Recent Progress in Flibe Chemistry Control, Corrosion, and Tritium Behavior

Recent Progress inFlibe Chemistry Control, Corrosion, and Tritium

Behavior

Phil SharpeFusion Safety Program

Idaho National Laboratory, USA

HAPL Program Meeting

ORNL

21-22 March 2006

Fusion Safety Program

Slide 2

• Research Program Logic

• Purification

• Mobilization

• REDOX

• Corrosion

• Deuterium/tritium PermeationBehavior in Flibe

Topics for Review


Slide 3

Integrated Program Logic for Flibe Activities

FlibePurification

REDOXTritium

Experiments

Corrosion Tests

Interim Set Of Impurities

Is REDOX working?

Reducing TF?Able to Establish

Viable TritiumControl

Strategy? Is Corrosion

Acceptable?

Successful REDOXTritium Control and

Acceptable Impurities

Need To reduce Impurities

yes

yesno

yes

yes

Iteration loop is not in program because

of budget limitation

Preparation of Starting Materials


Slide 4

Flibe Purification and AnalysisHydro-fluorination approach

• Bubble H2/HF/He thru melt (530ºC)

• Chemical reactions

BeO + 2HF = BeF2 + H2O

MF2 + H2 = M + 2HF

Chemical Analysis of Flibe

• Components

• Pre- post- purification

• Techniques:Metals: ICP-AES, ICP-MS, dissolutionC, N, O: LECO

O(ppm)

C(ppm)

N(ppm)

Fe(ppm)

Ni(ppm)

Cr(ppm)

BeF2 5700 <20 58 295 20 18

LiF 60 <20 78 100 30 4

Flibe 560 10 32 260 15 16

Control instruments & He-HF gas cabinet

Pot/heater assemblyTitration cellGas manifoldsHF traps

ProcessedFlibe


Slide 5

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


Slide 6

Mobilization Studies, cont.

1.E-05

1.E-04

1.E-03

1.E-02

1.E-01

1.E+00

1.E+01

1.E+02

1.E+03

1.E+04

1.E+05

400 600 800 1000 1200 1400

Temperature (°C)

Pressure BeF

2 (Pa)

ORNL data [1]ORNL extrapolationOlander, et al. [2]INEEL argon dataINEEL moist airINEEL air data

Differential manometer

1.E-05

1.E-04

1.E-03

1.E-02

1.E-01

1.E+00

1.E+01

1.E+02

1.E+03

1.E+04

1.E+05

400 600 800 1000 1200 1400

Temperature (°C)

Vapor Pressure (Pa)

ORNL data [1]ORNL extrapolationOlander, et al. [2]Buchler & Stauffer [4]INEEL Total pressureINEEL-BeF2INEEL LiBeF3

Differential manometer

Mass Spectrometry

• Approach: Expose molten Flibe to Ar, air, moist air at 500-800ºC, 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 BeF2 and LiBeF3 vapor species

• Tests in air and moist air show no significant differences from tests in argon


Slide 7

• 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

• BeF2 has the lowest free energy of formation for all metal fluorides except LiF. Thus, BeF2 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.

Fluoride Potential

-1200

-1000

-800

-600

-400

-200

0

450 550 650 750

Temperature (°C)

-RTlnp

F2

(kJ/mole)

CeF3/CeF4

NiF2

H2/HF=10

FeF2

H2/HF =20/lowpressure H2

CrF2

Si2F6

MnF2

AlF3

BeF2

LiF

Flibe REDOX Control Studies- Needs


Slide 8

Flibe REDOX Control Studies- Rationale

T+

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 10-100 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


Slide 9

Flibe REDOX Control Studies- Approach• Three key reactions:

Be +2HF <--> BeF2+ H2

H2 + MF2 <--> M + 2HFHF(g) <--> HF(s)H2(s) <--> H2(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 H2. If not, then REDOX is controlled by H2/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


Slide 10

€

VdC

dt= h C sat −C(t)( ),C(0) = 0

Wt% Be = Wt% Be( )sat

1− e−ht /V( )

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 Csat in the model

0

0.00005

0.0001

0.00015

0.0002

0.00025

0.0003

0.00035

0 10 20 30 40 50 60 70

time (min)

maximum X(Be) from titrator data

from QMS data

Series3

Linear (Series3)


Slide 11

Times on the plot are Be exposure times

0

200

400

600

800

1000

1200

1400

0 10 20 30 40 50 60 70

time (hr)

HF concentration (ppm)

REDOX-4

REDOX-5

REDOX-6

REDOX-7

REDOX-10

10 min

20 min

30 min 60 min

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


Slide 12

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”

€

f =1−yHFout (t)

yHFin

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

0 10 20 30 40 50 60 70

time (hr)

HF conversion

redox-6 (60 min)

redox-6 fit

redox-7 (30 min)

redox-7 fit

redox-8 (45 min)

redox-8 fit

redox-10 (20 min)

redox-10 fit

redox-11 (45 min)

redox-11 fit

redox-9 (1800 ppm, 30min)redox-9 fit


Slide 13

Simple REDOX Kinetics Modelingℜ=kyHFxBereaction rate plug flowreactor

f =1−e−αxBe

empirical relationship

relationship between xBe 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.


Slide 14

• When plotted in this dimensionless terms (f vs. xBe), the results are remarkably consistent

• Conversion (f) is based on mass spec data and Be mole fraction (xBe) is from titrator data

• Model predicts results very well

• Lower HF concentration data is currently being used to improve the model.

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

0 0.00005 0.0001 0.00015 0.0002

X(Be)

HF conversion

redox-6 data

redox-7 data

redox-8 data

redox-9 data

redox-10 data

redox-11 data

1st order, PFR fit

Simple REDOX Kinetics Modeling- Results


Slide 15

Corrosion Tests in Flibe

TC

He+H2+HF

HeHe

He

He

He+H2+HF

Pot-arrangement corrosion tests (stagnant fluid):


Slide 16

Planned Corrosion Tests

• Baseline Redox test with 5-hour Be exposure.HF=1075 ppmv, H2/HF=11, Flow Rate=140 sccmMeasure HF output with QMS and by titrationDissolve salt for H2 release, i.e., Be metal content

• Test with 5-hour Be dunk, then expose ferritic steelSample salt, ICP analyses for Fe, Cr, W and NiContinue 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


Slide 17

Exposure and Sampling Probesfor Corrosion Tests

Test positions for Be and FS sampleDepth of exposure: 2.3 cm


Slide 18

Consideration of Insulating Material

-300

-250

-200

-150

-100

-50

0

50

100

150

200

0 200 400 600 800 1000 1200

Temperature (C)

?G, kJ/mole F

2

Al2O3ZrO2MgOCaO

1/2 ZrO2 + 2 HF(g) = 1/2 ZrF 4 + H2O(g)

1/3 Al 2O3 + 2 HF(g) = 2/3 AlF 3 + H2O(g)

MgO + 2 HF(g) = MgF 2 + H2O(g)

CaO + 2 HF(g) = CaF 2 + H2O(g)

HF reactions with ceramics oxides

Alumina was selected.


Slide 19

Redox vs Corrosion Parameters

Be samplefor Redox tests

Diameter: 0.76 cmDepth: 1.9 cmArea: 4.99 cm2

Diameter: 0.51 cmDepth: 2.3 cmArea: 3.87 cm2

Be samplefor corrosion tests

Redox vs Corrosion

HF (ppm) = 500F.R. = 120 sccm

H2/HF = 10HF input rate=4.6E-8 mol/s

HF (ppm) = 1075F.R. = 140 sccm

H2/HF = 11HF input rate=1.12E-7 mol/s


Slide 20

Analyses of Salt Samples

Salt Sample: ~1.3 grams

Sulfuric acid dissolutions:H2 release, i.e.,

Be metal determinations

Nitric acid dissolutions:ICP-AES determinations

For Fe, Cr, W and Ni

~ 1 gram ~ 0.3 gram


Slide 21

Be Determinationed from Acid Tests

Acid Dissolutions of Flibe

0.001

0.010

0.100

1.000

10.000

0.00 0.20 0.40 0.60 0.80 1.00 1.20

Sample size (g)

Be metal in salt (mol%)

Detection liitBe6 saltBE10CenterBE10EdgeRedox test

(Redox 1-h test)

0.000

0.050

0.100

0.150

0.200

0.250

0 2 4 6 8 10 12

Be dunk time (h)

Be in salt (mol%)

Be Solubility


Slide 22

Chemical Analyses of Flibe following FS exposure

0

200

400

600

800

1000

1200

0 20 40 60 80 100 120

time (h)

Iron content (ppm)

REDOX control active

with REDOX control

w/o REDOX control

passivation or saturation

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 cm2

Increase in salt: ppm-(g)Assuming uniformsample loss: (um) Fe Cr W

5 40-(0.019) 4 (0.002) 1 (0.0004)10 80 (0.037) 8 (0.004) 2 (0.0008)50 400 (0.185) 40 (0.019) 9 (0.004)

100 800 (0.370) 80 (0.037) 18 (0.008)500 4000 (1.850) 400 (0.187) 88 (0.042)1000 8000 (3.700) 800 (0.374) 175 (0.083)

ICP-AES det. limits (ppm) 10 10 10

Predicted increases of Fe in Flibe


Slide 23

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


Slide 24

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


Slide 25

Results of experiments: without/with Flibe

1013

2

3456

1014

2

3456

1015

1.31.21.11.00.9

1000/T [K-1]

800 700 600 500

Robertson's correlation

Ni, t=0.5mmD2 introduction

P=1.23x1017exp(-50.5/RgT)

one-probe exp two-probe exp

0 100 200 300 400 500 600 700 800

Elapsed time (min)-reference to PB1 pressurization

0

100

200

300

400

500

600

700

800

900

1000

D2

Partial pressure (Pa)-Exp A (no Flibe)

0

10

20

30

40

50

60

70

80

90

100

D2

Partial pressure (Pa)-Exp B (with Flibe)

Evacuate PB1

Exp B

Evacuate PB1

Exp A

Above Flibe(D2)

Exp B

Probe2(D2)

Exp B

Probe2(D2)

Exp A

Test temperature: 600oCExp A: no Flibe

Probe1 D2 pressure:

9.4x104 Pa

Probe2 Ar flow: 25 sccm

Pot volume Ar flow:

ET< 150 min: 0 sccm

ET> 150 min: 100 sccm

Exp B: with Flibe

Probe1 D2 pressure:

9.1x104 Pa

Probe2 Ar flow: 25 sccm

Ar flow above Flibe: 25 sccm

Without Flibe Without & with Flibe

Derived permeabilities in empty pot show good agreement with Robertson’s correlation for Ni

Reduction in probe-2 concentration of D2

(due to low solubility in Flibe)Time delay for observation of permeation signal in probe 2 (due to slow diffusivity in flibe)

D2 P

artia

l Pre

ssur

e

D2 P

artia

l Pre

ssur

e


Slide 26

Correlation of D diffusivity and solubility in Flibe

0.80 1.00 1.20 1.40 1.60 1.80

Inverse temperature (1000/K)

10 -14

10 -13

10 -12

10 -11

10 -10

10 -9

10 -8

10 -7

10 -6

Diffusion coefficient (m

2/s)

400500600700800Temperature ( oC)

Ohmichi et al. (F self diffusion)

Moriyama et al. (T in solid Flibe)

Stokes-Einstein

JUPITER-II (this work)

from D2 buildup

in Probe2

Oishi et al. (T)

Ed = 128 kJ/mol

En = 31 kJ/mol

Ed = 111 kJ/mol

Ed = 44 kJ/mol

0.90 1.00 1.10 1.20 1.30 1.40

Inverse temperature (1000/K)

10 -8

10 -7

10 -6

10 -5

10 -4

10 -3

Solubility coefficient (mol/m

3-Pa) 500600700800

Temperature ( oC)

JUPITER-II (this work)

from D2 buildup

in probe 2

Malinauskas et al. (H2, D2)

Field et al. (DF)

Diffusion data• > D from viscosity estimate• < D from capillary experiment• activation E similar to F- diffusion

Solubility dataDerived solubilities are comparableto those reported by Field et al. for DF in Flibe

Sol

ubil i

ty C

o ef fi

c ie n

t

Diff

usio

n C

oeffi

cien

t (m

2/s)


Slide 27

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 theFLiBe Tritium Permeation Experiment


Slide 28

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 D2 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

Ni10060.32.82215010Ni30956.3512.76.359.52NiNiss3162409.52lid ss316114160306.35φ6.35 φ6.35 φ316ss12.720446.352411010812

replace photo with one showing GC- for SCM


Slide 29

D2,, T2 orAr gas in Flibe in ExhaustAr gas in Ar gas out

Flibe stagnant

Ni

Electric furnace

High temperaturesalt

s.s.316

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 D2 permeation experiment

• Suitable study for graduate student, but need to perform soon

Model basis for permeation pot by Fukada et al.


Slide 30

Activity 3: Design and testing of tritium handling anddiagnostics systems

• Tritium provided in pressurized vessel containing D2/T2 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

Ar D2

Flow meter

Flow meter

Flow meter

Gas chromatograph

or QMS

or ionization chamber

HF trap

exhaust

High temperature salt

Flinak or Flibe

Cap

Ni

T2

Vacuum pump

Pressure gauge

dip Be if Redox control is successful

Conceptual layout proposed by Fukada et al.

Recent Progress in Flibe Chemistry Control, Corrosion, and Tritium Behavior

Documents