Comprehensive tritium migration model for lithium metatitanate ($\mathrm{Li}_2\mathrm{TiO}_3$) breeder ceramics: damage-dependent trapping, gas–surface isotope exchange, and coupled kinetics

Lithium metatitanate ($\mathrm{Li}_2\mathrm{TiO}_3$) is a leading solid tritium-breeder ceramic for fusion blankets, and its fuel-cycle viability depends on predicting how bred tritium migrates out of the material. A 2013 experimental study by Kobayashi, Oya, and Okuno established grain-scale Arrhenius parameters for tritium diffusion, detrapping from irradiation defects, and hydroxyl (O–T) decomposition, and posed the objective of converting these pieces into a comprehensive tritium migration model, listing the specific gaps: tritium behavior under $\mathrm{H}_2$/$\mathrm{H}_2\mathrm{O}$-bearing purge gas and isotope exchange, the damage- and burn-up-dependence of trapping-site populations, the merely qualitative atomistic diffusion mechanism, the unvalidated discrepancy in detrapping pre-exponential factors between neutron-irradiated and ion-implanted samples, and the coupling of surface O–T bond chemistry with bulk diffusion. Thirteen years of follow-up work has substantially narrowed two components — the atomistic mechanism is now grounded in multiple first-principles studies, and the damage-density dependence has been experimentally mapped and modeled at grain scale — but the integrated objective remains open: no $\mathrm{Li}_2\mathrm{TiO}_3$-specific surface adsorption/desorption and gas–surface isotope-exchange kinetics have been measured, trap-population evolution under burn-up and annealing is unquantified, and no validated coupled model spanning grains, defects, pores, and purge-gas chemistry exists. The problem posed here is to construct and validate that comprehensive model against reactor-relevant release data.

Background

In a deuterium–tritium fusion reactor, tritium must be bred inside the machine: the breeding blanket contains lithium-bearing ceramics in which the nuclear reaction ${}^6\mathrm{Li}(n,\alpha)\mathrm{T}$ produces tritium that must then migrate through the ceramic grain, reach the pore space, and be swept out by a purge gas (typically helium, often with added hydrogen) for recovery. Lithium metatitanate ($\mathrm{Li}_2\mathrm{TiO}_3$) is a leading candidate breeder ceramic for the test blanket modules of ITER (the International Thermonuclear Experimental Reactor, a large tokamak experiment under construction) and for DEMO (the planned demonstration fusion power plant) blanket designs, because of its chemical stability, low activation, and favorable tritium release at relatively low temperature; in these designs the ceramic is typically fabricated as millimetre-sized spheres (pebbles) packed into a pebble bed. Predicting the tritium inventory (how much tritium stays trapped in the ceramic) and the release rate is a design-driving quantity, since the tritium breeding ratio (the number of tritium atoms bred per tritium atom burned in the plasma) barely exceeds unity and excess inventory is both a fuel-cycle loss and a safety hazard. The migration itself is studied mainly by thermal desorption spectroscopy (TDS, heating an irradiated sample and recording the released tritium as a function of temperature) and by isothermal release experiments; kinetic parameters are extracted with the Kissinger–Akahira–Sunose (KAS) model-free method, which infers activation energies from how desorption peak temperatures shift with heating rate, without assuming a specific kinetic model. Trapping of hydrogen isotopes is described conceptually by the McNabb–Foster framework, a set of coupled diffusion-plus-trapping rate equations in which mobile solute reversibly binds to trap sites; in lithium ceramics the relevant traps include irradiation defects (oxygen-vacancy-related paramagnetic centers detectable by electron spin resonance, ESR, such as F$^+$ centers — oxygen vacancies occupied by one electron — and O$^-$ centers, oxygen atoms with dangling bonds) and hydroxyl groups (O–T/O–D bonds, seen as stretching bands around 2500–2630 cm$^{-1}$ in infrared spectra). Lithium burn-up (consumption of ${}^6\mathrm{Li}$ by transmutation, which creates lithium vacancies) and displacement damage from recoil particles both accumulate these defects over reactor life. Before 2013, tritium release from neutron-irradiated $\mathrm{Li}_2\mathrm{TiO}_3$ was known to be mostly diffusion-controlled, but the reported diffusion activation energies scattered between about 0.6 and 1.6 eV, and additional desorption peaks pointed to trapping whose kinetics were undetermined; for other breeder oxides (e.g. $\mathrm{Li}_2\mathrm{O}$), Nishikawa and co-workers had shown that surface water and isotope exchange reactions between gas-phase hydrogen species and surface-bound tritium strongly affect release — an effect documented for $\mathrm{Li}_2\mathrm{TiO}_3$ itself in the surface-water study of Hanada et al. (2011) — but this chemistry was not yet incorporated into a kinetic migration model for $\mathrm{Li}_2\mathrm{TiO}_3$. The 2013 study of Kobayashi, Oya, and Okuno separated the elementary processes on thermal-neutron-irradiated $\mathrm{Li}_2\mathrm{TiO}_3$ (uniform tritium generation, long diffusion length) and on 3.0 keV $\mathrm{D}_2^+$-implanted $\mathrm{Li}_2\mathrm{TiO}_3$ (deuterium confined to ~25 nm, dense defects): they obtained the grain diffusivity $D = 6.88\times10^{-7}\exp(-1.07\ \mathrm{eV}/kT)\ \mathrm{m^2\,s^{-1}}$, first-order detrapping from E$^\prime$-centers (a further irradiation-induced paramagnetic defect detected by ESR) with $A = 7.54\times10^3\ \mathrm{s^{-1}}$, $E_a = 0.66$ eV, hydroxyl decomposition with $A = 8.39\times10^5\ \mathrm{s^{-1}}$, $E_a \approx 1.17$ eV (neutron-irradiated) versus $A = 4.05\times10^6\ \mathrm{s^{-1}}$ (ion-implanted), and proposed that hydrogen migrates via lithium vacancies interacting with adjacent oxygen atoms, consistent with molecular-dynamics barrier estimates of 0.5–1.0 eV for jumps between lithium vacancies and 1.2–1.6 eV for rotation about oxygen. By comparing the rate constants of diffusion ($k \propto D/d^2$ for grain diameter $d$) with the detrapping rate constants, they could identify the rate-determining step for each desorption peak. What the study deliberately left open — and what defines this problem — is that the experiments were all performed under inert helium or vacuum, so the effect of the molecular hydrogen and water vapor actually present in a blanket purge gas, entering through surface isotope exchange, was not measured; the trap concentrations were treated as fixed rather than evolving with neutron fluence and burn-up; the diffusion mechanism remained a qualitative proposal; the roughly fivefold difference in detrapping pre-exponential factors between neutron-irradiated and ion-implanted samples was attributed to the shorter diffusion length in implantation but not validated; and the formation and decomposition of surface O–D/O–T bonds and their coupling to bulk diffusion were unquantified. In short, all the ingredients of a predictive model existed as separate Arrhenius parameters, but no comprehensive, reactor-relevant tritium migration model for $\mathrm{Li}_2\mathrm{TiO}_3$ existed. Terminology used below: isotope exchange reactions (denoted EX1 and EX2 in the breeder literature) are gas–surface reactions such as $\mathrm{H}_2(\mathrm{g}) + \mathrm{T_2O(ads)} \leftrightarrow \mathrm{T_2(g)} + \mathrm{H_2O(ads)}$ and $\mathrm{H_2O(g)} + \mathrm{T_2O(ads)} \leftrightarrow \mathrm{T_2O(g)} + \mathrm{H_2O(ads)}$ that swap hydrogen isotopes between the purge gas and surface species; Knudsen diffusion is gas-phase diffusion through pores so narrow that molecule–wall collisions dominate; the nudged elastic band (NEB) method and climbing-image variant are standard density-functional-theory (DFT) techniques for computing atomic migration barriers.

Problem Statement

Construct and validate a comprehensive, predictive model of hydrogen-isotope (tritium) migration in lithium metatitanate ($\mathrm{Li}_2\mathrm{TiO}_3$) breeder ceramics, replacing the current patchwork of separately measured Arrhenius parameters with a single coupled description, such that tritium release rate and inventory can be predicted under reactor-relevant conditions (temperature range of blanket operation, purge-gas composition, neutron fluence, and lithium burn-up). The model must couple, with kinetic parameters measured on $\mathrm{Li}_2\mathrm{TiO}_3$ itself: (1) tritium diffusion in the crystal grains, taking as baseline the measured $D = 6.88\times10^{-7}\exp(-1.07\ \mathrm{eV}/kT)\ \mathrm{m^2\,s^{-1}}$ and the proposed mechanism of migration via lithium vacancies with interaction with adjacent oxygen atoms; (2) trapping/detrapping at irradiation defects and hydroxyl (O–T) groups, with trap populations that evolve with neutron fluence and displacement damage, with lithium burn-up (lithium-vacancy creation), and with thermal annealing of defects, rather than being fixed concentrations; (3) surface adsorption/desorption and gas–surface isotope exchange with molecular hydrogen and water vapor in the purge gas, quantifying the rate constants that current system-level models must import from other ceramics ($\mathrm{LiAlO}_2$, $\mathrm{Li}_2\mathrm{ZrO}_3$); and (4) formation and decomposition of surface O–T (O–D) bonds and their coupling with bulk diffusion. The construction must also settle the stated auxiliary issues of the source: validate (or refute) the interpretation that the roughly fivefold difference in detrapping pre-exponential factors between neutron-irradiated and ion-implanted samples (e.g. $8.39\times10^5\ \mathrm{s^{-1}}$ versus $4.05\times10^6\ \mathrm{s^{-1}}$ for hydroxyl decomposition) arises from the much shorter diffusion length in ion implantation; and reconcile the atomistic migration barriers with the measured effective activation energies. Validation requires quantitative agreement, within the uncertainties reported for those measurements, with the independent tritium release experiments reported in the cited studies — thermal desorption and isothermal release, in both inert (helium/vacuum) and $\mathrm{H}_2$/$\mathrm{H}_2\mathrm{O}$-containing atmospheres, over the damage-level range those experiments span — and the model must demonstrably resolve which process is rate-determining in each regime, extending to the breeder-pebble scale (grain, pore, pebble bed) where tritium inventory predictions for blanket design are made. The target is the integrated model spanning components (1)–(4) together with the validation, which are jointly necessary; the validate-or-refute item above stands in its own right and is resolved by a decisive resolution of the fivefold prefactor interpretation, whether that confirms or refutes the diffusion-length explanation.

References

1. Makoto Kobayashi, Yasuhisa Oya, Kenji Okuno, 'Migration of hydrogen isotopes in lithium metatitanate', Journal of Nuclear Materials 439 (2013) 159-167. DOI: 10.1016/j.jnucmat.2013.04.020, https://doi.org/10.1016/j.jnucmat.2013.04.020
2. Makoto Kobayashi, Kensuke Toda, Yasuhisa Oya, Kenji Okuno, 'Dependency of irradiation damage density on tritium migration behaviors in Li2TiO3', Journal of Nuclear Materials 447 (2014) 1-8. DOI: 10.1016/j.jnucmat.2013.12.001, https://doi.org/10.1016/j.jnucmat.2013.12.001
3. Makoto Kobayashi, Hiromichi Uchimura, Kensuke Toda, Yasuhisa Oya, 'Effects of helium and ambient water vapor on tritium release from Li2TiO3', Journal of Nuclear Materials 455 (2014) 735-738. DOI: 10.1016/j.jnucmat.2014.08.058, https://doi.org/10.1016/j.jnucmat.2014.08.058
4. Makoto Kobayashi, Yasuhisa Oya, Kenzo Munakata, Kenji Okuno, 'Developing a tritium release model for Li2TiO3 with irradiation-induced defects', Journal of Nuclear Materials 458 (2015) 22-28. DOI: 10.1016/j.jnucmat.2014.11.047, https://doi.org/10.1016/j.jnucmat.2014.11.047
5. Makoto Kobayashi, Yasuhisa Oya, 'Development of the Tritium Transport Model for Pebbles of Li2TiO3', Plasma and Fusion Research 13 (2018) 3405048. DOI: 10.1585/pfr.13.3405048, https://doi.org/10.1585/pfr.13.3405048
6. Makoto Kobayashi, Yasuhisa Oya, Kenji Okuno, 'Tritium trapping states induced by lithium-depletion in Li2TiO3', Journal of Nuclear Materials 487 (2017) 84-90. DOI: 10.1016/j.jnucmat.2017.02.008, https://doi.org/10.1016/j.jnucmat.2017.02.008
7. K. Katayama, Y. Someya, K. Tobita, S. Fukada, Y. Hatano, T. Chikada, 'Influence of hydrogen addition to a sweep gas on tritium behavior in a blanket module containing Li2TiO3 pebbles', Fusion Engineering and Design 113 (2016) 221-226. DOI: 10.1016/j.fusengdes.2016.08.031, https://doi.org/10.1016/j.fusengdes.2016.08.031
8. Samuel T. Murphy, 'Tritium Solubility in Li2TiO3 from First-Principles Simulations', The Journal of Physical Chemistry C 118 (2014) 29525-29532. DOI: 10.1021/jp508875y, https://doi.org/10.1021/jp508875y
9. Yanli Shi, Jianqi Qi, Yong Han, Tiecheng Lu, 'Anisotropic Diffusion of a Charged Tritium Interstitial in Li2TiO3 from First-Principles Calculations', Physical Review Applied 10 (2018) 024021. DOI: 10.1103/PhysRevApplied.10.024021, https://doi.org/10.1103/PhysRevApplied.10.024021
10. Kun Li, Wen Yang, Wei-Hua Wang, Yong-Tang Li, 'First Principles Study of Tritium Diffusion in Li2TiO3 Crystal with Lithium Vacancy', Materials 11 (2018) 2383. DOI: 10.3390/ma11122383, https://doi.org/10.3390/ma11122383
11. Kamal N. Goswami, Samuel T. Murphy, 'Influence of Lithium Vacancy Defects on Tritium Diffusion in β-Li2TiO3', The Journal of Physical Chemistry C 124 (2020) 12286-12294. DOI: 10.1021/acs.jpcc.0c02551, https://doi.org/10.1021/acs.jpcc.0c02551
12. Jin-Yang Su, Wen Yang, Jia-Hong Zhu, Wei-Hua Wang, Kun Li, Shu-Ping Liu, Yong-Tang Li, 'Tritium diffusion in a Li2TiO3 crystal terminated with the (001) surface from first-principles calculations', Physical Chemistry Chemical Physics 22 (2020) 27206-27213. DOI: 10.1039/d0cp04209a, https://doi.org/10.1039/d0cp04209a
13. Zhonghua Lu, Yanli Shi, Yuchen Liu, Huanhuan Liu, Xiuling Wang, Cong Zhang, Gaoyuan Wang, Jianqi Qi, Tiecheng Lu, 'Lowered tritium diffusivity by decreased site density along Li2TiO3 (001) twin boundary', Journal of Nuclear Materials 617 (2025) 156127. DOI: 10.1016/j.jnucmat.2025.156127, https://doi.org/10.1016/j.jnucmat.2025.156127
14. Akito Ipponsugi, Kazunari Katayama, Tsuyoshi Hoshino, 'The influence of the long-term heating under H2 atmosphere on the tritium release behavior from the neutron-irradiated Li2TiO3', Fusion Engineering and Design 170 (2021) 112495. DOI: 10.1016/j.fusengdes.2021.112495, https://doi.org/10.1016/j.fusengdes.2021.112495
15. T. Kulsartov, I. Kenzhina, A. Tolenova, Ye. Kenzhin, A. Shaimerdenov, Ye. Nesterov, Sh. Gizatulin, Ye. Chikhray, A. Gluchshenko, 'Modeling of hydrogen isotopes release from lithium ceramics Li2TiO3 during in-situ experiments using vacuum extraction method', Fusion Engineering and Design 170 (2021) 112705. DOI: 10.1016/j.fusengdes.2021.112705, https://doi.org/10.1016/j.fusengdes.2021.112705
16. Timur Kulsartov, Kuanysh Samarkhanov, Inesh Kenzhina, Zhanna Zaurbekova, Vadim Bochkov, Alexandr Yelishenkov, Yevgen Chikhray, 'High-temperature interaction of water vapor with lithium ceramics Li2TiO3', Nuclear Materials and Energy 38 (2024) 101612. DOI: 10.1016/j.nme.2024.101612, https://doi.org/10.1016/j.nme.2024.101612
17. Xuelei Xie, Shurui Shang, Wei Zheng, Jing Wang, Yifan Zhang, Wei Lu, Chang Chen, Shouxi Gu, Qiang Qi, Yingchun Zhang, Yasuhisa Oya, Haishan Zhou, 'Effect of low dose γ-ray pre-irradiation on the tritium desorption behavior of Li2TiO3 breeding pebbles', Nuclear Materials and Energy 48 (2026) 102177. DOI: 10.1016/j.nme.2026.102177, https://doi.org/10.1016/j.nme.2026.102177
18. Yiming Pu, Ziyi Zhang, Yasuhisa Oya, Zhenhua Yao, Qilai Zhou, 'Analysis of hydrogen isotope release from lithium ceramics with closed pores and different distributions', Nuclear Materials and Energy 48 (2026) 102208. DOI: 10.1016/j.nme.2026.102208, https://doi.org/10.1016/j.nme.2026.102208
19. T. Hanada, M. Nishikawa, T. Kanazawa, H. Yamasaki, N. Yamashita, S. Fukada, 'Effect of surface water on tritium release behavior from Li2TiO3', Journal of Nuclear Materials 417 (2011) 735-738. DOI: 10.1016/j.jnucmat.2010.12.130, https://doi.org/10.1016/j.jnucmat.2010.12.130
20. Kun Xu, Chao Qi, Bo Wang, 'Recent Progress in Research of Solid Tritium Breeder Materials Li2TiO3: A Review', Coatings 12 (2022) 1053. DOI: 10.3390/coatings12081053, https://doi.org/10.3390/coatings12081053
21. Timur Zholdybayev, Timur Kulsartov, Zhanna Zaurbekova, Yevgen Chikhray, Asset Shaimerdenov, Magzhan Aitkulov, Saulet Askerbekov, Inesh Kenzhina, Assyl Akhanov, Alexandr Yelishenkov, 'Determination of Tritium Transfer Parameters in Lithium Ceramics Li2TiO3 During Reactor Irradiation Based on a Complex Model', Materials 18 (2025) 4117. DOI: 10.3390/ma18174117, https://doi.org/10.3390/ma18174117
22. Timur Kulsartov, Inesh Kenzhina, Yergazy Kenzhin, Yevgeniy Chikhray, Asset Shaimerdenov, Zhanna Zaurbekova, Alexander Yelishenkov, Kuanysh Samarkhanov, Meiram Begentayev, Saulet Askerbekov, Aktolkyn Tolenova, 'Simulation of tritium release processes from lithium ceramics under reactor irradiation conditions with varying reactor power', Journal of Nuclear Materials 615 (2025) 155979. DOI: 10.1016/j.jnucmat.2025.155979, https://doi.org/10.1016/j.jnucmat.2025.155979