresponding contact radius was 498 nm, as calculated by Eq. (4). According to the suggestion by Chioct et al. that f = 1.44 [49], the radius of PAR (R_PAR) could be calculated to be 717 nm. This value was approximately equal to the maximum depth of the damage region obtained by SRIM calculations.

The appearance of “strain-softening” is also associated with the ISE [46-48, 50]. On one hand, the hardness decreases with the increasing penetration depth of the indenter (i.e., the indenter radius) in the initial hardening segment. On the other hand, for smaller indenter radii, the ISE shifts the indentation stress-strain curves upward [46], resulting in a higher indentation yield stress and critical indentation stress at the onset of “strain-softening” (σ_s).

H/He synergistic effects on mechanical properties

Compared with single He/H irradiation, a higher indentation yield stress was observed after He and H sequential irradiation, indicating that enhanced hardening had occurred as a result of the He and H irradiation. Based on triple-beam (Fe+He+H) irradiation experiments, Lee and Hunn et al. reported similar effects in F/M steels and stainless steels [7, 9]. They considered that the enhancement of the hardening by the synergy of H and He was due to the dislocation pinning by bubbles and dislocation loops. The production of dislocation loops resulted from the punching of over-pressurized bubbles or an accumulation of surplus SIAs left over during the formation of gas-vacancy clusters. They also speculated that the reason was related to the trapping of hydrogen by helium clusters. The present study confirmed that the formation of both bubbles and dislocation loops was enhanced by the synergy of He and H irradiation.

An atomistic simulation indicated that the trapping of hydrogen by helium bubbles is more beneficial for the production of dislocation loops by the “loop punching” of He bubbles [6]. More specifically, there are two kinds of primary crystal defects, namely vacancies and SIAs (Frank pairs), produced in the early stages of an irradiation event. At room temperature, the formation of defect-clusters is primarily determined by the fate of SIAs because of their higher mobility compared to vacancies [51]. A migrating interstitial atom could encounter a vacancy and they could annihilate each other, or it could encounter other interstitial atoms and form interstitial clusters and dislocation loops, or it could be absorbed by a defect sink (such as a grain boundary, precipitate, or dislocation). However, these processes are influenced by transmutation gas atoms (H/He). He can easily be trapped by vacancies, forming stabilized He-vacancy complexes, and leading to the decreased possibility of SIAs recombining with these vacancies. When the concentration of He aggregated in the vacancy clusters is high enough, the lattice atoms around the vacancy clusters will be emitted into the bulk as new SIAs or absorbed by interstitial dislocation loops, accompanied by the growth of bubbles. Furthermore, some studies [6, 31, 52] have shown that a He-vacancy cluster (or He bubble) is a strong trap for H. The binding of H and He-vacancy clusters is more favorable for vacancy-interstitial pair production. This means that the co-presence of H and He can lead to a lower possibility of SIAs recombining with vacancies and a stronger “loop punching” effect.

Compared to triple-beam irradiation, double-beam (He+H) irradiation at room temperature is closer to the special condition of a continuous helium but insufficient vacancy supply because of higher gas/dpa rates [53]. Therefore, it can be inferred that “loop punching” might have been a vital mechanism for the nucleation or growth of both the bubbles and dislocation loops in sample #3. In addition, a higher H to He concentration ratio (10:1) suggests that the contribution of H to “loop punching” was more outstanding because of the trapping of H by He bubbles.

In summary, the co-presence of H and He was more beneficial for the nucleation and growth of vacancy-type (bubbles or voids) and interstitial-type defect clusters (I-loops), resulting in more severe hardening.