Atomistic modelling of segregation and precipitation in fe cr alloys under irradiation

Atomistic Modelling of Segregation and Precipitation in Fe Cr Alloys under Irradiation 2nd Int Workshop Irradiation of Nuclear Materials Flux and Dose Effects November 4 6, 2015, CEA – INSTN Cadarache[.]

EPJ Web of Conferences 115, 03002 (2016) DOI: 10.1051/epjconf/201611503002 © Owned by the authors, published by EDP Sciences, 2016 2nd Int Workshop Irradiation of Nuclear Materials: Flux and Dose Effects November 4-6, 2015, CEA – INSTN Cadarache, France Centre of Excellence for Nuclear Materials Atomistic Modelling of Segregation and Precipitation in Fe-Cr Alloys under Irradiation Frédéric SOISSON1, Chu Chun FU1, Thomas JOURDAN1, Maylise NASTAR1, Oriane SENNINGER1, Enrique MARTINEZ2, Yves BRÉCHET3 CEA-DEN-DMN, Service de Recherches de Métallurgie Physique, SRMP (Saclay, France) Los Alamos National Laboratory, Materials Science in Radiation Dynamics Group (Los Alamos, USA) SIMAP, INP Grenoble, CNRS UJF (Saint Martin d’Hères, France) Iron-chromium alloys are the basis for ferritic and ferritic-martensitic steels that will be used in future fission (generation IV) and fusion nuclear reactors With Cr content between typically to 12% [1], or even 14% in the matrix of some oxide dispersion-strengthened steels [2], one can expect the precipitation of a Cr-rich α’ phase that can be strongly accelerated under irradiation, due to point defect supersaturation This precipitation can cause hardening and embrittlement Radiation-induced segregation (RIS) is another important technological problem It can lead to a Cr depletion at grain boundaries and therefore to a loss of corrosion resistance and again, to embrittlement RIS in austenitic steels is well known and presents almost systematic trends: depletions of Cr and enrichments of Ni at grain boundaries In ferritic steels, the experimental situation is far from being so clear: depletions and enrichments of Cr have been observed, without clear correlation with the irradiation conditions and the materials properties [3] Segregation and precipitation occur by formation, migration and elimination of point defects (vacancies and self-interstitials) We present a model that includes these mechanisms, and combines ab initio and atomistic kinetic Monte Carlo (AKMC) simulations Migration barriers and jump frequencies are computed by using effective pair interactions with concentration and temperature dependences Thanks to these dependences, the simulations are able to well reproduce the thermodynamic [4] and diffusion [5] properties of dilute and concentrated alloys, including the effects of the magnetic configurations and magnetic transitions, which are especially important in Fe-Cr alloys They also give kinetics pf precipitation during isothermal annealing in good agreement with experimental studies [5] Simulations of α’ precipitation under irradiation, predict that the kinetics can be considerably accelerated For example, irradiations of alloys with 15 to 18%Cr at 290°C and a dose rate of 3.4×10-7 dpa.s-1 produce a point defect supersaturations and therefore, an acceleration, of approximately 6-7 orders of magnitudes To get a more precise estimation of the evolution of sink densities and point defect concentrations, cluster dynamics [6] have been used: the physical time scale of the Monte Carlo simulations is rescaled accordingly and one gets precipitation kinetics in good agreement with the few available experimental results [7,8] The possible effect of ballistic mixing occurring in displacement cascades and the role of carbon, which is known to strongly interact with the vacancies, will be discussed RIS results from the elimination of excess vacancies (V) and self-interstitials (I) on sinks such as dislocations, grain boundaries, or free surfaces Permanent fluxes of point defects towards the sinks are then sustained, i.e permanent fluxes of chemical species, leading to a modification of the local composition near the sinks RIS can be analyzed properly in the framework of the thermodynamics of irreversible processes, that gives the flux of a specie α (atom or defect) as a function of the gradients of chemical potentials μβ : Jα =-∑Lαβ∇μβ The tendencies to solute enrichment or depletion near the sinks are controlled by the Onsager coefficients, Lαβ In Fe-Cr alloys, one must observe a depletion of Cr near the sinks if LCrV/LFeV > LCrI/LFeI, an enrichment of Cr in the opposite case Onsager coefficients can be directly estimated in Monte Carlo simulations (Fig 1a) Simulations have been also performed to model directly the evolution of point defects and solute distributions near a grain boundary (Fig 1b) As expected from the analysis of the Onsager coefficients, at 650 K the interstitial contribution dominates and one observes an enrichment of Cr On the contrary at 950K vacancy contribution is This is an Open Access article distributed under the terms of the Creative Commons Attribution License 4.0, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited 2nd Int Workshop Irradiation of Nuclear Materials: Flux and Dose Effects November 4-6, 2015, CEA – INSTN Cadarache, France Centre of Excellence for Nuclear Materials dominant and Cr depletion occur [9] We will compare these trends with experimental observations of RIS Finally, the AKMC enables to study the interaction between segregation and precipitation [9]: typical cases will be presented (b) (a) Fig 1: (a) Evolution of the ratios of Onsager coefficients with the temperature, in Fe-Cr alloys of different compositions, estimated by Monte Carlo simulations, (b) Monte Carlo simulation of radiation induced segregation of Cr near a grain boundary (steady-state concentration profiles of point defects and concentration profiles of Cr at two successive doses) References [1] R.L Klueh, A.T Nelson, Ferritic/martensitic steels for next-generation reactors, J Nucl Mater 371 (2007) 37 [2] J.-L Boutard et al, Oxide dispersion strengthened ferritic steels: a basic research joint program in France, J Nucl Mater 455 (2014) 605 [3] Z Lu et al, Irradiation-induced grain boundary chromium microchemistry in high alloy ferritic steels, Scripta Mater 58 (2008) 878 [4] M Levesque, E Martínez, C.-C Fu, M Nastar and F Soisson, Simple concentration-dependent pair interaction model for large-scale simulations of Fe-Cr alloys, Phys Rev B 84, 184205 (2011) [5] O Senninger, E Martinez, F Soisson, M Nastar and Y Bréchet, Atomistic simulations of decomposition kinetics in Fe-Cr alloys: influence of the magnetism, Acta Mater 73, 97–106 (2014) [6] E Meslin et al, Cluster-dynamics modelling of defects in α-iron under cascade damage conditions, J Nucl Mater 382 (2008) 190 [7] M Bachhav, G R Odette, and E A Marquis, α’ precipitation in neutron-irradiated Fe–Cr alloys, Scripta Mater 74 (2014) 48 [8] V Kuksenko, C Pareige and P Pareige, Cr precipitation in neutron irradiated industrial purity Fe–Cr model alloys, J Nucl Mater 432 (2013) 160 [9] O Senninger et al., Modeling radiation induced segregation in Fe-Cr alloys, submitted to Acta Mater Atomistic Modeling of Segregation and Precipitation in Fe-Cr Alloys under Irradiation C.-C Fu, T Jourdan, M Nastar, O Senninger, F Soisson SRMP, CEA Saclay E Martinez Los Alamos National Laboratory Y Bréchet SIMAP, Grenoble INP 2nd Int MINOS Workshop, Irradiation of Nuclear Materials: Flux and Dose Effects November 4-6, 2015, CEA – INSTN Cadarache, France Fe-Cr alloys: nuclear applications • Fe-Cr alloys : a model for ferritic and ferritic-martensitic steels (7-18%Cr) candidate materials for future nuclear reactors (Gen IV and fusion) 1200 Potential problems under irradiation: - α’ precipitation  hardening and embrittlement - Cr depletion at GBs  corrosion, embrittlement T (K) Due to: point defect supersaturation Acceleration of precipitation Radiation Induced Segregation (RIS) Radiation Induced Precipitation (RIP) TC 1000 σ α α' 800 α + α' 600 400 0.0 0.2 0.4 0.6 0.8 • Main objectives: c - quantify the acceleration of precipitation in supersaturated alloys - understand the mechanisms controlling RIS - study possible effects of equilibrium vs non-equilibrium segregation, ballistic mixing, carbon atoms,… on the precipitation and segregation kinetics • Our approach : Atomistic Kinetic Monte Carlo (AKMC) simulations CEA – DEN 2nd Int MINOS Workshop - November 4-6, 2015, CEA – INSTN Cadarache, France 1.0 Diffusion model - AKMC Simulations Pair interactions on stable (bcc) and saddle-point positions • - Diffusion by jumps of : Vacancies (V) Self-interstitial atoms (SIA) : dumbbells Direct interstitials (C) • The migration barriers are computed with a brokenbond model: - with composition and temperature dependent pair interactions - with saddle-point pair interactions - fitted on DFT calculations Γ AV Γ BV “grain boundary” = perfect pd sink Sink strength ktot  12 / L2 • Formation of isolated Frenkel with replacement collision sequences or small replacement cascades • Annihilation of V and SIA on a perfect sink CEA – DEN L = 36 to 104 nm 2nd Int MINOS Workshop - November 4-6, 2015, CEA – INSTN Cadarache, France Thermodynamics: effective pair interactions ( n) ( n) ( n) Concentration dependent pair interactions: hFeFe , hCrCr , hFeCr (x) fitted on DFT calculations of ΔHmix at 0K Magnetic and vibrational contributions: linear temperature dependence, fitted on T α-α’ (exp) M Levesque et al, Phys Rev B 84, 184205 (2011) PWSCF, GGA-PAW  Special Quasi-random structures (SQS)  Ordered structures Pair interaction model ( n) ( n) ( n) gFeCr (x ,T )  hFeCr (x)  TsFeCr (x) , with n  1,2 Hmix  (x) x(1  x) (x)   n zn ( n ) ( n) ( n) hFeFe  hCrCr  2hFeCr (x) x CEA – DEN 2nd Int MINOS Workshop - November 4-6, 2015, CEA – INSTN Cadarache, France Diffusion coefficients Tracer diffusion coefficients (pure iron) TαγTC -14 10 600°C Interdiffusion coefficients in Fe-Cr alloys 10-6 10-7 -15 10 DCr* D (cm2.s-1) -1 10 10-9 Fe -17 -18 -19 10 Fe ~ D (m s ) 10 DFe* -20 10 -21 10 -22 10 γ-Fe α-Fe  para 0.7 1453 K [2] 10-10 10-11 1124.2 K [1] -12 10 10-13 994.5 K [1] -14 10 914.8 K [1] 10-15 α-Fe ferro -16 10 -23 10 1713 K [2] 10-8 AKMC -16 10 [1] Braun & Feller-Kniepmeier, 1985 [2] Jönsson 1995 Exp [1,2] AKMC 500°C 850 K -17 0.8 0.9 1.0 1.1 1.2 1.3 -1 1.4 10 10 20 30 40 50 60 70 80 90 100 % Cr 1000/T (K ) O Senninger et al, Acta Mater 73, 97–106 (2014) - CEA – DEN Acceleration of the diffusion at the F/P transition The Curie temperature decreases with the Cr concentration 2nd Int MINOS Workshop - November 4-6, 2015, CEA – INSTN Cadarache, France Kinetics of α- α’ decomposition: isothermal annealing Fe-20%Cr T = 500°C AKMC (E Martinez et al, 2012) 3DAP (Novy et al, 2009) 1200 T (K) 1000  800 ' 600  ' 400 0.0 0.2 0.4 0.6 0.8 1.0 xCr E Martinez et al, Phys Rev B 86, 224109 (2012) S Novy et al, J Nucl Mater 384 (2009) 96–102 CEA – DEN 2nd Int MINOS Workshop - November 4-6, 2015, CEA – INSTN Cadarache, France Kinetics of α- α’ decomposition: isothermal annealing Small-angle neutron scattering experiments (SANS) 500°C: Bley (1992) 540°C: Furusaka et al (1986) 500°C 540°C  SANS experiments  AKMC simulations  AKMC simulations without magnetic acceleration O Senninger et al, Acta Mater 73, 97–106 (2014) CEA – DEN 2nd Int MINOS Workshop - November 4-6, 2015, CEA – INSTN Cadarache, France Radiation Accelerated Precipitation: experiments Experimental evidences (3DAP) in Fe-9% to 20%Cr alloys at 290-300°C - isothermal annealing: α’ precipitation is not observed due to slow kinetics - neutron irradiation: α’ precipitation Bachhav et al (Scripta Mater 2014) Fe-3%Cr to 18%Cr, T= 290°C 3.4 x 10-7 dpa/s, 1.82 dpa 3%Cr 6%Cr 9%Cr 12%Cr 15%Cr 18%Cr • P • Si Cr rich clusters V Kuksenko et al, JNM 432 (2013) 160 C Pareige et al JNM 456 (2015) 471–476 Precipitation under neutron irradiation in Fe-9%Cr and Fe-12%Cr, but not under ion irradiation at higher flux CEA – DEN 2nd Int MINOS Workshop - November 4-6, 2015, CEA – INSTN Cadarache, France Radiation Accelerated Precipitation: AKMC AKMC simulations Fe-18%Cr @ 563 K, 3.4 x 10-7 dpa.s-1 Point defect concentration profiles 104 nm GB Precipitate free zones near the GBs CEA – DEN 3DAP Kuksenko et al (2012) Fe-12%Cr 300°C – 0.6 dpa 2nd Int MINOS Workshop - November 4-6, 2015, CEA – INSTN Cadarache, France 10 Radiation Accelerated Precipitation: AKMC vs 3DAP • AKMC simulations: one GB in the simulation box (constant sink strength: ktot  12 / L2 ) • Cluster Dynamics: evolution of the sink strength ktot (in pure iron) estimation of the point defect concentration ci ,v  G / (ktot Di ,v ) c (AKMC )  rescaling of the Monte Carlo time scale t  tMC  i ,v ci ,v (CD)  better irradiation 3.4 x 10-7 dpa.s-1 3DAP (Bachhav et al, 2014) AKMC - k2tot = Cte 19 1020 Fe-15%Cr @ 563 K thermal ageing AKMC - k2tot (CD) 10 dp (cm-3) dp (cm-3) 1020 3DAP (Bachhav et al, 2014) AKMC - k2tot = 12/(L2) 2.5 2.5 2.0 2.0 1.5 1.0 0.5 0.0 10-1 100 101 102 103 104 105 106 107 108 109 1010 1011 1012 1013 t (s) thermal ageing 1019 1018 100 R (nm) R (nm) 1018 irradiation 3.4 x 10-7 dpa.s-1 Fe-18%Cr @ 563 K AKMC - k2tot (CD) 101 102 103 104 105 106 107 108 109 1010 1011 1012 1013 101 102 103 104 105 106 107 108 109 1010 1011 1012 1013 1.5 1.0 0.5 0.0 100 t (s) Strong acceleration by irradiation (x 106-107) Good agreement with the experiments of neutron irradiation (Bachhav et al, 2014) F Soisson, T Jourdan, Acta Mater 2016 CEA – DEN 2nd Int MINOS Workshop - November 4-6, 2015, CEA – INSTN Cadarache, France 11 Effect of ballistic mixing • Experiments: Cr precipitation under neutron irradiation, but not under ion irradiation at higher flux A ballistic dissolution of α’ precipitates, in displacement cascades ? • AKMC simulations - at different dose rates G (dpa.s-1) - with different numbers of replacements/displacement • At 290°C Γ G γ = bal  st Γth cv Dv 3 10  10 4 1020 dp (cm-3) Nrep/Ndis = (channeling) Nrep/Ndis = 10 (replacement collision sequences) Nrep/Ndis = 100 (cascades) 1019 3.4 x 10-7 dpa.s-1 10-3 dpa.s-1 Fe-18%Cr @ 563 K 18 10 ci ,v  G / (k Di ,v ) tot  ballistic effects 10-2 10-1 100 101 102 103 104 105 106 10-2 10-1 100 101 102 103 104 105 106 0.9 0.8 0.7 0.6 not explain the difference between ion and neutron irradiations • 10-3 1.0 R (nm) - no dissolution of precipitates - no effect of Nrep/Ndis on the precipitation kinetics - dose rate effects: a simple acceleration due to the point defect supersaturation 0.5 10-3 t (s) Possible ballistic effects at lower temperatures (below 100°C) CEA – DEN 2nd Int MINOS Workshop - November 4-6, 2015, CEA – INSTN Cadarache, France 12 10 10-12 Radiation-Induced SegregationCd 10-13 10-14 Ji   Lij μ j j Fe-10%Cr 650 K vacancies 10-15 -6 -1 JV/JCr : negative coupling  Cr depletion sinks, dominantself-interstitials at high T 10 at dpa.s -16 10 JSIA/JCr : positive coupling  Cr enrichment sinks, dominant at40low -40 at-20 20 GB 0.20 Fe-10%Cr T = 650 K, 10-6 dpa.s-1 0.007 dpa 0.294 dpa 10-9 0.15 10-10 0.10 -11 Cr 10 d 0.05 10-12 X C -13 0.00 10 10-14 Fe-10%Cr 950 K -3 -1 10 -40dpa.s-20 -40 vacancies self-interstitials 20 40 distance (nm) -20 20 40 GB 0.20 Fe-10%Cr T = 950 K, 10-3 dpa.s-1 0.012 dpa 0.636 dpa 0.15 XCr 0.10 0.05 Steady-state profile: 0.00 -40 -20 20  LCrV LCrI  distance (nm) O Senninger et al  C     C nd   V Int MINOS CEA – DEN Workshop - November 4-6,(2016) 2015, CEA – INSTN Cadarache, France Cr Acta Materialia 103, 1–11  LFeV LFeI  40 13 T Radiation-Induced Segregation Ji   Lij μ j j JV/JCr : negative coupling  Cr depletion at sinks, dominant at high T JSIA/JCr : positive coupling  Cr depletion at sinks, dominant at low T GB e seg  0.1 eV Cr 0.30 0.06 dpa Fe-10%Cr T = 650 K, 10-6 dpa.s-1 0.25 0.20 XCr 0.15 0.10 0.05 -40 -20 20 40 distance (nm) seg Cr e  0.1 eV GB 0.20 0.338 dpa Fe-10%Cr T = 950 K, 10-3 dpa.s-1 0.15 XCr 0.10 Steady-state profile: 0.05  LCrV LCrI  distance (nm)  C      CV 2nd Int MINOS Workshop - November 4-6, 2015, CEA – INSTN CEA – DEN Cadarache, France Cr W-shape profile  LFeV LFeI  -40 -20 20 40 14 Radiation-Induced Precipitation (AKMC) Undersaturated alloys, at low T: strong Cr enrichment on sinks  RIS  radiation induced precipitation 10-10 1200 10-11 T = 563 K Fe-9%Cr 10-6 dpa.s-1 vacancies self-interstitials 10-13 10-14 10-15 10-16 T (K) 10 Cd TC 1000 -12 σ α α' 800 α + α' 600 400 -17 10 -40 -20 20 40 d (nm) 0.0 0.2 0.4 c 0.6 0.8 1.0 0.20 0.20 dpa 0.15 CCr 0.10 0.05 -40 -20 20 40 d (nm) CEA – DEN 2nd Int MINOS Workshop - November 4-6, 2015, CEA – INSTN Cadarache, France 15 CONCLUSIONS AKMC simulations with thermodynamic and point defect parameters fitted on DFT calculations Good description of driving forces, diffusion properties and nucleation In Fe-Cr alloys • Magnetic effects are important (impact on thermodynamic and diffusion properties) • Radiation Induced Segregation is controlled by a balance between opposite effects of V and SIA the Lij are very dependent of the details of migration barriers  may explain the variability of experimental studies (≠ RIS in austenitic steels) • Radiation accelerated precipitation - a good agreement with neutron irradiations (with rescaling of the sink strength using CD) - no effect ballistic mixing in the experimental conditions (relatively high temperatures, high sink densities)  does not explain the difference between neutron an on irradiation - precipitation kinetics is less sensitive than Radiation Induced Segregation to the details of migration barriers • Related work, perspectives - Phase-Field model for RIS (J.B Piochaud, L Thunier, A Legris – UMET, Lille) - Precipitation under electron irradiation (O Tissot, et al) - Effects of C (or O, N) on the kinetics of precipitation and segregation CEA – DEN 2nd Int MINOS Workshop - November 4-6, 2015, CEA – INSTN Cadarache, France 16 ... Pareige and P Pareige, Cr precipitation in neutron irradiated industrial purity Fe? ? ?Cr model alloys, J Nucl Mater 432 (2013) 160 [9] O Senninger et al., Modeling radiation induced segregation in Fe- Cr. .. segregation in Fe- Cr alloys, submitted to Acta Mater Atomistic Modeling of Segregation and Precipitation in Fe- Cr Alloys under Irradiation C.-C Fu, T Jourdan, M Nastar, O Senninger, F Soisson SRMP,... Pair interaction model ( n) ( n) ( n) gFeCr (x ,T )  hFeCr (x)  TsFeCr (x) , with n  1,2 Hmix  (x) x(1  x) (x)   n zn ( n ) ( n) ( n) hFeFe  hCrCr  2hFeCr (x) x CEA – DEN 2nd Int