Synthesis details for the monazite ceramics and single crystals and irradiation details are given in the Methods section and in the supplementary information (SI), sections Density determination and Irradiation details, Supplementary Tables 1 and 2, Supplementary Figs. 1–3. A combination of diffraction and spectroscopic techniques was used to confirm that the composition of the ceramics and single crystals were LaPO4 monazite. A representative diffraction pattern of a pristine, unirradiated monazite ceramic is presented in Supplementary Fig. 4. Additionally, Raman spectra and luminescence spectra of both single crystals and ceramics agree with data reported of corresponding monazite-type samples in the literature27,28.Irradiation details are included in the Methods section. The Stopping and Range of Ions in Matter (SRIM) Monte Carlo simulation code29 with the “full cascade” option was used to estimate the penetration depth of 14 MeV Au ions into LaPO4 (Supplementary Figs. 1–3)30. A depth of approximately two micrometers was calculated and considered when performing all analyses of the irradiated samples.Grazing-incidence X-ray diffraction (GIXRD)To assess the preservation of long-range order, GIXRD experiments were performed to probe the top layer of the samples where, based on the SRIM calculations, the irradiation damage is expected to occur (see Supplementary Figs. 2 and 4). A suitable alignment of the LaPO4 single crystals with respect to the incident grazing beam was not possible due to their irregular shapes. Therefore, GIXRD data were collected only for dense ceramics of LaPO4. An example of an irradiated pellet is shown in Supplementary Fig. 5. As expected, pristine LaPO4 ceramics were found to be fully crystallized (Supplementary Fig. 6). In contrast, the data from irradiated pellets were dominated by diffuse scattering, although reflections from crystalline phases could still be observed in all the samples, even those treated with the highest fluences (Fig. 1). The signal obtained from the crystalline phases present in the irradiated samples is very similar to the signal obtained from the non-irradiated sample (Fig. 1a, b). Specifically, peaks have the same peak shape and broadening. Therefore, Bragg peaks present in the irradiated samples can be attributed to the pristine component not affected by irradiation. Note, some Bragg reflexes are missing, due to the absence of diffracting grains at the probed region of the very textured ceramic pellets. All present diffraction peaks can be assigned to LaPO4 monazite, i.e. the samples do not contain impurity phases. A residual fraction of the parent crystalline phase in the irradiated samples can be estimated by comparing relative intensities of the same peak series. In the case of the sample irradiated at the lowest fluence (5 × 1013 ions/cm2, F1) (Fig. 1c), the corresponding fraction of the residual parent phase is on the order of 10–15%.Fig. 1: GIXRD data for LaPO4 collected at incident angle α = 1°.Diffraction data of the LaPO4 sample irradiated at F2 (a); sample irradiated at F3 (b); and comparison of the integrated data for samples irradiated at all fluences at incident angle α = 1° (c).The LaPO4 sample irradiated at the intermediate fluence (1 × 1014 ions/cm2, F2) features similar fractions of amorphous and crystalline phases as the sample irradiated at the F1 fluence (Fig. 1c). The sample treated with the F3 ion fluence (1 × 1015 ions/cm2) shows a slightly weaker contribution from the amorphous part (Fig. 1c) and, consequently, exhibits a somewhat higher fraction of crystalline phase compared to the samples irradiated with the F1 and F2 fluences. It is known that irradiation with swift heavy ions increases the temperature of the target material via generation of local temperature spikes31,32. This in turn can affect the microstructural and structural properties of the irradiated materials. For example, it can cause an increase in grain size or even induce structural transformations33,34. While the irradiated samples in this study were kept at 77 K with liquid N2 during irradiation, it is not excluded that short (ps time scale) thermal spikes (up to 104 K)31 may still induce local and rather limited annealing. Indeed, an increase in the effective fluence from F1 to F3 is expected to increase the local thermal load inside the irradiated pellets under the same cooling conditions with a possible influence on the local microstructure, i.e. partial stabilization of the parent crystalline phase. During GIXRD experiments, the penetration depth of the X-rays strongly depends on the incident angle, α. Here, α = 1° which corresponds to a penetration depth of ~0.42 μm (Supplementary Fig. 7). This value was calculated using the GIXA package35,36,37. While the theoretical penetration depth of 14 MeV Au5+ ions into the LaPO4 material was estimated to be 2 μm using SRIM29 (see Supplementary Table 2), the parent crystalline phase can already be observed at a low grazing angle α = 1°. Conventionally, the penetration depth is defined as the depth at which the intensity of the incoming radiation is reduced to 1/e (~37%) of the original intensity at the surface. Using the 1/e law, the fraction of the synchrotron beam that can penetrate 2 μm into LaPO4 is found to be close to 1%. Although this number may seem negligible, 1% out of typical synchrotron flux of ~1013 photons/s available at the ROBL38 still may produce a signal that is orders of magnitude higher than that obtained with a laboratory source. Therefore, a small amount of strongly diffracting pristine crystalline LaPO4 may provide a significant contribution to the experimental data dominated by weak diffuse scattering of amorphized LaPO439.A closer look at the scattering data of the irradiated LaPO4 ceramics reveals that the diffraction pattern from the sample irradiated at the lowest fluence of 5 × 1013 ions/cm2 features a few diffuse maxima present on top of the broad diffuse base signal (Fig. 2a). This is in contrast with the diffuse signal from the samples irradiated at the intermediate (Fig. 2b) and highest fluences which exhibit only broad diffuse signals. Careful masking of Bragg reflections on the F1, F2 and F3 ceramics scattering data with a subsequent smoothing of the profile allowed us to obtain qualitative reduced signal from the amorphized part of the irradiated LaPO4 samples (Fig. 3a). Samples irradiated at the intermediate and highest fluences, F2 and F3 respectively, feature near identical diffuse signals; two broad peaks around 20 and 30° 2θ (Fig. 3a) with a low-angle shoulder on the 20° peak. As mentioned above, the diffraction pattern of the sample irradiated with the lowest fluence displays features in addition to the base diffuse scattering signal observed for the middle F2 and highest F3 fluences (Fig. 3a). The presence of these additional well-defined diffuse features indicates that the amorphous LaPO4 phase retains more structured short-range order correlations after irradiation at the lowest F1 fluence compared to irradiation at the higher F2 and F3 fluences. Consequently, the higher F2 and F3 fluences induce a higher degree of amorphization of the parent crystalline LaPO4 phase.Fig. 2: Comparison of diffuse scattering of irradiated LaPO4 ceramics.a Scattering from the sample irradiated at F1 featuring structured diffuse scattering on top of broad diffuse signal. b Scattering from the sample irradiated at F2 showing only broad diffuse signal.Fig. 3: Extracted amorphous scattering signal.Normalized qualitative amorphous scattering signal from the irradiated LaPO4 ceramic samples, low-r part of pair distribution function (a), G(r), obtained from the ceramic irradiated at F3 scattering data (b).Information on short-range order correlations can be obtained from diffuse scattering experiments by calculating the pair-distribution function (PDF). In this study PDFs, G(r), were calculated with the PDFGetX3 package40. The analysis of the LaPO4 diffuse scattering signal is limited due to the low accessible Q range up to 4.44 Å–1. Indeed, experiments in GI mode require utilization of low photon energies in order to limit the penetration depth. In contrast, a high photon energy is required to collect high-Q data suitable for a PDF analysis. Nevertheless, three strong peaks can be seen on the experimental G(r) function (Fig. 3b, F3 irradiated sample is shown as an example) and they correspond to scattering pairs with c.a. 2.5, 4.3, and 6.7 Å distances. The origin of these correlations can be determined by examining the three-dimensional crystal structure of the parent crystalline LaPO4 phase.At ambient conditions the LaPO4 phase adopts the monazite-type structure. In standard setting it corresponds to the P21/n space group and cell parameters of a ~ 6.8 Å, b ~ 7.1 Å, c ~ 6.5 Å, β ~ 103.3°. It consists of edge-sharing LaO9 and PO4 polyhedra that form chains along the crystallographic a axis (Supplementary Fig. 7). It can be seen that the peak around 2.5 Å on the G(r) plot corresponds to the average La-O distance in the LaO9 polyhedra. The 4.3 Å peak corresponds to the closest La-La distances along the c direction. The third peak at 6.7 Å matches the a unit cell parameter of LaPO4 and, therefore, originates from La-La distances along the [1 0 0] crystallographic direction mediated by the PO4 tetrahedra (Supplementary Fig. 8). Consequently, the presence of peaks on the G(r) function at 2.5 and 6.7 Å indicates the preservation of the LaO9 polyhedra and stronger structural rigidity along the a axis, which is the direction of edge sharing LaO9 and PO4 units. In addition, stacking of the LaO9-PO4 chains is also correlated by a 4.3 Å peak on the experimental PDF analysis. Since the sample irradiated at the F1 fluence possesses additional features on top of the base diffuse signal observed for the higher F2 or F3 irradiated samples, the lower fluence induces smaller structural damage in LaPO4, as expected. It may, for instance, indicate better preservation of the LaO9 structural units which feature nine independent La-O distances in the parent crystalline LaPO4. However, the corresponding fine details cannot be obtained from the limited scattering data of the F1 irradiated sample.Vertical scanning interferometry (VSI) and Scanning electron microscopy (SEM)VSI images were taken of select LaPO4 monazite ceramics and single crystals (Fig. 4). VSI is a powerful technique for probing a surface topography at the nanoscale to observe features such as surface roughness and grain height41,42. Fig. 4a shows the surface of the monazite ceramic irradiated at the lowest fluence. The primary feature is the swelling that is concentrated at the grain boundaries of the ceramic. In contrast, at the highest fluence (Fig. 4b), the swelling of entire grains is observed, instead of being concentrated at the boundaries of the grain. However, the damage across the irradiated region is not uniform, as smooth areas surround swollen grains. For further comparison, VSI images of the boundary between pristine and irradiated regions of ceramics are given in Supplementary Figs. 9 and 10. For individual monazite single crystals where grain boundaries are not present, no topographical changes are observed between the pristine single crystal (Fig. 4c) and an irradiated one (Fig. 4d).Fig. 4: VSI images of LaPO4 monazites.VSI images of LaPO4 monazite ceramics exposed to F1 (a) and F3 (b) and of single crystals without irradiation (c) and exposed to F3 (d).SEM images were taken of the irradiated ceramics to further probe changes in surface topography due to ion-irradiation. Due to the unevenness and inhomogeneity of the surfaces of the single crystals, SEM images that clearly resolved the surface features could not be obtained. Representative images of the ceramics irradiated at the highest and lowest fluences are shown in Fig. 5. Energy dispersive spectroscopy (EDS) (Supplementary Figs. 11 and 12) and additional SEM images (Supplementary Figs. 13–50) have been compiled in the SI. Results from EDS analyses are summarized in Supplementary Tables 3–7. The surface of the ceramic irradiated at the highest fluence (F3) exhibits significant damage as observed in Fig. 5a–c. Grain breakage from the surface of the irradiated side of the ceramic signifies severe topographical damage. At the border between the pristine and irradiated parts of the ceramic, differences in grain height indicate that the irradiated side is swollen. Enlarged grains with visible grain boundaries and even grain breakout in the irradiated region of the ceramic are also observed (Supplementary Figs. 39 and 40). The circles on the irradiated surface may be nucleation sites from the ion irradiation (Fig. 5b). It is important to note that the signs of irradiation damage are not uniform across the sample surface, as was observed in TEM studies of AmPO418. In another irradiated region, grains appear smooth as if their surface has melted, with less well-defined grain boundaries and rounded edges (Fig. 5c).Fig. 5: SEM images of irradiated LaPO4 ceramics.SEM images of irradiated LaPO4 ceramics, the surface of ceramic irradiated at F3 (a), the irradiated-pristine boundary of the ceramic irradiated at F3 (b), grain irradiated at F3 (c), surface of ceramic irradiated at F1 (d), the irradiated-pristine boundary the ceramic irradiated at F1 (e), and a grain irradiated at F1 (f).A study by Meldrum et al. reported that irradiation appeared to enhance diffusion and crystallization processes9,43. Additionally, an activation energy of 0.064 eV of recrystallization was calculated, which corresponds to a temperature of about 470 °C43. It is also worth noting that thermal spikes, caused by the transfer of energy of the implantation ions to the electrons in the target material, can cause local heating in the sample44. This could also account for possible recrystallization processes in the irradiated monazite samples, as the critical temperature of monazite is between 350 and 485 K9. Also, a study of electron-irradiation of monazite by ref. 15 reported recrystallization, even under mild conditions. Studies suggest this ‘alpha annealing’ is dependent on the ratio of energies of electronic to nuclear interactions, however more studies are needed to confirm these findings15,20. This corroborates signs of recrystallization observed in the LaPO4 ceramic irradiated at the highest fluence. Notably, this phenomenon was not observed in the other two monazite ceramics irradiated at lower fluences.The ceramic irradiated at the intermediate fluence (F2) exhibited signs of irradiation damage less severe than the ceramic irradiated at the highest fluence. SEM images showed signs of swelling and disintegration of the surface (Supplementary Figs. 22–35). Again, the irradiation damage is not the same across the entire surface. Other irradiated regions of the ceramic are rough compared to the pristine regions, and spalling of the surface is evident. Finally, the ceramic irradiated at the lowest fluence (F1) shows the least damage due to irradiation of all the ceramics. The irradiated regions are rough and exhibit a thin, highly porous surface layer (Fig. 5e). As with the other two ceramics, the irradiation damage does not appear the same throughout the irradiated surface. Differences in grain height and breakouts of grains from the surface are suggested by both VSI and SEM data and suggest radiation-induced swelling (Fig. 5f), as has been reported previously13,15. Fig. 5f also shows visible, rather large pores in one of the grains, which are clearly larger in size than pores found on the pristine side of the ceramic pellets (visible in e.g. Fig. 5d (top) and 5e (top)).SEM images indicate that the structural response to irradiation damage is influenced by topology and that the presence of grain boundaries likely plays a role in mediating the effects of the ion irradiation33,45. This could explain the difference in spectroscopic results between the irradiated monazite single crystals and ceramics discussed in the following section.Raman spectroscopyRaman spectroscopic measurements of irradiated samples were used to probe changes in local coordination environments of the ceramics and single crystals. It is likely that the laser penetrated into the samples beyond the ~2 micron irradiated layer, giving non-negligible contributions from the pristine, unirradiated layers below26.The Raman spectra of both ceramic samples and single crystals show the appearance of a shoulder on the ν1 peak (symmetric stretch vibration) at ~960 cm–1 upon irradiation of the samples (Fig. 6). This is observed in both the ceramics and single crystals, although it is more pronounced in the Raman spectra of the irradiated ceramics. These spectra demonstrate a disruption in the local coordination environment of these samples, however, the Raman bands characteristic of the monazite structure are still readily observed14. Interestingly, when Raman data were collected of pores on the surface of the ceramic or in regions close to grain boundaries, the spectra were featureless and indicated amorphization (see Supplementary Fig. 51). This suggests that the damage is concentrated at the grain boundaries in the ceramics. In addition, swollen grains show a larger contribution of the shoulder at 960 cm–1 than smooth surfaces, which speaks for a heterogeneous damage or damage distribution in the ceramics. Similar, clear differences in the collected Raman spectra are not seen for the single crystals, implying a more homogenous (albeit small) overall damage in these specimens (Supplementary Fig. 51).Fig. 6: Raman spectra of pristine and irradiated LaPO4 monazites.Raman spectra of pristine and irradiated LaPO4 monazite ceramics (a) and single crystals (b).Additionally, in the ceramic samples, the ν2 band at ~465 cm–1 increases in intensity with decreasing fluence for the ceramic samples, whereas no clear trend is observed for the single crystals (Fig. 7). This Raman band is attributed to a bending mode of the phosphate tetrahedra27,28. For the monazite ceramics, the ratio of intensities of the ν2 to ν1 modes, averaged over multiple Raman spectra from different spots on the sample surfaces, were found to increase with decreasing fluence, as shown in Fig. 7a. This same trend was not observed in the ν2 to ν1 ratios of the monazite single crystals (Fig. 7b), however it is worth noting that numerous factors affect the peak intensities in Raman spectra in single crystal studies, especially sample orientation46. A study from ref. 47 posits that the intensity of this peak associated with the phosphate bending mode is dependent on the Ln:P ratio and that intensity decreases with decreasing Ln:P ratios47. Finally, the small Raman bands at ~1065 cm–1 and 1073 cm–1, visible in the pristine sample and the LaPO4 ceramic subjected to the highest fluence (F3), are absent in both samples irradiated at the lower fluences. These bands have been assigned to internal stretching vibrations of the PO4 tetrahedra14. Their absence at lower fluences is another indication for the partial recrystallization of the monazite at the highest fluence, which corroborates the results obtained in our diffraction investigations and the SEM micrographs.Fig. 7: Normalized Raman spectra.Raman spectra normalized to the ν1 band intensity (lower wavenumber region) of the irradiated LaPO4 monazite ceramics (a) and single crystals (b).Many discrepancies exist in the literature for Raman measurements of irradiated monazites. For example, in studies of Ce-monazite lamellae irradiated using three different Au ion energies, Raman measurements demonstrated substantial structural damage to short-range order indicated by a loss of intensity and shifting of Raman bands to lower wavenumbers17,26. However, other reports of Au-ion irradiated monazite ceramics showed no significant broadening or shifting of the Raman bands, especially those associated with the PO4 tetrahedra coordination environments13. Raman measurements of AmPO4 polycrystalline samples showed an increase in the full-width at half maximum (FWHM) of certain peaks, such as the band associated with the symmetric stretch of the phosphate tetrahedra, however peak positions remained unchanged18. It is worth noting that Raman spectra of irradiated monazites appear dependent on a number of factors, such as sample form, type of irradiation (self-irradiation versus external ion-irradiation), and the setup of the Raman instrument as confocal or non-confocal.Luminescence spectroscopyAs mentioned previously, all ceramics and single crystals were doped with ~500 ppm Eu(III). Eu(III) is a luminescent lanthanide that is known to be incorporated into the LaPO4 monazite structure and gives distinctive luminescence spectra that provide information about its local coordination environment48. The low concentration of the Eu-dopant was used to avoid self-quenching of the luminescence signal due to two luminescent ions being in close proximity to one another as well as to study how the LaPO4 endmember and not a LaxEu1-xPO4 solid solution reacts to heavy ion-irradiation49.Luminescence data include excitation and emission spectra, as well as luminescence lifetime decay plots, which can be used to determine the number of non-equivalent sites in a given matrix, their respective site symmetry, and the lifetimes of each species48,49,50. It is important to note that luminescent probes, such as Eu(III), still give a spectrum even when in amorphous environments, making it well-suited for a sample in which irradiation-induced amorphization is likely to occur.The integrated excitation spectra, i.e. the luminescence emission intensity as a function of excitation wavelength, of irradiated monazite ceramics and single crystals normalized to maximum intensity are presented in Fig. 8. The excitation peak maximum for Eu(III)- doped LaPO4 has been shown to occur at 578.40 nm49,50. The excitation peak maxima for each of the ceramic and single crystal samples studied here agrees well with this reported value for pristine monazite. For the irradiated samples, the trends in peak maximum relative to each other are different for the two different sample types. For the ceramics, no systematic trend in peak maximum as a function of fluence is observed (Fig. 8a). The peak maxima of the samples irradiated at the F1 and F2 fluences are both slightly blue-shifted compared to that of the pristine sample. A blue-shift has been reported for lanthanide monazite endmembers and solid solutions of La-Gd monazites, where the shift to higher energies in both cases was attributed to the lengthening of the Ln-O bond and the weaker exertion of the ligand field on the Eu(III) dopant cations49,50. Thereby, analogous to these studies, we assign the minor blue shift in these ceramic samples to a slight increase in the Eu(III)/La-O interatomic distance49,50. The sample irradiated at the highest fluence has a peak maximum almost identical to that of the pristine, suggesting a similar crystal field strength in both samples types, and consequently a very similar Eu(La)-O interatomic distance, compared to the samples irradiated at the lower fluences. This is similar to the findings of the SEM images, in which the surface of the sample irradiated at the highest fluence appeared recrystallized and less damaged. Unlike the ceramics, the integrated excitation spectra show a different trend for the monazite single crystals (Fig. 8b). There is a systematic blue-shift with increasing fluence, again corresponding to an increased interatomic distance between the La atoms and the coordinating oxygen atoms. Spectra of a Ne calibration lamp were taken before the measurement of each sample and confirmed that these shifts, while small, are significant (Supplementary Fig. 52).Fig. 8: Eu(III) excitation spectra of LaPO4 monazites.Eu(III) excitation spectra (7F0 ← 5D0 transition) of LaPO4 monazites doped with 500 ppm Eu(III): polycrystalline ceramics (a) and single crystals (b).The difference in excitation spectra between irradiated monazite ceramics and single crystals could be due to the presence or absence of grain boundaries. Studies have demonstrated that at elevated irradiation temperatures, more defect recombination will occur, but at lower temperatures of irradiation, migration of defects to sinks, such as grain boundaries is more prominent44,51. Additionally, studies have suggested that grain boundaries can act as sinks for point-defects, allowing for defect recombination21,22,23,24.Emission spectra of all samples are presented in Fig. 9. In an unirradiated monazite matrix with monoclinic C1 lattice symmetry, the Eu(III) dopant cations occupy low symmetry sites that give threefold splitting of the 7F1 band and fivefold splitting of the 7F2 band48. In this low symmetry system, the 5D0 → 7F1 transition has predominantly magnetic dipole character and therefore does not change significantly with disturbances in ligand environment48. Conversely, the 5D0 → 7F2 transition has predominantly electric dipole character and is hypersensitive to changes in ligand environment. The typical splitting pattern observed in LaPO4 monazite doped with 500 ppm Eu(III) is observed in all samples here regardless of sample type and radiation dose. However, band intensity and sharpness change with fluence.Fig. 9: Eu(III) emission spectra of irradiated LaPO4 monazites.Emission spectra of irradiated LaPO4 monazite ceramics (a) and single crystals (b) doped with 500 ppm Eu(III).Changes in the 7F1 and 7F2 bands in the ceramics and single crystals with irradiation are subtle. To probe for asymmetry in the coordination environment of the Eu(III) dopant cations, the 7F2/7F1 ratios were calculated both for the emission spectra collected after excitation at the corresponding excitation peak maxima of each sample (these emission spectra are shown in Fig. 9) as well as at the FWHM (see Supplementary Fig. 53 for more details). As broadening of excitation peaks should increase with increasing disorder in the sample, the 7F2/7F1 ratios should be larger at the FWHM than at the excitation peak maximum. These results are given in Fig. 10. For the ceramics, no clear trend in the F2/F1 ratio of the irradiated samples is observed when compared to that of the pristine ceramic. All 7F2/7F1 ratios lie between 0.73 and 0.75, and no conclusions of the relative asymmetry can be drawn from these ratios. However, in the case of the single crystals, the F2/F1 ratios increase systematically from 0.84 to 0.94 with increasing fluence, corresponding to an increase in asymmetry of the Eu dopant coordination environment. This difference in the F2/F1 ratios appears dependent on sample form, again suggesting that grain boundaries in the ceramics affect radiation response. However, the 7F2/7F1 ratio is lower for the emission spectra corresponding to the peak maximum of integrated excitation energy, compared to that of the emission spectra corresponding to the FWHM for all samples, regardless of sample form. This corresponds to more structural disorder and a greater contribution of the irradiated layer in the spectroscopic signal.Fig. 10: Trends in the 7F2/7F1 ratios.Trends in the 7F2/7F1 ratios of emission spectra collected after excitation at the peak maximum and at the full width at half maximum of the integrated excitation peaks for the ceramics (a) and the single crystals (b).Lifetimes were collected for the LaPO4 monazite ceramics and single crystal samples and correspond to values reported previously in the literature (Supplementary Fig. 54, Supplementary Table 7)49. However, the lifetimes of the samples do not provide additional information about their microstructure, specifically with regard to radiation damage. Additional details are offered in the SI.
