Tuning Plasma by the Nanoscale Lithium-Dopped Surfaces

Transcription

1 Tuning Plasma by the Nanoscale Lithium-Dopped Surfaces Predrag Krstić Joint Institute of Computational Sciences, U. of Tennessee Ex: Physics Division, Oak Ridge National Laboratory Support from: US DOE (OFES), ORNL LDRD program NCCS(DOE) & NICS(NSF) Computing for the Department of Energy LTP, June 01/2012

2 Theory: Carlos Reinhold PD, ORNL Steve Stuart Clemson U. Paul Kent ORNL Alain Allouche CNRS, Fr Stephan Irle Nagoja U, Japan J. Jakowski, NICS K. Morokuma, Kyoto U. CMD CMD Mat. PWDFT Close collaborators: TBDFT modeling Students Jonny Dadras (UTK) Eric Yang (Purdue) Chris Ehemann (MSTD) Jae Park (ORISE) Experiment: Fred Meyer (PD, ORNL) JP Allain (Purdue) Chase Taylor (Purdue) Eric Hollmann (UCSD) Jeff Harris (FED, ORNL) Rick Goulding (Fed, ORNL) beam beam beam plasma plasma plasma H. Witek, Taiwan Our thanks to John Hogan (FED, ORNL) Chris (MTSU) for the 2 Department of Energy OFES LTP, June Review, 01/2012 ORNL, March 02, 2010

3 1. Why PLASMA-MATERIAL INTERFACE is such an important problem? for the Department of Energy OFES LTP, June Review, 01/2012 ORNL, March 02, 2010

4 All energy from D-T fusion reactions passes through first wall Flux of (particles + heat + 14 MeV neutrons) ~10 MW/m 2 Supercon ducting magnet Schematic magnetic fusion reactor Plasma a Plasma heating (rf, microwave,...) Turbine generator Why is PMI important? Why lithium? Carbon? Shield Vac. Blanket Unlike nuclear fission where energy is volume-distributed A FUSION REACTOR IMPLIES MANY INTERFACES BETWEEN THE PLASMA AND MATERIALS Particles and surfaces Key role of PMI in fusion research well recognized in US and internationally for the 4 Department of Energy OFES LTP, Review, June 01/2012 ORNL, March 02, 2010

5 Guiding principle: If Edison had a needle to find in a haystack, he would proceed at once with the diligence of the bee to examine straw after straw until he found the object of his search I was a sorry witness of such doings, knowing that a little theory and calculation would have saved him 90% of his labor. Nikola Tesla, New York Times, October 19, 1931 The traditional trial-and-error approach to PMI for future fusion devices by successively refitting the walls of toroidal plasma devices with different materials and component designs is becoming prohibitively slow and costly Need bottom-up approach arising from the fundamental atomistic and nano science for the 5 Department of Energy OFES LTP, Review, June 01/2012 ORNL, March 02, 2010

6 2.Why bottom-up science? or How to build an effective science for PMI? for the Department of Energy OFES LTP, June Review, 01/2012 ORNL, March 02, 2010

7 Probing the PMI requires integration of many experimental and theoretical techniques spanning orders of magnitude in time, length, and energy scales e.g., quartz crystal microbalance e.g., secondary neutral mass spectrometry e.g., low energy ion scattering, x-ray photoelectron spectroscopy Monte-Carlo techniques Diffusion; transport for the 7 Department of Energy OFES LTP, June Review, 01/2012 ORNL, March 02, 2010 e.g., Rutherford backscattering, elastic recoil detection Courtesy of J.P. Allain

8 What does flux of particles/m 2 s mean (ITER) for a typical atomistic (MD) simulation? for the Department of Energy LTP, June 01/2012 At a box of surface of 3 nm lateral dim? a few thousands atoms (carbon) The flux is 0.01 particle/nm 2 ns 1) 1 particle at the interface surface of the cell each 10 ns. But for deuterium with impact energy less then 100 ev: Penetration is less than 2 nm, typical sputtering process takes up to 50 ps Is each impact independent, uncorrelated? Each particle will functionalize the material, change the surface for the subsequent impact! Processes essentially discrete Atomistic approach!!! OFES Review, ORNL, March 02, 2010

9 3. Why is PMI so difficult problem? Interfacial physics, when the two worlds meet : traditionally the most challenging areas of science Dynamical surface communicates between two worlds: Plasma and Material for the Department of Energy OFES LTP, June Review, 01/2012 ORNL, March 02, 2010

10 PMI has many fundamental processes & synergies When an ion or neutral arrives at a surface it undergoes a series of elastic and inelastic collisions with the atoms of the solid. elastic reflection trapping/detrapping retention diffusion, permeation implantation Drivers: Plasma Material Multi -T, -n, -species, plasma irradiation, re-emission & neutrons sputtering & sheath acceleration chemistry Give rise to synergistic effects for 10 the Department of Energy OFES LTP, June Review, 01/2012 ORNL, March 02, 2010 Damage Effects: Vacancies, bubbles, blisters, dislocations, voids, neutrons? Re-deposition Co-deposition Erosion Ablation Melting (metals)

11 Materials exposed to plasma are modified, resulting in a dynamical surface Deuterium impact of carbon Amorphization depends on penetration depth rather than of deposited energy Methane sputtering requires H loading of the surface sp 3 density (1/A 3 ) 7.5 ev/d 15 ev/d ev/d 20 ev/d impact Chemical sputtering of hydrocarbons surface 30 ev/d Depth (A) Plasma irradiation results in a z<5 different surface Sputtering Yield/D Counts of R-CD x moieties, x=1,2,3 1x10-2 8x10-3 6x10-3 4x10-3 2x sp 3 sp 2 CD 3 CD 2 CD sp E (ev/d) Impact energy (ev/d) Krstic et al, New J. of Phys. 9, 219 (2007). for 11 the Department of Energy OFES LTP, June Review, 01/2012 ORNL, March 02, 2010 Normalized Mass 20 signal number density (a.u.) ev / D Surface morphology wall contribution to saturated signal: ~ 10 % Beam ON Time (s) ev / D 1.00 wall contribution to impact saturated signal: ~ 20 % Beam ON background level saturated C density signal background level saturated signal Time (s) 0 Fluence (10 20 m -2 ) surface depth (A) H density Reaching steady state

12 4. How to validate theory with experiments (and vv)? for the Department of Energy LTP, June 01/2012 OFES Review, ORNL, March 02, 2010

13 What have we learned from the next door beam-surface experiments? Beam-surface exp t: precision control of projectiles & targets enabled development & validation of MD approach Remarkable agreement of theory & exp t when simulation mimics exp t. No fitting parameters! Key: simulation prepares surface by bombardment! Fluence (not flux) like that in experiment Type, internal state, energy, angle as in exp t Beam-surface experiments: Prepared beam & target Meyer et al, Physica Scripta T128, 50 (2007). for 13 the Department of Energy OFES LTP, June Review, 01/2012 ORNL, March 02, 2010

14 New physics from theory Counts rotational distribution of ejecta Fit based on Maxwell of 0.34 ev D 2 with D impact (10 ev) H 2 with D impact (10 ev) energy (ev) for the Department of Energy OFES LTP, June Review, 01/2012 ORNL, March 02, 2010 Ejection energy (ev) Hydrocarbon ejection CD 3 ejection (a) Translational Rovibrational Rotational Vibrational (b) D Impact energy (ev) Ejection energy Approximate thermalization indicated Functions of hydrocarbon mass and impacting D energy Ejection temperature Krstic et al, J. Appl. Phys. 104, (2008).

15 5. Is the PMI phenomenology with plasma Irradiation different than with a beam irradiation? Can we derive plasma-irradiated PMI from the beam-pmi? Can we model plasma irradiation? Answer is: Yes, No, Yes for the Department of Energy LTP, June 01/2012 OFES Review, ORNL, March 02, 2010

16 Goal: Study synergy in the evolution of surface irradiated by plasma Experiments with Ar+ and H: ( physical ) Sputtering = (chemical) + Surface preparation by H impact for chemical sputtering Impurity atoms in plasma are efficient precursors for erosion PM processes very dependent on inventory of H in the material He suppresses H retention in W He penetrates deeper than H Strong dependence on energy He bubbles: barrier to H diffusion? T=653 K He: 0.1% He: 0% Ion flux = 3.5 * cm -2 s -1 H flux = 1.4 * cm -2 s -1 H and Ar + physical H alone Ar + alone Hopf & von Keudell, 2003 for the Department of Energy OFES LTP, June Review, 01/2012 ORNL, March 02, 2010

17 Counts V Z Simulations of the plasma irradiation (D atoms) T=10,000K sqrt(e) exp(-e/kt) E (ev) T=10,000K All V Z >0 changed sign uniform V X HOW? Maxwell Impact distributions V Y Yield 0.8 D ejected plasma beam D 2 sputtered Impact energy of D (ev) Krstic et al, 2009 With plasma irradiation: Reflection significantly higher than beam, sputtering suppressed!!! Krstic et al, AIP Conf. Proc. 1161, 75 (2009). for 17 the Department of Energy OFES LTP, June Review, 01/2012 ORNL, March 02, 2010

18 6. What is outreach of the PMI science at the plasma facilities? or Do we need dedicated PMI plasma facility? What do we want to do with it? Difficult to study PMI in thoroidal facilities! Importance of the dedicated PMI facilities Pisces-B, Magnum, ) = Bridge!!! for the Department of Energy LTP, June 01/2012 OFES Review, ORNL, March 02, 2010

19 QuickTime and a Integration of theory & experiment basis for PMI research Beam-surface experiments: Prepared beam & target Molecular Dynamics (MD) & Monte Carlo (MC) simulation Material Science High-flux linear PMI experiment: Well-diagnosed plasma & target MD and MC with plasma synergy Potential models Quantum-classical MD Increases in computational power Toroidal confinement experiments decompressor are needed to see this picture. PMI Design & Qualification ITER, DEMO Predictive science!!! for the Department of Energy OFES LTP, June Review, 01/2012 ORNL, March 02, 2010

20 What have we learned from carbonfacing plasma? PMI extremely difficult interfacial problem (Material mixing create SURFACE entity; its scale depends on impact energy: for sub-100ev => nm-ns scales PMI science can be built from bottom-up recognizing its multiscale character and building from shortest time/spatial scales (fs/angstrom) up Theory&modeling of PMI has to be validated by experiment (and v.v.) Irradiation create dynamical surface, changing interface, cumulative bombardment is the key for agreement with experiment Surface responds to synergy in plasma irradiation (angles, energies, particles), NOT following linear superposition principle; NEED plasma irradiation modeling and experiments; dedicated plasma devices a must for the Department of Energy LTP, June 01/2012 OFES Review, ORNL, March 02, 2010

21 Lithium wall conditioning improves confinement! Why? We know from in-situ experiments labs, and more than 7 different tokamak machines (TFTR, CDX-U, FTU, DIII-D, TJ-II, EAST, and NSTX ) that work with graphite with thin lithium coatings have a "significant" effect on plasma behavior and more specifically on hydrogen recycling. Controlled experiments demonstrated reduced recycling, improved energy confinement time t E, and a reduction of edge instabilities known as edge localized modes (ELMs) Notice the ration of the diemsnions of the plasma and Li layer!!! Initially the experimentalists conjecture was that there was some "functionality" that governed the behavior of the Li-C-O-H system observed indirectly by analyzing the O(1s) and C(1s) peaks. D+ For "some reason" the Li(1s) peaks didn't show much information. ~ 1-10 s nm for the Department of Energy OFES LTP, Review, June 01/2012 ORNL, March 02, 2010 ~ 1 s m

22 Lithium dynamics: Problem to study theoretically because Li polarizing features when interacting with other elements Electronegativity is chemical property of an element defining its tendency to attract electrons: Li has it exceptionally low in comparison to H, C, O, Mo, W. Consequence: Bonding between Li and other atoms covalent and polar; Long-range nonbonding: Coulomb :1/R Lennard-Jones :1/R 6, 1/R 12 for the Department of Energy LTP, OFES June Review, 01/2012 ORNL, March 02, 2010 Electronegativity and size of atoms related!

23 How to treat the dynamics in quaternary system Li-C-O-H Li-C, Li-H, Li-O are of very different electronegativities and therefore these could be considered as ionic solids/liquids As a result of partial charge transfer from one element to the other, the dominant long-distance binding force between particles is the Coulombic attraction between opposite charges, including multipole interactions Classical MD is not a good answer, cannot self-consistently calculate charges (EEM+Tersoff-Brenner covalent models too expensive) Quantum-Classical MD based on Self-Consistent-Charge Density-Functional Tight-Binding (SCC-DFTB) method (developed by Bremen Center for Computational Mat. Science, Germany) a possible answer for qualitative phenomenology is our choice for the Department of Energy OFES LTP, Review, June 01/2012 ORNL, March 02, 2010

24 Simulation Goals: Chemistry and sputtering/reflection/retention dynamics in lithiated&oxidized carbon material, bombarded by slow deuterium atoms is studied. The objectives of this research are two-fold: 1) To develop the realistic methods for computational simulation of the Li-C-O-H, validated by experiments. 2) Experiments from Purdue and NSTX (PPPL) indicate higher retention and lower erosion rate with D whenever Li present in C, however XPS diagnostics show dominating D-O-C chemistry. Why is the question now? C.N. Taylor, B. Heim, J.P. Allain, Journal of Applied Physics 109, (2011) for the Department of Energy LTP, OFES June Review, 01/2012 ORNL, March 02, 2010

25 Simulation of deuterium impact to lithiated and oxidized carbon surface (quantum-classical approach, DFTB) Cell of a few hunreds atoms of lithiated and oxidated amorphous carbon (~20% of Li, and/or ~20% of O), at 300K How? By random seed of Li and O in amorphous carbon and energy minimization, followed by thermalization bombarded by 5 ev D atoms, up to 500fs for the full evolution Perpendicularly to the shell interface 5004 random trajectories (embarrassingly parallel runs at Jaguar, Kraken); Time step 1 fs; several 10,000 CPU hours per run for the Department of Energy OFES LTP, June Review, 01/2012 ORNL, March 02, 2010

26 Slabs studied: Periodicity in x-y C-Li-O Only C+H for the Department of Energy LTP, OFES June Review, 01, ORNL, 2012 March 02,

27 What do experiments teach us? II From experiments: There was correlation between hydrogen irradiation and the behavior change of the O(1s) and C(1s) peaks ONLY IN THE PRESENCE OF LITHIUM. The Li(1s) peak was always invariant???? Intensity (a.u.) Normalized Intensity (a.u.) O 1s C 1s Li 1s x O 1s Post D Bombardment C 1s Li 1s Lithiated Graphite Virgin Graphite Binding Energy (ev) for the Department of Energy OFES LTP, Review, June 01/2012 ORNL, March 02, 2010

28 10 2 Does Li bond more D s? Counts 10 1 C Li D (a) Normalized cumulative distribution Partial charge (e) (b) (a) Distribution of average charges of D, Li and C, with impact of D on a-c:li surface (~23% Li concentration, Li/(Li+C)). D has a slight preference for interacting with Li rather than with C. Krstic et al, FE&D, 2011 for the Department of Energy OFES LTP, Review, June 01/2012 ORNL, March 02, 2010

29 What is model teaching us (I)? Large percentage of impact D (40%) prefer closeness of Li to settle down Not much attention to O (already bonded to Li) But there is only 20% Li in C, < 5% of O The first try of the model had a nice story. The electropositive nature of Li gave us a clear picture on its influence on hydrogen retention when we compared "with Li" and "without Li" cases in the graphite matrix. The model remained somewhat unconnected to the XPS data that clearly showed it was the O(1s) that was really active and correlated to hydrogen irradiation. BUT ONLY IN THE PRESENCE OF LITHIUM. What was missing...??? for the Department of Energy OFES LTP, June Review, 01/2012 ORNL, March 02,

30 What do experiments teach us? III Is it the impact energy? QM model used 5 ev and experiment (cannot afford higher energy ) Comes very interesting and theoretically anticipated result: C. Taylor shows that the chemistry is incident particle energy independent... as expected. But, again a result came by observing the O(1s); in presence of Li Not even the C(1s) showed much. No problem with impact energy!!! for the Department of Energy OFES LTP, Review, June 01/2012 ORNL, March 02, 2010

31 Quantum-mechanical, PWDFT static calculations finger -point in the same direction, Qualitatively benchmark the DFTB findings: graphene bilayer with Li and H on the surface Bonding of H (ev) d=1.73a When a lithium atom is co-adsorbed on surface bonding energy of H grows up to values ranging from -2.2 to -2.5 ev, with decreasing the Li-H distance. (compared with -1.9 ev for pure graphite) -5.0 A B C D E The bonding E enhancement is also observed when Li is sandwiched 1 layer below the surface layer (conf. E) Allouche&Krstic, Carbon (2012) 1.88A Configuration PWDFT 5.71A DFTB 3.53A Charges (e) A B C D E H Configurations Li for the Department of Energy LTP, OFES June Review, 01/2012 ORNL, March 02, 2010

32 What do experiments teach us? IV Result: The quantitative amount of lithium deposition in the in-situ Purdue studies correlated DIRECTLY with results by Maingi et al on amounts of Li deposited in NSTX and subsequent changes on plasma behavior nm THIN FILM only OF LITHIUM MADE A SIGNIFICANT CHANGE IN PLASMA BEHAVIOR of spatial scales of meters!! NO ELMS. But this behavior was limited in time and for some reason saturated. Then comes another pioneering result by C. Taylor showing how this saturation is attained in lithiated graphite: Connected with O saturation for the Department of Energy OFES LTP, Review, June 01/2012 ORNL, March 02, 2010

33 What do experiments teach us? V Here comes the experiment again (Chase): 1) At most 5% oxygen content on the surface of NON- LITHIATED graphite... AS EXPECTED. 2) With lithium you get 10% of Oxygen 3) IMPORTANT: with LOW-ENERGY IRRADIATION one gets over 20% oxygen on the surface.... B/C LITHIUM BRINGS IT THERE WHEN LITHIATED GRAPHITE IS IRRADIATED. for the Department of Energy OFES LTP, Review, June 01/2012 ORNL, March 02, 2010

34 And more: Effective surface exposed to Li Large Polished graphite samples have surface height variations of >1 µm (for a 100 x 100 µm domain) in addition to local height variations on smaller domains, as shown in the inset atomic force micrographs (AFM) in (b-c). Courtesy of C. Taylor The effective film thickness of Li is significantly less than the nominal dose due to the increased effective surface area. $"# 100 µm µm µm 0 nm 100 µm Li rapidly intercalates into graphite following deposition, causing a modest rise in surface oxygen content (10%). 50 µm 50 µm!"# 1 µm 0 µm 0 µm nm nm 0 nm 1 µm But in some samples, D ion bombardment dramatically increases the surface oxygen content to as much as 45%. 0.5 µm 0.5 µm 0 µm 0 µm for the Department of Energy OFES LTP, Review, June 01/2012 ORNL, March 02, 2010

35 Counts % Li 0% O <d> = SD = % Li 5% O <d> = SD = % O 5% Li <d> = SD = % Li 20% O <d> = SD = Penetration depth (Angs) Quantal (250 atoms) What is modeling teaching us? II Multipeak structure is consequence of the discrete structure of the bonding centers Penetration depth? sp 3 density (1/A 3 ) impact surface 5 ev/d 7.5 ev/d 30 ev/d 15 ev/d 20 ev/d Depth (A) Classical (2500 atoms) Interestingly: QM and classical give similar penetrations for a:c (deut.) for the Department of Energy OFES LTP, Review, June 01/2012 ORNL, March 02, 2010

36 How do Li and O compete? Matrix composition: MODEL Nearest Neighbor Occurrence (%) C Li O H A 100% 100% C Li O H Matrix Composition B C D 80% 60% 52% 20% 20% 16% 20% 16% 16% 9% 91% 27% 5% 67% 9% 16% 3% 71% E 80% 20% 30% 70% A B C D E Integrated distribution Normalized Counts (a.u.) Matrix composition: A B C D a) c) e) g) i) b) d) (c) f) h) j) 9% Courtesy of C. Taylor E Indicate that D has a preference for interacting with O and C-O structures rather than with Li or Li-C structures when there is enough O Partial Charge (e) for the Department of Energy OFES LTP, Review, June 01/2012 ORNL, March 02, 2010

37 What do the first neighbors to D say? Again: O preference! (c) for the Department of Energy OFES LTP, Review, June 01/2012 ORNL, March 02, 2010

38 Distribution of binding energies of D in various mixtures consistent with O dominance <E b > = SD = % Li & O <E b > = SD = % O & 5% Li <E b > = <E b > = SD = % Li & O Binding energy (ev) Widths are signatures of various bonding sites SD = % Li & 5% O 40 (b) Counts (d) 0% Li 0% O 20% Li 5% O 20% O 5% Li 20% Li 20% O Final kinetic energy of bound D (ev) Most of the bound D s in the ground Vib state width ~0.2 ev) for the Department of Energy OFES LTP, Review, June 01/2012 ORNL, March 02, 2010

39 Experiment confirms: Li brings enough oxygen into the surface O 1s Surface Concentration (%) Virgin Graphite D + bombarded O-graphite D + bombarded Li-graphite D + bombarded virgin graphite Post-O irradiation Post-Li deposition 6x x10 16 D + Fluence (cm -2 ) D+ bombardment of virgin or oxidized graphite surfaces suppresses O content in time Lithiated graphite increases O-content upon D+ irradiation to 20-40% Surface Concentration (%) C 1s % F 1s % Li 1s % O 1s % N 1s % Irradiation Time (hr) C. Taylor et al, 2012 for the Department of Energy OFES LTP, Review, June 01/2012 ORNL, March 02, 2010

40 Simulation: P. Krstic et al, Nature SR, 2012 Probability per D Sputtering Yield (%) a) b) 0.0 A B C D E Matrix Composition CONCLUSIONS Retention Reflection Carbon Total Li is an excellent oxygen getter: Lithiation of C brings A LOT OF Oxygen inside C and this the main role of Li. Oxygen and Oxygen-Carbon bond D strongly: suppressing erosion & increasing D retention. Li mainly ineffective in this process. for the Department of Energy OFES LTP, Review, June 01/2012 ORNL, March 02, 2010 Nearest Neighbor Occurrence (%) C Li O H A 100% 100% C Li O H Matrix Composition B C D 80% 60% 52% 20% 20% 16% 20% 16% 16% 9% 91% 27% 5% 67% 9% 16% 3% 71% E 80% 20% 30% 70% A B C D E 9%

41 What is model teaching us? It is not Lithium that suppresses erosion of C, and increases retention of H OXYGEN plays the key role in the binding of hydrogen. Lithium is the oxygen getter If there is a SIGNIFICANT amount of oxygen on surface with lithium present in the graphite matrix, OXYGEN becomes the main player; NOT LITHIUM!!!... consistent with the XPS data!! A beautiful music of full harmony of theory and experiments for the Department of Energy OFES LTP, Review, June 01/2012 ORNL, March 02, 2010