Recent Progress in Flibe Chemistry Control, Corrosion, and Tritium Behavior Phil Sharpe Fusion Safety Program Idaho National Laboratory, USA HAPL Program Meeting ORNL March 2006

Fusion Safety Program Slide 1 Research Program Logic Purification Mobilization REDOX Corrosion Deuterium/tritium Permeation Behavior in Flibe Topics for Review

Fusion Safety Program Slide 2

Fusion Safety Program Slide 3 Flibe Purification and Analysis Hydro-fluorination approach Bubble H 2 /HF/He thru melt (530ºC) Chemical reactions BeO + 2HF = BeF 2 + H 2 O MF 2 + H 2 = M + 2HF Chemical Analysis of Flibe Components Pre- post- purification Techniques: Metals: ICP-AES, ICP-MS, dissolution C, N, O: LECO O (ppm) C (ppm) N (ppm) Fe (ppm) Ni (ppm) Cr (ppm) BeF < LiF60< Flibe Control instruments & He-HF gas cabinet Pot/heater assembly Titration cell Gas manifolds HF traps Processed Flibe

Fusion Safety Program Slide 4 Experiments on Flibe mobilization in Ar, air and humid air were completed. Obtained vapor pressures and mobilization estimates. Observed interesting behavior of the salt in different air and crucible environments Mobilization Studies

Fusion Safety Program Slide 5 Mobilization Studies, cont. Approach: Expose molten Flibe to Ar, air, moist air at ºC, quantify and characterize mobilized material Tests in argon agree with Knudsen mass spec data but are less than ORNL extrapolations and calculations Mobilized deposits were analyzed by ICP-AES. Vapor pressures were derived assuming BeF 2 and LiBeF 3 vapor species Tests in air and moist air show no significant differences from tests in argon

Fusion Safety Program Slide 6 Flibe under neutron irradiation will generate free fluorine and/or TF which can be quite corrosive Need to control fluorine potential to minimize corrosion Use Beryllium as the redox agent to tie up the free fluorine BeF 2 has the lowest free energy of formation for all metal fluorides except LiF. Thus, BeF 2 is the most stable with respect to other metal fluorides and TF as well Be is also useful for neutron multiplication in a blanket system. Flibe REDOX Control Studies- Needs

Fusion Safety Program Slide 7 Flibe REDOX Control Studies- Rationale T+T+ F-F- n+LiF Metal Wall, M Be + (surface or dissolved) M (surface) Key issue is chemical competition between T +, Be + and metal wall for the free fluorine. REDOX control is expected to tie up F - and minimize formation of TF and MF Reactor calculations by DK Sze and APEX team suggest F - and/or TF concentration would be ~ 10 wppb per pass in the reactor This level corresponds to ~ 10 Pa which is times below measurement sensitivity Evaluate REDOX behavior of Be in Flibe at different concentration levels of HF Evaluate Be solubility behavior in Flibe Develop kinetics model to guide experiments at very low HF level that are more representative of fusion blanket conditions Establish data needed for key checkpoint prior to tritium/corrosion testing

Fusion Safety Program Slide 8 Flibe REDOX Control Studies- Approach Three key reactions: Be +2HF BeF 2 + H 2 H 2 + MF 2 M + 2HF HF(g) HF(s) H 2 (s) H 2 (g) Inject HF into the Flibe and measure change in HF in the outlet gas as a signature of the REDOX potential Insert Be rod into Flibe for a specified time and then remove If Be solubility in Flibe is enough to provide REDOX control, then HF will be converted to H 2. If not, then REDOX is controlled by H 2 /HF reaction itself and we would expect to see no change in the gas Inject HF into the Flibe On line measurement of HF in the gas with titrator and mass spectrometer allows dynamic time dependent information to be obtained Can change HF level, temperature, Be exposure time and see dynamic change in system Measure HF in the gas phase as a signature of REDOX potential Be rod

Fusion Safety Program Slide 9 Simple Mass Transfer Model Be Dissolution Kinetics: Preliminary Data Both titrator and QMS used to estimate Be concentration in the salt after dunking with comparable results Even after 1 hr, the Be concentration is an order of magnitude less than that measured from dissolution test samples Need long term Be dunk in redox system to more accurately determine C sat in the model

Fusion Safety Program Slide 10 Times on the plot are Be exposure times Redox Results: HF Concentration vs time Be dunked in the salt for varying lengths of time HF concentration in the gas phase measured via QMS HF feed for these experiments was nominally 1000 ppm HF is initially reacted almost entirely by the Be As Be is depleted via reaction with HF, reaction rate slows

Fusion Safety Program Slide 11 REDOX Results: HF conversion versus time HF conversion, f, is defined by: High conversion while Be remains immersed in Flibe Reduction in conversion as Be dissolved in Flibe is consumed Shape of curve is “inverted S”

Fusion Safety Program Slide 12 Simple REDOX Kinetics Modeling reaction rate plug flow reactor empirical relationship relationship between x Be and time First attempt at a model neglected mass transfer limitations, rather assuming the kinetics are effectively reaction limited. The data best fit a reaction rate law first order in HF and Be, coupled with an unmixed reactor. Latest results appear to suggest that the reaction is in fact limited by diffusion of HF into the salt. The model is, thus, being reworked.

Fusion Safety Program Slide 13 When plotted in this dimensionless terms (f vs. x Be ), the results are remarkably consistent Conversion (f) is based on mass spec data and Be mole fraction (x Be ) is from titrator data Model predicts results very well Lower HF concentration data is currently being used to improve the model. Simple REDOX Kinetics Modeling- Results

Fusion Safety Program Slide 14 Corrosion Tests in Flibe TC He+H 2 +HF He He+H 2 +HF Pot-arrangement corrosion tests (stagnant fluid):

Fusion Safety Program Slide 15 Planned Corrosion Tests Baseline Redox test with 5-hour Be exposure. HF=1075 ppmv, H 2 /HF=11, Flow Rate=140 sccm Measure HF output with QMS and by titration Dissolve salt for H 2 release, i.e., Be metal content Test with 5-hour Be dunk, then expose ferritic steel Sample salt, ICP analyses for Fe, Cr, W and Ni Continue test beyond Be Redox control point Remove metallic impurities by electrodeposition Test another ferritic steel coupon without the Be Redox pretreatment. Post-test Analyses: fracture, clean and examine sample by various methods, e.g., SEM, AES, XPS, XRD, and RBS

Fusion Safety Program Slide 16 Exposure and Sampling Probes for Corrosion Tests Test positions for Be and FS sample Depth of exposure: 2.3 cm

Fusion Safety Program Slide 17 Consideration of Insulating Material HF reactions with ceramics oxides Alumina was selected.

Fusion Safety Program Slide 18 Redox vs Corrosion Parameters Be sample for Redox tests Diameter: 0.76 cm Depth: 1.9 cm Area: 4.99 cm 2 Diameter: 0.51 cm Depth: 2.3 cm Area: 3.87 cm 2 Be sample for corrosion tests Redox vs Corrosion HF (ppm) = 500 F.R. = 120 sccm H 2 /HF = 10 HF input rate= 4.6E-8 mol/s HF (ppm) = 1075 F.R. = 140 sccm H 2 /HF = 11 HF input rate= 1.12E-7 mol/s

Fusion Safety Program Slide 19 Analyses of Salt Samples Salt Sample: ~1.3 grams Sulfuric acid dissolutions: H 2 release, i.e., Be metal determinations Nitric acid dissolutions: ICP-AES determinations For Fe, Cr, W and Ni ~ 1 gram ~ 0.3 gram

Fusion Safety Program Slide 20 Be Determinationed from Acid Tests Be Solubility

Fusion Safety Program Slide 21 Chemical Analyses of Flibe following FS exposure Flibe Batch: 475 g (14.7 moles) Size of salt sample: 1.4 gram Ferritic steel sample: Composition: 89Fe-9Cr-2W Exposed area: 0.65 cm 2 Predicted increases of Fe in Flibe

Fusion Safety Program Slide 22 Post-test Examination of FS Samples FS sample with flibe coating Weigh and fracture sample Bottom section (INL) (re-weigh) SEM of cross-section: Flibe to salt interface Remove flibe: molten KCl:LiCl, then rinse with water, re-weigh Surface analyses at INL: SEM, XPS and AES Measure loss in thickness Top section (Japanese) (re-weigh) Baseline samples (thickness, mass) Send to Japan: XRD, RBS, XPS and Moessbauer analysis

Fusion Safety Program Slide 23 Permeation experiments: Interrelated transport processes and chemical interactions characterize the behavior of hydrogen isotopes in molten salts Integral test approach: Dual permeation probes assembly Combine experiment and modeling One-dimensional diffusion Nickel probes(0.5 mm) are Flibe resistant Diffusion in Flibe is rate-limiting 400 cc of Flibe Tests at 600 and 650ºC Transport parameters Diffusion, solubility, convection in melt Recombination at metal surfaces Liquid/gas phase transport Chemical interactions HT or HF Trapping of T at impurities

Fusion Safety Program Slide 24 Results of experiments: without/with Flibe Without FlibeWithout & with Flibe Derived permeabilities in empty pot show good agreement with Robertson’s correlation for Ni Reduction in probe-2 concentration of D 2 (due to low solubility in Flibe) Time delay for observation of permeation signal in probe 2 (due to slow diffusivity in flibe) D 2 Partial Pressure

Fusion Safety Program Slide 25 Correlation of D diffusivity and solubility in Flibe Diffusion data > D from viscosity estimate < D from capillary experiment activation E similar to F - diffusion Solubility data Derived solubilities are comparable to those reported by Field et al. for DF in Flibe Solubility Coefficient Diffusion Coefficient (m2/s)

Fusion Safety Program Slide 26 Activity 1:Installation and testing of permeation chamber in pot furnace arrangement Activity 2:TMAP modeling of permeation chamber Activity 3:Design of tritium handling and diagnostics systems Activities for the FLiBe Tritium Permeation Experiment

Fusion Safety Program Slide 27 Activity 1: Setup of Permeation Pot Permeation chamber received from Japan in October 2005; testing revealed pinhole leaks in several welds, repairs were made Chamber is designed to fit within pot furnace placed in glovebox; same system used for D 2 permeation studies Salt bath is optional if thermal gradients persist or wall leakage is substantial New batch of Flibe is being prepared; hydro-fluorination purification to proceed following completion of corrosion studies replace photo with one showing GC- for SCM

Fusion Safety Program Slide 28 Activity 2: TMAP modeling of Permeation Chamber Straightforward modeling tool will help optimize experiment layout, e.g. required sweep gas flow rates, need for use of salt bath, thickness of Flibe for appropriately timed experiments, etc. 1-D axial model with sink terms to simulate radial loss of T Builds on success of TMAP modeling with D 2 permeation experiment Suitable study for graduate student, but need to perform soon Model basis for permeation pot by Fukada et al.

Fusion Safety Program Slide 29 Activity 3:Design and testing of tritium handling and diagnostics systems Tritium provided in pressurized vessel containing D 2 /T 2 mixture Glovebox setup to contain potential leaks Localized tritium cleanup will be connected GC column for H isotope separation has been tested with tritium; works well but needs calibration Develop DF/TF generator if schedule permits ArD2D2 Flow meter Gas chromatograph or QMS or ionization chamber HF trap exhaust High temperature salt Flinak or Flibe Cap Ni T2T2 Vacuum pump Pressure gauge dip Be if Redox control is successful Conceptual layout proposed by Fukada et al.