SEARCH · Engineering Papers

Results for “Grain Boundary”

Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.

SourceResearch connectionFrom yearThrough year

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 577 records · Page 32

All sources NASA NTRS DOE OSTI

DOE OSTI2021

Theoretical Analysis of Experimental Data of Sodium Diffusion in Oxidized Molybdenum Thin Films

In this work, the diffusion process of sodium (Na) in molybdenum (Mo) thin films while it was deposited on soda lime glass (SLG) was studied. A small amount of oxygen was present in the chamber while the direct-current (DC) magnetron sputtering was used for the deposition. The substrate temperatures were varied to observe its effect. Such molybdenum films, with or without oxidations, are often used in thin film solar cells, either as back contact or as hole transport layers. Secondary ion mass spectrometry (SIMS) was used to quantify the concentration of the species. A grain diffusion mechanistic model incorporating the effect of grain and grain boundary geometrical shape and size was developed. The model was used to provide an in-depth theoretical analysis of the sodium diffusion in molybdenum thin films that lead to the measured SIMS data. It was observed that not only diffusion coefficients should be considered when analyzing diffusion processes in thin films but also the ratio of grain boundary size to grain size. Both depend on substrate temperature and directly affect the amount of diffused species in the film. The data were analyzed under the light of the film growth speed versus diffusion front speed, the effect of oxygen content, and the effect of substrate temperature on the overall diffusion process. The temperature inversely affects the ratio of grain boundary size and grain size and directly affects the diffusion coefficient, which leads to a preferable temperature at which the highest amount of alkali can be found in the film.

36 MATERIALS SCIENCE↗

DOE OSTI2021

Hall–Petch strengthening limit through partially active segregation in nanocrystalline Ag-Cu alloys

The breakdown from grain-size strengthening to softening mechanisms is generally well understood for high-purity nanocrystalline materials when the mean grain size decreases to the nanometer range. In nanocrystalline alloys, however, the stabilization of nanosized grains by grain-boundary solute segregation complicates the above mechanisms. Moreover, current segregation models have little predictive power for determining the optimal solute content that maximizes Hall-Petch strengthening effects. In this article, using large-scale hybrid Monte-Carlo/molecular dynamic simulations, we present a systematic study of the Hall–Petch breakdown in Cu-segregated Ag alloys with grain sizes ranging from 8 nm to 59 nm, where three concentration-dependent regimes of plasticity are described: (1) Classical segregation strengthening behavior at low solute contents, (2) shear band-induced softening at high solute contents, and (3) a previously unknown, but extended plateau of maximum strengths for intermediate solute contents from 4 to 15 at.%, which we term as nanocrystalline Sterling silver. We find that flow strengths in nanocrystalline Sterling alloys naturally exhibit a zero-slope limit at the smallest grain sizes that is well below the ideal Hall–Petch strengthening trend. This phenomenon results from partially active grain-boundary segregation that acts to influence interfacial plasticity in some, but not all, grain boundary regions. Our findings amplify the atomic nature of solute segregation and interaction at grain boundaries and its complex roles on grain boundary-mediated plasticity mechanisms in nanocrystalline alloys.

36 MATERIALS SCIENCE↗

NASA NTRS1990

Effect of high temperature hydrogen exposure on the strength and microstructure of mullite

The corrosion of near stoichiometric mullite (3Al2O3-2SiO2) by pure dry hydrogen gas was studied at 1050 and 1250 C for times up to 500 hr. The hydrogen preferentially attacked the grain boundaries of the mullite where an aluminosilicate glass was present. Corrosion of the mullite grains was observed after 125 hr at 1250 C. The hydrogen reaction removed SiO2 from the glassy grain boundaries and the mullite grains resulting in a porous alumina rich surface. At 1250 C the strength increased after short exposure times (at least up to 125 hr) and decreased by 53 percent after 500 hr. At 1050 C, all exposure times (25 to 500 hr) decreased the strength. At 500 hr room temperature strength of mullite exposed to 1050 C was reduced by 22 percent. The strength reduction after short exposure times at 1050 C is attributed to crystallization of the grain boundary glass phase.

Herbell, Thomas P.↗

DOE OSTI2020

Comparative study of helium bubbles in a Ti-Ta alloy and a Ti/Ta nanocomposite

Here, we investigated the size, density, and distribution of helium (He) bubbles in a Ti-Ta single-phase alloy and a Ti/Ta dual-phase nanocomposite using transmission electron microscopy. The Ti/Ta nanocomposite was fabricated via phase separation during high-temperature annealing of the single-phase Ti-Ta alloy. He ion implantation in the Ti-Ta single-phase alloy leads to the formation of nano-scale He bubbles (~1.7 nm in size, ~2.6×10 5 /μm 3 in volumetric density). He bubbles were found to segregate to grain boundaries. Under identical implantation conditions, numerous He bubbles also formed in the Ti/Ta nanocomposite. Closer inspection revealed that He bubbles in the Ti-rich phase are smaller and of higher volumetric density (~1.2 nm, ~8.4×10 5 /μm 3 ) than those in the Tarich phase (~3 nm, ~1.8×10 5 /μm 3 ) and the single-phase alloy. He bubbles were also observed to decorate phase boundaries, but they were generally smaller than those on grain boundaries. These results suggest that Ti/Ta phase boundaries are less conducive to the growth of large He bubbles than grain boundaries in the Ti-Ta single phase alloy.

36 MATERIALS SCIENCE↗

NASA NTRS2022

A Record of Low-Temperature Asteroidal Processes of Amoeboid Olivine Aggregates from the Kainsaz Co3.2 Chondrite

Amoeboid olivine aggregates (AOAs) from the Kainsaz CO3.2 chondrite were analyzed using transmission electron microscopy in order to gain a more complete understanding of thermal metamorphism on the parent body and the role of fluids during metamorphic heating. The Kainsaz AOAs are dominated by strongly zoned, fine-grained, olivine grains (Fa2-31) with heterogeneous Fe enrichments along the grain boundaries, which are interpreted as the result of 26Fe2+-Mg2+ interdiffusion with the matrix during thermal metamorphism. However, our diffusion calculations show that such AOA olivine zoning and compositions cannot be produced by a simple diffusional exchange during metamorphic heating, unlike chondrule olivine zoning and compositions. In addition, fine-grained ferroan olivine overgrowths occur heterogeneously in crystallographic continuity with olivines on the AOA exteriors. The overgrowths (Fa33-36) are compositionally distinct from the underlying AOA olivines and are not fully equilibrated with the matrix olivines (Fa~20-55).The ferroan olivine overgrowths likely formed by precipitation from fluids in an epitaxial relationship with forsteritic olivine on the edges of AOAs. Texturally and compositionally diverse chromite grains are also observed along olivine grain boundaries, in olivine grains, and in pore spaces between olivine grains. They share a similar crystallographic orientation relationships with adjacent olivine, suggestive of their formation by exsolution and/or epitaxial growth. Collectively, these observations provide evidence for the mobilization of Fe, Mg, Si, Cr, and Al in the presence of fluids along olivine grain boundaries and into olivine grains during thermal metamorphism. We conclude that in Kainsaz AOAs, the strong zonation development in individual olivine grains and the formation of ferroan olivine overgrowths and chromite grains were a fluid-driven process that occurred at relatively low temperatures (<500°C), during the cooling history of the CO3 chondrite parent body, following the peak of thermal metamorphism.

Amoeboid olivine aggregates↗

DOE OSTI2024

Identifying multiple synergistic factors on the susceptibility to stress relaxation cracking in variously heat-treated weldments

The 347H austenitic stainless steel has been widely used for pressure vessels and pipeline (PVP) applications due to its excellent creep and corrosion resistance, which fit ideally to the harsh conditions in petrochemical industries, fossil fuel or nuclear power plants, and modern energy storages. However, a failure mode has been commonly observed with cracks emerging at the heat affected zone (HAZ) of weldments during post-weld heat treatment (PWHT) or under intermediate to high temperature service conditions. This phenomenon is termed as Stress Relaxation Cracking (SRC) since the purpose of PWHT is to relieve the welding-induced residual stress fields, or as Stress Age Cracking (SAC) if failure happens during service. A leading literature explanation of this failure suggests that the residual stress relaxation and the precipitation dissolution and/or re-precipitation occur in the same temperature range, which can lead to locally high strains and thus to crack at the grain boundaries. Since in situ spatial measurements of residual stress fields, microstructural evolution, and failure processes are nearly infeasible, this work recourses to a micromechanical finite element framework that models the high temperature failure as the nucleation and growth of grain boundary cavities, whereas various parameters such as thermomechanical loading history and its evolution, the competition of grain-interior dislocation creep and grain-boundary diffusion in failure lifetime, and microstructural heterogeneities (such as the precipitate free zone near grain boundaries) can be quantitatively incorporated. It can be concluded from these microstructure-explicit simulations that an accurate knowledge of residual stress evolution and a carefully calibrated set of material constitutive parameters are the essential prerequisites for lifetime predictions. The understanding of individual governing factors also leads to a mechanistic interpretation of the observed SRC susceptibility C-curves. In conclusion, these results suggest that the criticality of residual stress evolution, but not the precipitation-induced local strains, be the leading factor for SRC.

347H stainless steel weldments↗

DOE OSTI2025

Production of Creep-Resistant FeCrAl-ODS

Four ODS FeCrAl alloys with base alloy composition Fe-16Cr-4Al-2Mo-0.2Y (wt.%) were mechanically alloyed with minor additions of CeO 2 and Y 2 O 3 additions and were successfully extruded into rectangular-shaped bars. Two oxide compositions were explored, which were designated NC with oxide addition of 0.5CeO 2 plus 0.2Yin solution of the base alloy powder and PC with additions of 0.5CeO 2 and 0.7Y 2 O 3 to elevate the O level. The microstructure of both alloys revealed by SEM and EBSD to consist of very small grains with sizes of ~450-520 nm and weak texture components. Sub-size tensile specimens fabricated from both alloys were tensile tested from room temperature to 800ºC. The results for both alloys from room temperature to 400ºC showed very high strengths, but low ductilities, especially uniform strain. From 500ºC to 800ºC, the strengths decreased rapidly, and the ductilities increased significantly. The reason for the rapid decreases in tensile strengths above 500ºC for both alloys is not clear, but will require advanced characterization techniques including atom probe tomography, to determine the size and number density of the nano-size oxide particles, and TEM/EFTEM, for investigating the dispersion of nano-size oxide particles on grain boundaries to assess the Zener grain boundary pinning effect, which will hinder grain coarsening at high temperatures that may enhance the Hall Petch strengthening mechanism.

36 MATERIALS SCIENCE↗

DOE OSTI2026

Ambient Aging of Epitaxial and Polycrystalline Rhenium Thin Films

Extending the lifetime of superconducting qubits requires materials that are resistant to surface oxidation. Rhenium has emerged as one such promising candidate that forms a sub-nanometer oxide. Grain boundaries are known to promote material degradation, and thus, we compare epitaxial (single crystal, no grain boundaries) and polycrystalline (many grain boundaries) rhenium thin films after three years of ambient environmental exposure. Epitaxial rhenium remained free of undesired particles, while polycrystalline rhenium developed surface oxide crystallites. Surface characterization using scanning electron microscopy, energy dispersive x-ray spectroscopy, and x-ray photoelectron spectroscopy reveals localized oxygen and nitrogen enrichment, together with a Re⁷⁺ oxidation state in the polycrystalline surface particles, consistent with the proposed formation of ammonium perrhenate. Epitaxial rhenium demonstrated greater resistance to aging and oxidation, making it a more promising candidate for superconducting qubits. Ongoing cleaning and XPS depth profiling will further determine the identity and origin of the surface crystallites.

Wiacek, Wiktoria [Fermilab]↗

DOE OSTI2026

Ambient Aging of Epitaxial and Polycrystalline Rhenium Thin Films

Extending the lifetime of superconducting qubits requires materials that are resistant to surface oxidation. Rhenium has emerged as one such promising candidate that forms a sub-nanometer oxide. Grain boundaries are known to promote material degradation, and thus we compared epitaxial (single crystal, no grain boundaries) and polycrystalline (many grain boundaries) rhenium thin films after three years of ambient environmental exposure. Epitaxial rhenium remained free of undesired particles, while polycrystalline rhenium developed surface oxide crystallites. Surface characterization using scanning electron microscopy, energy dispersive x-ray spectroscopy, and x-ray photoelectron spectroscopy reveals localized oxygen and nitrogen enrichment together with a Re⁷⁺ oxidation state in the polycrystalline surface particles, consistent with the proposed formation of ammonium perrhenate. Epitaxial rhenium demonstrated greater resistance to aging and oxidation, making it a more promising candidate for superconducting qubits. Ongoing cleaning and XPS depth profiling will further determine the identity and origin of the surface crystallites.

Wiacek, Wiktoria [William Rainey Harper Coll.]↗

DOE OSTI2025

Understanding the effect of minor alloying elements on helium bubble formation in ferritic-martensitic steels

Ferritic-martensitic steels are promising structural materials for advanced nuclear reactors. To minimize long-term radioactivity, reduced-activation ferritic-martensitic steels have been developed by substituting high-activation elements like Ni and Mo with low-activation elements such as W. However, the impact of these alloying modifications on helium bubble formation, which plays a key role in material swelling, remains unclear. Here, in this study, we compared helium bubble formation in ferritic-martensitic steel T91 and reduced-activation ferritic-martensitic steel F82H. Both materials were irradiated with sequential 100 keV, 150 keV, and 200 keV helium ions to a dose of 0.5 dpa and a helium concentration of 9,000 appm at 500°C. The helium bubbles in F82H exhibited a larger average size and a lower density than those in T91, suggesting differences in minor alloying elements may influence the bubble growth. Here, to investigate the effects of these alloying elements, we characterized radiation-induced segregation near bubbles and grain boundaries. Prominent Ni-Mn-Si enriched clusters were found near bubbles in T91, while only Mn-Si enriched clusters were found near bubbles in F82H. In addition, the obvious Cr enrichment near grain boundaries was absent around bubbles in both steels. The different segregation trends among elements revealed the variations in element diffusion mechanisms and the different sink biases between bubbles and grain boundaries. Cr enrichment near grain boundaries is mostly driven by interstitial-mediated diffusion. However, since bubble growth relies on net vacancy flux, vacancy-mediated diffusion plays a dominant role in controlling element segregation near bubbles. Therefore, Cr enrichment was not found near bubbles. Because of preferential vacancy-drag diffusion for Ni, Si and Mn, these elements were enriched near bubbles. Due to the strong binding energies of vacancies with these solute atoms, the vacancy diffusivity can be reduced near these solutes. Therefore, the more prominent Ni-Si-Mn clustered near helium bubbles in T91 lead to stronger suppression of helium bubble growth compared to F82H.

36 MATERIALS SCIENCE↗

DOE OSTI2023

Halide Substitution Effects on Lithium-Ion Diffusion in Protonated Antiperovskites

Solid-state electrolytes (SSEs) for all-solid-state lithium-ion batteries are generating intense interest because these batteries can improve the safety and performance compared with devices fabricated with conventional, flammable liquid electrolytes. In these SSEs, it has been suggested that Li + ion diffusion in the grain boundaries is hindered and is a critical determinant of the overall ionic conductivity (σ). However, Li + ion diffusivities in the grain (D G ) and the grain boundary (D GB ) are difficult to determine experimentally, with few techniques capable of distinguishing the individual contributions. Here, we distinguished the D G and D GB for the protonated lithium antiperovskites (pLiAPs) SSEs: Li 2 OHCl, Li 2 OHBr, Li 2 OHF 0.1 Cl 0.9 , Li 2 OHF 0.1 Br 0.9 , and Li 2 OHCl 0.3 7Br 0.63 . The measurements were obtained directly from 7 Li pulsed-field gradient nuclear magnetic resonance (PFG-NMR) at 353 K. The 7 Li PFG-NMR echo profiles were composed of two primary components with additional secondary oscillatory components – the so-called NMR diffraction phenomenon. The length scale separating the two main components corresponds to a diffusion length of ~1.7 µm, which is thought to be the average grain size (by diameter). The short-range (≤ 1.7 µm) diffusion component associated with D G (≈10 -11 m 2 /s) varied minimally with halide substitution, while the long-range (≥ 1.7 µm) component D GB (≈10 -12 to 10 -15 m 2 /s) was highly sensitive to the substitution of halides and closely correlated with s. In addition, from the comparison of the ratio D GB /D G to D t (the Li + ion diffusion coefficient estimated from the rotational correlation time, t c ), it was determined that the contribution of D G to σ is negligible; 0.01 ~ 0.04 in the pLiAPs studied here. Finally, these insights provide fundamental understanding of the halide substitution effects on Li + ion grain versus grain boundary diffusion, and suggest that careful engineering of the grain boundaries at the microscopic level is necessary to achieve high-performance pLiAP SSEs.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

NASA NTRS1987

Permeation of oxygen through high purity, large grain silver

The permeation of high purity, large grain Ag membranes by oxygen has been studied over the temperature range 400 to 800 C. The permeability was found to be quite linear and repeatable, but the magnitude was 3.2 times smaller than that determined by past research. Since previous investigators studied substantially less pure Ag and conducted experiments within much poorer vacuum environments (which indicates that their grain boundary density was much greater), the data presented here suggest oxygen transport through the membrane is primarily by grain boundary diffusion. The diffusivity measurements were found to exhibit two distinct linear regions, one above and one below a critical temperature of 630 C. The high-temperature data have an activation energy (11.1 kcal/mole) similar to that reported by others, but the low-temperature data have a higher activation energy (15.3 kcal/mole), which can be explained by impurity trapping in the grain boundaries. Vacuum desorption of the oxygen-saturated Ag was found to occur at a threshold of 630 C, which is consistent with the onset of increased mobility within the grain boundaries.

Outlaw, R. A.↗

DOE OSTI2025

Additive Manufactured Composite Phase-Change Material for Thermal Energy Storage Applications

Phase-change materials play a critical role in industrial energy storage applications to drive efficiency improvements, thermal energy management, and carbon emissions reductions. Recently, it has been shown that rapid solidification of alloys with metastable immiscibility in the liquid phase has the potential to form unique microstructures in which a low-melting phase is uniformly distributed in a high-melting matrix. This feature can be exploited using additive manufacturing to produce components with complex geometries containing such unique phase-change microstructures. Phase-field simulations utilizing high-performance computing were used to provide a detailed description of the evolution of the active phase during service in terms of their morphology and composition in different polycrystalline matrix grain morphologies that are typically produced during additive manufacturing. Phase field simulations were performed using, MEUMAPPS-SL (Microstructure Evolution Using Massively Parallel Phase-field Simulations – Solid Liquid) code that was developed in-house by the Oak Ridge National Laboratory. The simulations utilized the capabilities of the Kestrel supercomputer at the National Renewable Energy Laboratory. The simulation results were compared with experimental results generated at Siemens Energy, Inc. The results indicate that the kinetics of liquid spreading along grain boundaries is largely determined by the mobility of the triple line along the intersection of the grain boundary liquid and the grain boundary plane.

25 ENERGY STORAGE↗

NASA NTRS1994

A scanning defect mapping system for semiconductor characterization

We have developed an optical scanning system that generates maps of the spatial distributions of defects in single and polycrystalline silicon wafers. This instrument, called Scanning Defect Mapping System, utilizes differences in the scattering characteristics of dislocation etch pits and grain boundaries from a defect-etched sample to identify and count them. This system simultaneously operates in the dislocation mode and the grain boundary (GB) mode. In the 'dislocation mode,' the optical scattering from the etch pits is used to statistically count dislocations, while ignoring the GB's. Likewise, in the 'grain boundary mode' the system only recognizes the local scattering from the GB's to generate grain boundary distributions. The information generated by this instrument is valuable for material quality control, identifying mechanisms of defect generation and the nature of thermal stresses during the crystal growth, and the solar cell process design.

Sopori, Bushnan L.↗

DOE OSTI2021

A study on the creep behavior of alloy 709 using in-situ scanning electron microscopy

In this research, an experimental evaluation of creep properties of Alloy 709 in the temperature range of 750–850 °C was undertaken. Alloy 709 is a novel austenitic stainless steel with 20% Cr and 25% Ni by wt% that was developed for application in structural components of nuclear power plants. Creep rupture tests were conducted in an in-situ heating-loading and Scanning Electron Microscope (SEM) unit equipped with Electron Backscatter Diffraction (EBSD) detector and Energy Dispersive Spectroscopy (EDS). “Real-time” creep damage mechanisms of Alloy 709 at various stresses and temperatures using a flat, un-notched sample with continuously reducing cross-section is studied so that the failure and maximum creep damage occurred at the center of the sample where the in-situ SEM imaging could be focused. Accelerated creep tests at temperatures and stresses above service conditions were performed by employing multiple blocks of constant loads where the loads were increased once the sample attained constant creep rate, indicating a secondary creep regime. This technique ensures multiple data points can be obtained from the same test, saves the time required for an otherwise long-term creep test and usage of SEM. Further, Coincident Site Lattice (CSL) boundary maps were collected as control maps before testing, and the grain boundaries were observed during the creep test to understand the effect of grain boundary character on the creep damage mechanism. Void growth, grain boundary separation, and sliding were found to be the main creep mechanisms whose rate is dependent on stress and temperature. Failure mechanisms studied on the fracture surface using SEM fractography were correlated to the sample surface observations to create complementary information to better understand the underline creep mechanism of Alloy 709.

36 MATERIALS SCIENCE↗

NASA NTRS2006

Molecular-dynamics Simulation-based Cohesive Zone Representation of Intergranular Fracture Processes in Aluminum

A traction-displacement relationship that may be embedded into a cohesive zone model for microscale problems of intergranular fracture is extracted from atomistic molecular-dynamics simulations. A molecular-dynamics model for crack propagation under steady-state conditions is developed to analyze intergranular fracture along a flat 99 [1 1 0] symmetric tilt grain boundary in aluminum. Under hydrostatic tensile load, the simulation reveals asymmetric crack propagation in the two opposite directions along the grain boundary. In one direction, the crack propagates in a brittle manner by cleavage with very little or no dislocation emission, and in the other direction, the propagation is ductile through the mechanism of deformation twinning. This behavior is consistent with the Rice criterion for cleavage vs. dislocation blunting transition at the crack tip. The preference for twinning to dislocation slip is in agreement with the predictions of the Tadmor and Hai criterion. A comparison with finite element calculations shows that while the stress field around the brittle crack tip follows the expected elastic solution for the given boundary conditions of the model, the stress field around the twinning crack tip has a strong plastic contribution. Through the definition of a Cohesive-Zone-Volume-Element an atomistic analog to a continuum cohesive zone model element - the results from the molecular-dynamics simulation are recast to obtain an average continuum traction-displacement relationship to represent cohesive zone interaction along a characteristic length of the grain boundary interface for the cases of ductile and brittle decohesion. Keywords: Crack-tip plasticity; Cohesive zone model; Grain boundary decohesion; Intergranular fracture; Molecular-dynamics simulation

Yamakov, Vesselin I.↗

DOE OSTI2020

Texture and precipitation behavior of friction stir welded Inconel 825 alloy

The microstructure, texture pattern, and precipitation behavior of friction stir butt welded Inconel 825 alloy are highlighted here. The microstructure of bright regions and black-banded regions in the stir zone (SZ) is characterized by SEM, EBSD, and TEM with EDS. EBSD analysis justifies that the SZ exhibits a dynamic recrystallized structure. Grain boundary mapping corroborates the insignificant difference in the low- and high-angle grain boundaries fraction between the SZ, the thermo-mechanically affected zone, and the base metal. TEM analysis confirms the formation of nano twins and dispersion of nano-size precipitations in the grains, and at the grain boundaries as well. Grain refinement and nano-size precipitation formation in the SZ significantly enhanced both microhardness and tensile strength.

Das, Hrishikesh↗

NASA NTRS1986

The effect of diffusive hydrogen segregation on fracture of polycrystalline nickel

Hydrogen embrittlement of polycrystalline nickel was studied under conditions which allowed the diffusion and segregation of hydrogen to grain boundaries to be controlled. It was shown that even at high solute concentrations of hydrogen, the fracture mode was ductile shear rupture, if transport of hydrogen to the grain boundaries was not allowed. As the amount of hydrogen segregation to the grain boundaries was increased, the fracture mode became increasingly intergranular. Analysis of the temperature dependence of this transition from ductile rupture to transgranular fracture yields a hydrogen segregation enthalpy of 11.6 kJ/mol. The transition in the fracture mode was also observed at a constant aging temperature by increasing the solute hydrogen concentration. The kinetics of this fracture transition suggest that the hydrogen segregation layer extends about 35 nm from the grain boundary. It is suggested that the fracture mechanism is a softening of the material in this segregation region due to the high hydrogen concentration.

Lassila, D. H.↗