This project has allowed to advance the fundamental understanding of radiation-induced effects in both biological matter and functional materials components of spacecraft, has provided direct cross sections that can be used as input to MC track structure codes for biological damage and has introduced a new way of counting defects in irradiated materials.
For the radiation effects in biological matter, water, small biological molecules and DNA have been considered. For components of spacecraft, the currently used triple junction solar cells and the next generation hybrid organic-inroganic perovskite-based solar cells have been considered.
The results for the radiation effects in biological matter can be summarized as follows:
1. a new method to compute accurately the electronic stopping power in disordered materials at a reduced computational cost for RT-TDDFT has been developed [B. Gu, B. Cunningham, D. Muñoz Santiburcio, F. Da Pieve, E. Artacho and J. Kohanoff, J. Chem. Phys. 153, 034113 (2020)]. The approach is based on selecting an optimal set of short, nanometer-size trajectories beforehand using a purely geometric criterion. We applied this new scheme for protons in liquid water, and the electronic stopping converges very quickly with the number of short trajectories, requiring at most eight 20 Å-long trajectories. The resulting stopping curve is comparable with existing experimental data at high velocity and with SRIM and PSTAR tables at low velocity. The main difference between our RT-TDDFT calculations and existing empirical stopping curves is located in the Bragg peak region. Similar to calculations by other authors in different systems, the position of the RT-TDDFT peak is red shifted, while the maximum value of the stopping lies between SRIM and PSTAR and close to ICRU 49.
2. In moving to biological molecules, we studied at first the validity of certain common assumptions for multi-chemical systems. The studies on proton irradiation of water vapour to study the validity of the Bragg's rule and on the validity of the Core and Bond approach scheme in SRIM have shown that [B. Gu, D. Muñoz-Santiburcio, F. Da Pieve, F. Cleri, E. Artacho, J. Kohanoff, ubmitted to Radiation Physics and Chemistry https://arxiv.org/abs/2111.14572] the Bragg's additivity rule for this system is applicable to the RT-TDDFT electronic stopping values without scaling when Ek>40 keV/amu. The results indicate that the scaling factor depends on velocity, hence suggesting that the constant 6% scaling proposed by SRIM for water may not be the most suitable approximation.
3. The charge scaling for deriving the stopping of different ions from the one for protons has been investigated. The case of proton and alpha-irradiation of the condensed amorphous phase of glycine has been considered [B. Gu et al, in preparation] and the results have shown that the pre-sampling scheme proved successful also for amorphous glycine to achieve a quicker and more accurate sampling of impacting events between ions and the targets atoms, than randomly placed ion trajectories and that the velocity dependence of the scaling factor used to derive the stopping of alpha-particles from the one for protons is qualitatively similar for SRIM/Core and Bond scheme and RT-TDDFT calculations, with some differences (lower or higher) depending on the energy regime; more importantly, it is observed that the instantaneous charge q depends on the chosen partition method, as there is no univocal form to partition the charge in a polyatomic system and each method has its own advantages and drawbacks.
4. The RT-TDDFT study on solvated DNA allowed to highlight the role of the solvating water molecules, which is rather remarkable [D. Muñoz-Santiburcio et al, in preparation]. The excitation distribution, informing on the depopulation of the different energy levels of the system and the population of the different excited levels at a given instant, and a further detailed analysis of the depopulation showed that the water molecules are always more depopulated than those of DNA, that the lone pairs and bonding pairs in the water molecules seem statistically equally affected and that solvation water does change in a considerable way the hole population in the different atoms and bonds with respect to the same dry target and projectile path. Remarkably, at Bragg peak velocities, the hole population of the water subsystem is much higher than that of the DNA itself for almost all the studied projectile paths, which demonstrates a proper assessment of radiation damage in biomolecules must necessarily include some of their solvation shells.
5. A full report for possible studies to connect to MC track-structure codes, and in particular to Geant4-DNA has been delivered.
6. For the inelastic scattering of electrons, we obtained linear response TDDFT-Random Phase Approximation (RPA, no exchange-correlation effects in the response) results in good agreement with experimental data for the Energy Loss Function (ELF) in the optical limit as well as at finite values of the momentum transfer [N. Koval et al, Inelastic scattering of electrons in water from first-principles: cross sections and inelastic mean free path for use in Monte Carlo track-structure simulations of biological damage, in preparation]. Some differences with semi-empirical models used in MC track-structure codes [D. Emfietzoglou et al., Radiation Research, 180(5):499 (2013)] and with recent TDDFT-Adiabatic LDA (ALDA) results [S Taioli et al.,J . Phys. Chem. Lett., 12((1):487 (2021)] highlighted how RPA performed compared to the Drude-like models (essentially based on the Homogeneous Electron Gas) and to ALDA in terms of peak positions and slope. We computed the single-differential cross-section, total inelastic cross-section, inelastic mean free path, and the electronic stopping power from the ELF. The investigated quantities have the potential to be of direct use in open-source MC track-structure codes like Geant4-DNA.
7. We have used RT-TDDFT to calculate the electronic stopping power for protons and electrons in water [N. Koval et al, in preparation] and compared our results with those obtained from semi-empirical ELF models [D. Emfietzoglou et al., Radiat. Res., 164:202 (2005); R. Garcia-Molina et al., Surface and Interface Analysis, 49(1):11 (2017)], ESTAR [M.J. Berger et al., ESTAR, PSTAR, and ASTAR: Computer Programs for Calculating Stopping-Power and Range Tables for Electrons, Protons, and Helium Ions] (only available at high electron velocities) and with linear-response TDDFT at the previous point. The linear-response and semi-empirical results, they largely underestimate the stopping power in a wide range of velocities up to 8 a.u. The results indicate that the linear theory works better for a proton than for an electron and suggests that the semi-empirical approach underestimates the stopping power at low and intermediate projectile velocities considerably for both protons and electrons. This means that the perturbation scheme used in Geant4-DNA could lead to an underestimation of the deposited energy in those energy ranges where also the accuracy of the cross-sections is questionable. First-principles results for the stopping power should be considered in the benchmark studies by the Geant4-DNA community.
9. A roadmap for the ab-initio chemical-physics community has been developed [F. Da Pieve et al., in preparation], highlighting the current status in different aspects of the physical and chemical physics steps of radiation damage in biological matter, the recent developments, and the challenges ahead. Such roadmap focuses mainly on TDDFT and ab-initio MD studies, with contributions also on classical MD studies to tackle large length scales and longer time scales. Connections to the developments in the MC track structure community are also done.
The results for the radiation effects in functional materials, notably spacecraft solar cells, can be summarized as follows:
1. The electronic stopping power calculations for impacting protons on the sub-junctions of the currently used triple-junction solar cells (Ga0.5In0.5P/GaAs/Ge), considered in the keV range (relevant in the transport of trapped radiation through the full stack), have shown that [N. Koval, F. Da Pieve and E. Artacho, R. Soc. Open Sci. 7: 200925 (2020)]: i) depending on the channel and on the materials the electronic stopping is sensitive to the trajectory to a different extent. In GaAs and Ga0.5In0.5P the electronic stopping varies more along different trajectories, affected by varying electron density; ii) the dependence of the stopping on the (centered or off-centered) channelled trajectory becomes higher at higher energies; iii) The effect of the interface between the layers of the lattice-matched multilayer solar cell and the effect of strain is negligible, which is understood, given the very similar chemistry and lattice constants of the materials. The energies used for this study were checked to be relevant for a 3-year mission on an ISS-like orbit, considering the NASA AP-8 model and studying the passage of the radiation through the different layers of the solar cell via the MC particle transport code MULASSIS.
2. The ab-initio MD studies on the effect of electronic excitations (brought by a projectile or a knock- on atom) on the threshold displacement energy, a key quantity in the Non Ionizing Energey Loss (NIEL) model for solar cells degradation) have shown that [D. Muñoz Santiburcio, N. Koval, E. Artacho and F. Da Pieve, Influence of electronic excitation on defect formation in irradiated GaAs, in preparation; F. Da Pieve, M. Koval, D. Muñoz Santiburcio J. Teunissen, E. Artacho, J. Kohanoff and F. Cleri, Fundamentals of Monte Carlo particle transport and synergies with quantum dynamics for applications in ion-irradiated materials in Space and radiobiology, Book chapter, book of the Cost Action TUMIEE 17126, https://oa.upm.es/69109/] the presence of electronic excitations clearly favours the formation of defects on GaAs, but that a small mitigating effect due to a more facile healing in some cases also exists. Both effects are related to the weaker bonding in the structure caused by the electronic excitations. There may occur some local phase transitions from the regular semiconducting, tetrahedrally coordinated GaAs phase to a metallic, octahedrally coordinated phase. The occurrence of such transitions should however be further checked by models able to consider the localized nature of the electronic excitations brought by the projectile or knock-on atom.
3. The RT-TDDFT + MD calculations for electronic stopping for heavy ion impact (the Primary Knock-on Atoms, PKAs) in GaAs via a plane waves implementation of RT-TDDFT were then given as input to MD calculations [J. Teunissen, T. Jarrin, N. Koval, D. Munoz Santiburcio, N. Richard, E. Artacho, J. Kohanoff, F. Cleri and F. Da Pieve, in preparation; T. Jarrin, J. Teunissen, N. Richard, F. Da Pieve and A. Hemeryck, Phys. Rev. B 104, 195203 (2021)]. The results of our work are:i) the Electron-Phonon model (EPH) provides a good description of the electronic stopping in GaAs and it is possible to fit the EPH model for multiple elements materials; ii) the description of the electronic stopping and electron-phonon coupling influences the number of defects, and are thus crucial to correctly model the radiation damage; iii) the temperature dissociation in the Two Temperature Model (TTM) was proven to be much faster, while the EPH model dissipated the heat spike much slower. Nevertheless, the number of defects in the TTM is lower, indicating that more energy goes into nuclear stopping; iv) different methods to count defects, the cut-off method and the Wigner-Seitz method, give rise to a higher or lower number of defects compared to the commonly used Norgett Robinson Torrens (NRT) model. The result based on the Wigner-Seitz method suggests that, independently on the occurrence or not of a thermal spike (whose occurrence in GaAs is put forward in Ref. [F. Gao et al., Journal of Applied Physics 121, 095104 (2017)]), the "ease" of displacing an atom due to the consideration of the electronic excitations induced/brought by the PKA (or another recoil atom) is overcome by the fact that the PKA (or another recoil atom) actually loses energy more efficiently (as part of its energy is lost to the electronic subsystem), making the overall efficiency in displacing atoms attenuated. v) although the "passage" of the non-adiabatic cloud during the very displacement process does affects the counting of defects, this does not exhibit any non-linearity in the number of defects vs the recoil energy. This does not exclude that further non-linear effects may arise because of amorphization/local change of structure or because of any "thermal spike" -like effects.
4. The DFT + compressed sensing-symbolic regression studies on hybrid organic-inorganic perovskites focused on the fundamental bond and composition properties that most affect the halogen-cation interactions in hybrid perovskites in a large chemical space [J. Teunissen and F. Da Pieve, J. Phys. Chem. C 125, 45, 25316 (2021)]. This allowed us: i) to give a rationale of previous results, based on smaller chemical spaces; ii) to unveil the reasons of an improved stability given by specific halogens, the origin of the higher stability offered by certain organic cations compared to others (formamidinium vs methylammonium) and the reason behind the exceptional case of overall stronger bonds for guanidinium; iii) to highlight in a quantitative and first-principles manner the importance of non-covalent interactions, neglected in the majority of studies; iv) to highlight how weak interactions have a significant role to bind the halogen atoms, and thus are expected to be important to prevent halogen migration; v) to design new mathematical, physically meaningful descriptors that identify the factors that dominate the coupling between the halogens and cations.
5. The studies on the NIEL for 3 different types of hybrid organic-inorganic solar cells served to investigate the two classes of experiments done in the field (one focusing on irradiation with protons of several tenths of MeV and another one with ~100 keV). We showed that a uniform damage profile comparable with the damage profiles experienced in space can be obtained with irradiation of 5 MeV protons, and that for the often used 68 MeV irradiation orders of magnitude higher doses are needed to reproduce the NIEL profile as induced in space [J. Teunissen, T. Jarrin, D. Lambert, M. Raine, N. Richard and F. Da Pieve, in preparation]. Input spectra were considered for an ISS-like mission, SEP in interplanetary travel and the surface of Mars, the latter derived by the radiation environment at sites of interest for the landing of ESA ExoMars20221 mission (Oxia Planum andMawrth Vallis) [F. Da Pieve , G. Gronoff , J. Guo, C. J. Mertens , L. Neary , B. Gu , N. E. Koval, J. Kohanoff, A. C. Vandaele , and F. Cleri, Journal of Geophysical Research: Planets 126, e2020JE006488 (2021)].