**4. Results**

#### *4.1. PLEPS Measurement*

PLEPS measurement of one investigated specimen was used to confirm the ability of protons to deliver significant lattice damage and observation of the defect depth profile up to 520 nm. The specimen in the as-received state and after the implantation were compared in terms of defect size and defect presence in the surface and subsurface layers.

Figure 4 shows that positron lifetimes in bulk (LT1) of the as-received specimen remained approximately at the same level through all measured depths, which is in perfect agreemen<sup>t</sup> with expectations because positron annihilation in bulk should remain the same independently of the state of the material. However, LT1 for proton-irradiated specimens compared to the non-irradiated ones increased from 140 ps to 160–170 ps. A slight increase of the positron lifetime LT1 in our measurements compared to positron lifetime LT1 in pure Fe (~107 ps [95]) might be attributed to the fact that these spectra were also decomposed into two components instead of the usual three components; thus, the DTM model was applied.

Because of the two components' decomposition, we believe the evaluation software attributed some annihilations to minor lattice defects such as dislocation and monovacancies to the annihilation in bulk, which resulted in the increase of LT1. The same explanation applies also to the ~20–30 ps difference between the non-irradiated and the irradiated states—the presence of larger defects in lattices after proton irradiation led to

their attribution to annihilation in bulk and therefore the increase of LT1 compared to the as-received state.

The intensity of annihilation in bulk (I1) for the depth range of 150–520 nm for the as-received state remains constant at values around 90%; only for depths of 50 and 90 nm is I1 is somewhat lower. This decrement is naturally accompanied by a proportional increase in the intensity of annihilation in defects (I2), as seen in Figure 5. This can be attributed to the fact that positrons after implantation and thermalization move randomly, and during this process, they can also return to the entrance surface.

**Figure 5.** Positron lifetimes (LT2) and intensity of annihilation (I2) in defects for various energies of incident positrons (penetration depth) for both non-irradiated and proton-irradiated specimens.

"Positrons can be trapped at the surface where the electron density is lower than in the bulk. Additionally, they can be trapped at surface defects or positronium, i.e., a bound state between positron and electron can be created" [96]. Another factor that might be contributing is residue defects from specimen preparation—ones that remained even after polishing. All of this causes the I1 value to be lower and the I2 value to increase at depths closer to the surface than in the bulk (Figure 5). This effect is even more significant for proton-irradiated specimens. As seen in Figure 4, I1 decreases from ~95% for depths deep in the bulk to 45% for depths close to the surface (also accompanied by a proportional increase in I2 in Figure 5). We believe that this significant decrease in I1/increase in I2 for proton-irradiated specimens was caused mainly by storage of the specimens (creation of an oxidation layer on the surface, see below) and by manipulation and transportation of specimens during proton irradiation and PLEPS measurements (surface damage).

As seen from Figure 5, positron lifetimes in defects (LT2) for depths of 50, 100, and 150 nm is 20–40 ps higher for the proton-irradiated specimens compared to the as-received state. At depths this close to the surface, the possibility that this disproportion would be created by proton irradiation is excluded because, as shown by SRIM simulations, the lattice damage starts at depths of ~100–150 nm (Figure 5). This increase in LT2 was probably caused by the creation of an oxidation layer on the surface of the specimen. The assumption of the oxidation layer creation is supported by the results from other papers focused on the study of corrosion-related defects by slow positron beam. These papers reported similar values of LT2 in the surface oxidation layer [97,98]. However, the creation of the surface oxidation layer was expected because the period between specimen preparation and PLEPS measurements of proton-irradiated specimens on this occasion was prolonged because of various problems with the FRM-II reactor to more than two years. Figure 5 also shows that the surface oxidation layer reaches a depth of around 200 nm, where the positron lifetime in defects (LT2) for proton-irradiated specimens exhibits the same values as for the non-irradiated state (without the oxidation layer).

For depths starting at ~270 nm, a progressive increase in LT2 for the proton-irradiated state is clear compared to the non-irradiated state. This progressive increase in LT2 is in perfect agreemen<sup>t</sup> with SRIM simulations because, as seen in Figure 2, the number of vacancies produced starts to increase at this depth. At maximum depth, LT2, reaches ~460 ps, which indicates the presence of vacancy clusters containing more than 40 vacancies [99]. It is important to note that the maximum depth of positrons reached by PLEPS is, as seen in Figures 4 and 5, around 0.52 μm, but according to SRIM simulations, the maximum lattice damage is at 2.3 μm; therefore, even bigger vacancy clusters are expected to form deeper in the material. Unfortunately, we were unable to probe deeper into the specimen because 0.52 μm is the maximum depth that can be reached by positrons in PLEPS. By using 500 keV protons, we hit the spot where this energy is too low for effective usage of conventional PALS and simultaneously too high for PLEPS to see the overall lattice damage. Nevertheless, PLEPS data confirmed that lattice damage was introduced to the material by proton irradiation.

#### *4.2. CDB Measurement*

The CDB measurements using standard positron energy spectra were preliminarily applied only for five as-received and five implanted specimens as confirmation of PLEPS results and the test of sensitivity to the observed defects by a conventional positron source. The CDB results characterizing the material at a depth of 150 μm are shown in the S-W diagram (Figure 6), where the W parameter describes positron annihilation with core electrons (annihilation in bulk) and the S parameter with valence electrons located mostly in open volume (annihilation in vacancy defects). These two parameters are extracted from each spectrum, and each carries different information about the measured materials [4].

As seen in Figure 6, proton-irradiated specimens exhibit a significant increase in the S parameter and a decrease in the W parameter compared to the non-irradiated specimens. The total amount of annihilating positrons is fixed, and in general, an S parameter increase is necessarily accompanied by a W parameter decrease. The S parameter can be affected by both the number density and the size of vacancy-type defects [100], so in our case, this enhancement of the S parameter indicates defect accumulation in the structure due to the implantation. These results are in good agreemen<sup>t</sup> with the PLEPS results above, and they also indicate that conventional positron sources can be useful for the detection of implantation-accumulated defects in our structure.

**Figure 6.** W parameter as a function of the S parameter for the as-received and proton-irradiated state.

It is important to note that the data's variance for all as-received specimens is 0.35%, and for all irradiated ones, it is 0.11%, i.e., the scattering of the measured data is very low. This proves that the results are not very sensitive to material inhomogeneity, clusters of carbides, disparities in specimen preparation, or inaccuracies during experimental measurements and during proton irradiation of the specimens. Therefore, the standard trapping model [78] can be applied for the evaluation of the PALS spectra.

The average S parameter of the as-received specimens is 0.6108 ± 0.0007, and the average W parameter is 0.24998 ± 0.0007. The average CDB parameters changed due to the implantation, and they are S = 0.617 ± 0.0003 and W = 0.2444 ± 0.0003. The implantation caused the S parameter to increase about 1.01 times, which indicates an accumulation of very small defects (mono- or di-vacancies).

#### *4.3. PALS Measurement*

All five investigated specimens were observed by the PALS technique in the asreceived state, after the implantation, and after the gradual annealing. They showed good homogeneity in structure, and the response to the experimental loads was also similar, as can be seen in Figure 7. Specimen no. 5 is measured only up to the annealing temperature of 500 ◦C due to the specimen sustaining mechanical damage during the experiment.

PALS data were evaluated in the program LT9 by the standard trapping model [90] and separated into two or three components including lifetimes (LT) and intensities (I). For the first lifetime (LT1), the bulk was fixed to 100 ps (calculated values for a pure iron range from 97 ps [95] to 110 ps [48]) for better comparison of the second components for all measurements. The second component involves small defects and can describe their size as proportional to a lifetime (LT2, mostly dislocations and vacancies) and their concentration which is proportional to the intensity (I2). The third component (LT3, I3) was found only for implanted specimens and annealed specimens at temperatures out of range (300–500) ◦C, where we expected the most visible change in structure. The LT3 was on average 1.4 ± 0.5 nm with I3 of approx. 1% and describes annihilation in the source or infight annihilation not fully compensated during the process of fitting.

The LT2, I2, and positron Mean Lifetime values (MLT, describing the whole spectra while minimizing the influence of the data treatment) are presented in Figure 7 for all measurements. For the individual treatments of specimens, the average of LT2 and I2 as well as their standard deviations were calculated considering the uniformity of the specimens and their good structural homogeneity. These average values are described below in more detail.

**Figure 7.** PALS data for as-received, implanted, and annealed specimens: Lifetime 2 proportional to the size of the defects; Intensity 2 proportional to defect concentration; and Mean lifetime describing the defect presence in the specimens.

In the as-received state, the average LT2 = 158 ps ± 3 ps probably represents dislocations together with the smaller presence of mono-vacancies. The average intensity of as-received specimens was 59.7% ± 2.5%. The Mean Lifetime for the as-received specimens

was between 134 and 137 ps, and the average value was 135 ps ± 2 ps. It is the lowest value for all our measurements, which indicates the lowest total defect volume (Figure 8), as was expected.

**Figure 8.** Average values of defect concentration and defect volume for the as-received, implanted, and annealed specimens at 200, 300, 400, 450, 475, 500, 525, and 550 ◦C.

After the implantation, LT2 increased, with the average value of 188 ± 3 ps representing a mixture of mono-vacancies and di-vacancies with intensity (I2) 42.8 ± 1.6%. The third lifetime appeared with average values of 1 544 ± 820 ps and intensity of 0.4 ± 0.1%. The MLT of the implanted specimen was between 142 and 146 ps, and the average value was 143 ± 2 ps. This is a much higher MLT value than for the as-received specimens. The MLT change was 8 ± 4 ps, which proves there was defect accumulation due to H+ implantation in the specimen. The mono-vacancies partially changed into di-vacancies during the process of irradiation. Although previously found dislocations did not disappear from the structure, only the positrons were more attracted and trapped in the larger defects, di-vacancies, which could be near the dislocations [45].

The second components of PALS spectra and MLT were progressively changed after the gradual experimental annealing at 200, 300, 400, 450, 475, 500, 525, and 550 ◦C, as is seen in Figure 7. The annealing temperature of 200 ◦C did not show significant change from the perspective of MLT, 142 ± 1 ps, but there were found mostly mono-vacancies, probably together with dislocations again. The average value of the second component for the specimens annealed at 200 ◦C is LT2 = 165 ± 3 ps and I2 = 59.1 ± 1.5%. The existence of the third component affected the MLT value, and thus its change could probably not be so evident. The average values for third component are LT3 = 992 ± 190 ps and I3 = 0.4 ± 0.09%.

The specimens annealed at temperatures from 300 to 500 ◦C showed a small decrease in defect presence compared to the implanted data as well as the specimens annealed at 200 ◦C. The average MLT values for all annealed specimens are similar, being 140 ± 2 ps (Figure 7). Only the specimens annealed at temperatures between 475 and 500 ◦C had MLT values of 139 ± 1 ps, which is a practically negligible difference. However, the temperature of 475 ◦C is commercially used for the recovery of reactor steels in older VVER reactors. Our radiation-induced changes had much lower levels and were only present in the thin layer of the specimens, and thus the change here is not so significant.

The second components for specimens annealed at temperatures between 300 and 500 ◦C are almost the same and showed the presence of mostly mono-vacancies together with dislocations: LT2 = 169 ± 2 ps and I2 = 57.3 ± 0.9% (at 300 ◦C), LT2 = 169 ± 2 ps and I2 = 56.8 ± 1.4% (at 400 ◦C), LT2 = 168 ± 2 and I2 = 57.6 ± 1.3% (at 450 ◦C), LT2 = 167 ± 2 ps and I2 = 57.1 ± 1.8% (at 475 ◦C), LT2 = 166 ± 2 ps and I2 = 56.68 ± 0.9% (at 500 ◦C).

The MLT for specimens annealed at higher temperatures than 500 ◦C indicates an increase in the defect presence at a more significant rate than during the implantation. The second component for specimens annealed at 525 ◦C has average values LT2 = 176 ± 3 ps (mono-vacancies) and I2 = 57.3 ± 0.9%. The second component for specimens annealed at 550 ◦C is LT2 = 177 ± 2 ps (mono-vacancies) and I2 = 56.3 ± 0.6%. These specimens have also third components with average values: LT3 = 1 228 ±150 ps and I3 = 1.19 ± 0.2 (at 525 ◦C), LT3 = 1 865 ± 180 ps and I3 = 1.3 ± 0.2 (at 550 ◦C).

From the positron data, the defect concentrations and defect volume were calculated according to equations published in [101–103] and presented in Figure 8. The average values of defect concentration and defect volume for the individual treatments are seen in Figure 8. The calculation was applied only for the material visible by positrons up to a depth of approximately 150 μm, which is a volume of around 0.012 cm<sup>3</sup> with ≈ 10<sup>21</sup> atoms.

The average defect concentration (CD) decreased about ΔCD = − (2.85 ± 0.89) ppm after the implantation due to the formation of larger defects (from dislocation with monovacancy to mostly di-vacancy). However, the average defect volume (VD) observed by positrons increased by ΔVD = + (2.88 ± 1.05) × 10−<sup>8</sup> cm<sup>3</sup> due to the implantation.

After annealing at 200 ◦C, the size of defects decreased, and the defect concentration increased back by ΔCD = + (3.84 ± 0.83) ppm compared to the implanted specimens and by ΔCD = + (0.99 ± 0.73) ppm for to the as-received specimens. This indicates the presence of more defects than in the two previous stages of the specimens. The average defect volume decreased by ΔVD = − (2.00 ± 0.97) × 10−<sup>8</sup> cm<sup>3</sup> compared to the implanted specimens due to partial recombination of di-vacancies and a shift of the average defect size from di-vacancy to a vacancy. The change of the defect volume compared to the as-received specimens is ΔVD = + (0.88 ± 0.72) × 10−<sup>8</sup> cm3, and thus the annealing did not recover the material.

The decline of defect volume due to annealing was the biggest at 200 ◦C, although the defect volume still gradually decreases up to 500 ◦C, almost only within the error bars. However, the absolute lowest values for defect concentration and defect volume within the annealed specimens seem to occur at temperatures 450 and 475 ◦C. The values ΔCD ≈ − 3.19 ppm and ΔVD ≈ − 2.57 × 10−<sup>8</sup> cm<sup>3</sup> showed the biggest decrease compared to the implanted specimens for both temperatures (the same values). The changes in the values for 450 and 475 ◦C compared to the as-received specimens are ΔCD ≈ + 0.34 ppm and ΔVD ≈ + 0.30 × 10−<sup>8</sup> cm3, which shows that no full recovery of the structure occurred due to the annealing.

At the higher temperatures of 500, 525, and 550 ◦C, the defect concentration and the defect volume visibly increased more than ΔCD ≈ 0.85 ppm and ΔVD ≈ 0.75 × 10−<sup>8</sup> cm<sup>3</sup> compared to the as-received specimens, but the defect volume is still lower than for the implanted specimens due to smaller defects. This could be due to thermal strain and the formation of new vacancies due to structural changes (precipitates, defect mobility, thermo-vacancies).
