ORNL-TM-2978.txt

 

ORNL-TM-2978

Contract No. W-7405-eng-26

METALS AND CERAMICS DIVISION

COMPATIBILITY OF FUSED SODIUM FLUOROBORATES AND BF; GAS
WITH HASTELLOY N ALLOYS

J. W. Koger and A. P. Litman

JUNE 1970

 

LEGAL NOTICE

This report was prepared as an account of Guvernment aponsored work. Neither the United
States, nor the Commission, nor any person acting on behalf of the Commission:

A. Makes any warranty or representation, expressed or implied, with reapect to the accu-
racy, completeness, or uaefulnesa of the information contained in this report, or that the use
of any information, spparatus, method, or process discloged in this repor! may oot infringe
privately owned rights; or

B. Assumes any liabilities with respect to the use of, or for damages resuiting from the
use of any information, apparatus, method, or process disciosed in this report.

Ag used in the above, ‘‘person acting on behalf of the Commisgaton” includes any em-
ployee or contractor of the Comwlission, or employee of such contractor, to the extent that
such employee or contracter of the Commission, or employee of such contractor prepares,
disseminates, or provides access to, any {nformation pursuant to his employment or contract
with the Commigsion, or his employment with such contractor.

 

OAK RIDGE NATIONAL LABORATORY

Oak Ridge, Tennessee

operated by

UNION CARBITE CORPORATION

for the

U.3. ATOMIC ENERGY COMMISSION

SEITHINIRION OF THIS DO 700 [T UNLIMITED

¥
5

\ i

 
iii
CONTENTS

Abstract .
Introduction .
Experimental Methods and Materials .
Results .o
Effect of Salt Composition and BF; Pressure (Series I)
Weight Changes .
Salt Chemistry .
Specimen Chemistry .
Metallography
Effect of Temperature (Series II) .
Interpretation of Corrosion Rates
Conclusions

Acknowledgment .

Page

Qo ~1 ~1 ~1 WM

10
1l
13
14
16
17
COMPATIBILITY WITH FUSED SODIUM FLUOROBORATES AND
BF; GAS OF HASTELLOY N ALLOYS

 

 

J. W. Koger and A. P. Litmanl

ABSTRACT

Corrosion by NaBF,-NaF mixtures (4—8 mole ¢ NaF) low in
oxygen and water was studied with 680C0-hr static tests at
605°C in standard Hastelloy N capsules with titanium-modified
Hastelloy N specimens. Boron trifluoride gas was added to
overpressures up to 400 psig. Specimens in the molten salts
and at the interface showed both weight gains and losses. The
gains decreased with increasing NaBF, content, and losses up
to 0.03 mg/em? (< 0.0l mil/year) were observed for specimens
in 4 mole ¢ NaF. Weight changes of specimens in vapor were
very small and independent of overpressure. The main reaction
was the oxidation of chromium in the alloy by iron fluoride in
the salt to form Na;CrF, and deposit iron on both specimens
and capsule. Metallographic examination showed only minor
attack and no difference in the two Hastelloy N alloys.

The rate of chromium depletion at 605°C was consistent
with the rate of solid-state diffusion of chromium to the
alloy surface. Additional tests from 538 to 760°C showed
Arrhenius behavior that confirmed this mechanism,

Fluoroborate salt mixtures were found compatible with
Hastelloy N alloys under the conditions of these experiments,
and comparison with other experiments showed that increases
in water and oxygen content increased the chromium uptake
and corrosion.

 

INTRODUCTION

The successful use of molten fluoride salts in a reactor system, as

demonstrated by the Molten Salt Reactor Experiment (MSRE) at the Qak

Ridge National Laboratory, has led to development studies for a Molten

Salt Breeder Reactor (MSBR). One of the most promising features of

molten salts has been the low corrosion rates experienced with materials

of construction, principally nickel-base alloys.

 

INow with AEC Division of Space Nuclear Systems, Washington, D.C.
Theory and experiment?® have shown that the corrosion resistance of
metals to molten fluoride salts varies directly with the "nobility" of
the metal; that is, the resistance to attack varies inversely with the
magnitude of the free energy of formation of the fluorides of metals in
the container material, Therefore, the container alloy for the fluoride
salt must have only a small percentage of constituents that are easily
oxidized by the components of the salt. Considering these facts and
utilizing information gained in corrosion testing of commercial alloys,
ORNL developed® a high-strength nickel-base alloy, Hastelloy N, containing
174, Mo, 7% Cr, and 59 Fe, for use in the MSRE. These early experiments,
as well as many recent corrosion tests* ™ ® in LiF-BeF,-ThF,, LiF-BeF,-UF,,
and LiF-BeF,-UF,~ThF,, have shown that Hastelloy N is very resistant to
attack and corrosion rates are below 0.1 mil/year.

Recently the selection of a fluoride salt that can be used as a
coolant to transfer heat from the fuel salt to a steam power conversion
system has been seriously considered. Cost and melting point considera-
tions favor the sodium fluoroborate mixture, NaBF,—8 mole ¢ NaF. However,
little information exists on corrosion by fluoroborates., Initial corro-
sion results were obtained from an impure fluorcborate salt (high in
oxygen and water) in thermal convection loop tests,’ but these results
were not considered indicative of behavior with purer salt. Since that
time, new methods of preparation have greatly increased the purity of

available salt with respect to oxygen and water.

 

¢J. H. DeVan and R. B. Evans III, Corrosion Behavior of Reactor
Materials in Fluoride Salt Mixtures, ORNL-TM-328 (Sept. 19, 1962).

*W. D. Manly, J. H. Coobs, J. H. DeVan, D. A. Douglas, H. Inouye,
P. Patriarca, T. K. Roche, and J. L. Scott, "Metallurgical Problems in
Molten Fluoride Systems," Progr. Nucl. Energy Ser. IV 2, 164-179 (1960).

 

“J. W. Koger and A. P. Litman, MSR Program Semiann. Progr. Rept.
Feb. 29, 1968, ORNL-4254, pp. 218-221.

°J. W. Koger and A. P. Litman, MSR Program Semiann. Progr. Rept.
Aug. 31, 1968, ORNL-4344, pp. 257-264.

®J. W. Koger and A. P. Litman, MSR Program Semiann. Progr. Rept.
Feb. 28, 1969, ORNL-4396, pp. R43-246.

7J. W. Koger and A. P. Litman, Compatibility of Hastelloy N and
Croloy 9M with NaBF,-NaF-KBF, (90-4-C mole ¢,) Fluoroborate Salt,
ORNL-TM-2490 (April 19€9).
The overall objective of this study was to determine the corrosion
characteristics of pure (low oxygen and water) NaBF,-NaF mixtures under
isothermal conditions. Of particular concern was the temperature-
dependent dissociation of the salt into liquid NaF and BF, gas. Thus,
we first determined the effects of varying the concentration of our salt
mixture and the BF; pressure over it in 6800-hr tests at 605°C. 1In a
later series of experiments we determined the rate of chromium uptake

by NaBF,—8 mole ¢ NaF between 538 and 760°C in 1200-hr tests.

EXPERIMENTAIL METHOD AND MATERIALS

The capsule design used in studying the effect of salt composition
and BF; overpressure (Series I) is pictured in Fig. 1. Each capsule con-
tained three specimens of titanium-modified Hastelloy N, one located in
the vapor space, one in the salt, and one at the liquid-vapor interface,
The capsule was constructed of commercial Hastelloy N and was 2 in. in
diameter X 6 in. high. The titanium-modified alloy is considered

because of its superior radiation-resistant properties.

PHOTC 76207A

AS AND FILL LINES
BAS AN DRAIN TANK

CAPSULE

 
 
     

SPECIMENS

Fig. 1, Hastelloy N Capsule Assembly for Studying Effects of Salt
Composition and BF; Overpressure.

The capsule used in studying the effect of temperature (Series II)
was basically the same. However, the test specimens were commercial
Hastelloy N in the form of 0,030-in.-thick strips to provide more surface
area., FEach capsule was 2 in. in diameter X 8.5 in. high and contained
16 specimens. The capsule 1s pictured in Fig. 2, and the test specimens
are shown in Fig. 3. Table 1 lists the compositions of the alloys used

for the capsules and specimens.
 

 

 

 

 

 

 

 

 

 

PHOTO 76436A

5 F

   

 

 

 

TANK

 

 

 

 

 

______ ‘,i:".".°fl " HASTELLOY N
- PLATES '

   

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

) 7 5*?“ A ‘

Fig. 3. Standard Hastelloy N Specimens Made to Provide Maximum
Surface Area for Study of Temperature Effects. (a) End view. (b) Side
view. ' - ’ ' : - |

 
Table 1. Composition of Hastelloy N

 

 

 

o a
sgiizs Component Composition (wt ¢)
Mo Cr e Si Mn Ti
I Capsules - 15.8 7.4 2.4 0.2 0.3 0.02
Specimens 12.0 7.3 < 0.1 <0.,01 0.14 0.5
IT Capsules and 17.0 7.2 4.3 0.45 0.5 0.02
Specimens

 

aBalance nickel.

The salt for these tests was processed by the Fluoride Frocessing
Group of the Reactor Chemisfry Division. The very pure (> 99.94) starting
materials were evacuated to about 380 torr, heated to 150°C in a vessel
lined with nickel, and then held for about 15 hr under these conditions.
After the rise in vapor pressure was observed to be not excessive (indi-
cating no volatile impurities), the salt was heated to 500°C while still
under vacuum and agitated by bubbling helium through the capsule for a
few hours. The salt was then transferred to the fill vessel and forced
into the capsules with helium pressure. Salt chemistry 1s discussed
under "Results" in this report.

A1l capsules were heated in vertical muffle furnaces (Fig. 4) and
monitored with Chromel-P vs Alumel thermocouples, which had been spot
welded to the capsule and covered with shim stock. In our first series
of tests all capsules were operated at 605°C, close to the maximum tem-
perature proposed for the coolant salt in the MSBR. Capsule 1 of this
series contained a helium overpressure of 4 psig and no added BFj.
Capsules 2, 3, and 4 contained BF; corresponding to overpressures of 50,
100, and 400 psig, respectively. Initially the added BF, dissolved in
the molten salt or combined with the NaF, but after a few minutes the
pressure in the capsule gradually increased. However, when the capsules
had been pressurized to the operating pressures, no changes in BF; con-
tent were necessary during the test. The compositions of the salts

calculated from amounts of BF; added are shown in Table Z.
 

 

Photo 76727

 

 

 

 

 

 

 

Furnace Assembly

4.

Fig.

 
Table 2. Effect of Salt Composition and BF; Overpressure
on Weight Changes of Hastelloy N

 

 

 

Salt . 5
BF, Composition Weight Change (mg/cm?)
Capsule Overpressure (mole )
(psig) _(mole %) Vapor Interface Liquid
NaBF, NaF
1 0 92 8 —-0.03 +0.5 +1.3
2 50 94 6 —0.3 +0.4 +1.2
3 100 | 95 5 —0.4 +0.2 +0.06
A 400 96 & 0.0 —0.03 —0.03

 

After 6800 hr the salt in each capsule was removed and sampled. The
capsules were opened and the specimens were removed, washed in warm dis-
tilled water, dried with ethyl alcohol, and weighed., The specimens and
capsules were examined metallographically and by x-ray fluorescence and
microprobe analysis. The salts were analyzed before and after the tests.

In our second series of experiments, the capsules were operated at
427°C (800°F), 538°C (1000°F), 649°C (1200°F), and 760°C (1400°F) with
5 psig He overpressure and calculated equilibrium BF,; pressures of 0,005,
0.07, 0.5, and 2.5 psia, respectively. After 1200 hr the salt in each
capsule waS'removed and sampled for analysis. The salt had also been

analyzed before test.
RESULTS

Effect of Salt Composition and BF, Pressure (Series I)

Weight Changes

The compositional changes in the salt effected by BF; gas additions
in our first series of tests caused a measurable difference in the weight
'changes of our test specimens. As shown in Table 2, the weight changes
were relatively small, and some specimens showed weight gains instead of
the expected losses. There was no correlation between the pressure of

BF; and the weight change of the specimens exposed to the vapor.
 

| The speciméns immersed_in salt,in capsules 1, 2, and 3 showed pro-
gressively decreasing'weight‘gains. The salt specimen -in capsule 4 showed
- a weight loss equivalent to a corrosion rate less than 0.0l mil/year.
The same weight change pattern was observed for the specimens at the
interface. Thus, the weight gain of specimens in the liquid and at the
interface increased as the amount of NaBF, in the salt decreased.
Figure 5 shows the specimens éfter test. Very little corrosion was

noted at any position on the specimens or capsules.

 

" PHOTO 76206A

VAPOR

 

 

  
 

| ~—— INTERFACE

 

 

SALT

 

Fig. 5.  Titanium-Modified Hastelloy N Specimens Aftter Exposure to
NaBF,—8 mole 4 NaF at 605°C for 6800 hr. Numbers identify capsules.

Salt Chemistry
Chemical analyses of the salt before end after test for each

 

capsule are given in Table 3. The most significant changes are an
increase in chromium concentration and a decrease in iron concentration.

No titanium concentrations are included in Table 3 because, as expected,
Table 3. Salt Analyses Before and After Test (Series I)

 

 

 

 

Concentration
Element As Capsule Capsule Capsule Capsule
Received 1 2 3 2
7 Welght Percent
Na 21.9 21.4 21.0 21.8 21.0
9.57 9.54 9.48 9.48 9.49
68.2 £9. 2 68.5 68.7 68.8
Parts Per Million
Cr 19 75 75 73 72
Ni 28 < 10 < 10 < 10 < 10
Fe 223 24 28 29 22
Mo < 10 <5 < 5 < 5 < 5
0 459 194 576 372 1420%
H,0 400 400 460 350 460

 

aQnestionable result.

no change was noted; only the test specimens contained titanium, and
they constituted only a small part of the total system. The only change
in the water and oxygen concentrations occurred in capsule 4. The high
oxygen concentration reported for this capsule would normally have pro-
duced highly oxidizing conditions® and high corrosion rates. The
reported value is believed to be in error because close examination of
the system disclosed no leaks or signs of oxidation. Apparently some
BF3; escaped from the capsules after cooling and during opening, since
the concentrations of sodium, boron, and fluorine in the salt were about
the same in the final analysis of each capsule.

The increase in chromium and the decrease in iron in the salt sug-

gest a reaction of the type

 

8. E. McCoy, J. R. Weir, Jr., R. L. Beatty, W. H. Cook, C. R. Kennedy,
A. P. Litman, R. E. Gehlbach, C. E. Sessions, and J. W. Koger, Materials
for Molten-Salt Reactors, ORNI-TM-2511 (May 19692).

 
10

Cr(s) + FeF,(d) 2 CrF,o(d) + Fe(deposited) , (1)

where (s) indicates solid solution in Hastelloy N and (d) indicates
dissolved in salt. Any water or oxygen-type impurity in the salt would
also have caused some oxidation of the container materials, but these
impurities appear to have been negligible in this series of tests.
'Because the salt in these experiments contained NaF, the chromium fluo-
ride in the corrosion product was found as NasCrFg. This material has
also been identified in other corrosion experim.ents9 and its crystal
structure has been determined by x-ray diffraction analysis.lo Also,
the iron fluoride in these experiments may have existed as NasFeF¢, but

this compound was not identified.

Specimen Chemistry

 

The proposed corrosion reaction was verified further by x~-ray
fluorescence analysis of the specimens. Table 4 gives the composition
of the near-surface region of a specimen from capsule 4. The composi-
tion was determined by comparing x-ray intensities with those of a
specimen not exposed to salt or vapor. For any given region in the
capsule, the qualitative results from all the specimens were quite
similar. These results show a depletion of chromium and enrichment of
iron in the near-surface region. All specimens also showed a loss of
titanium. Microprobe analysis showed essentially the same surface
changes. A chromium gradient was seen but was so small that we could
not distinguish between edge effects and an actual gradient.

1

According to Cantor, ! NaBF,—8 mole % NaF may also oxidize chro-

mium by the reactions

 

°J. W. Koger and A. P. Litman, Compatibility of Hastelloy N and
Croloy 9M with NaBF,-NaF-KBF, (90-4-6 mole %) Fluoroborate Salt,
ORNL-TM-2490 (April 1969).

105, Brunton, "The Crystal Structure of NasCrF¢," Mater. Res. Bull. 4,
621—626 (1969). =

11s. Cantor, MSR Program Semiann. Progr. Rept. Aug. 31, 1968,
ORNI~4344, p. 160.

 
11

Table 4. Composition of Near-Surface Regions of Test Specimens
as Determined by X-Ray Fluorescent Analysis

 

 

 

 

Exposure Concentration (wt %)
Conditions Mo Ni Fe Cr T4
Unexposed 12.0 79.9 0.1 7.5 0.5
Vapor 12.3 79.2 0.13 7.1 0.19
Interface 13.7 69. 6 13.4 3.0 0.2
Liquid 18.0 22.7 56. 4 2.7 0.16
(1 + x) Cr(s) + NaBF,(d) & NaF{(d) + CrFs(d) + CrXB(s) , (2)
3NaF(d) + CrFs(d) 2 NasCrFg(s) . (3)

However, no borides were identified either on our specimens or in the
salt, and the experimental evidence that iron replaced the chromium
indicates that reaction (1) rather than (2) was the predominant mode of
chromium oxidation. If essentially no iron fluoride compound were
present in the melt, reaction (2) could well control the corrosion

process.

Metallography

Figure 6 shows the microstructure of the specimen from the vapor
region of capsule 1. There was some indication of a very thin, discon-
tinuous deposit at the surface in the as-polished condition. After
etching, this deposit was no longer visible; Figure 7 shows the speci-
men that was exposed to salt in capsule 1. A deposit is visible on the
as-polished specimen. Etching again removed the deposit but in this
case produced considerable attack at the exposed surface. Metallographic
observation of the capsule wall showed similar behavior. This deposit is
apparently iron rich, and the area revealed by the etchant is chromium-
depleted Hastelloy N. Thus, no difference in the behavior of titanium-
modified and standard Hastelloy N was noted in this test. The interface
between the salt and gas could be seen on the capsule, but no attack was

noted there.
 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

12

 

 

 

 

 

 

 

Fig. 6. Titanium-Modified Hastelloy N from Capsule 1 Exposed to
BF3 Vapor for 6800 hr at 605°C. Weight change, +0.03 mg/cm?. 500x.
(a) As polished. (b) Etched with glyceria regia.

 

 

 

 

 

 

 

-
i

TN

i e A

Fig. 7. Titanium-Modified Hastelloy N from Capsule 1 Exposed to
NeBF;~8 mole % NaF for 6800 hr at 605°C. Weight change, +1.3 mg/cm?.
500x. (a) As polished. (b) Etched with glyceria regia.

 

 

 
 

 

oh

 

X

wy

 

 

 

 

 

13

- Effeet‘ef Tehperature (Series II)

 

Our second serles of tests was des1gned to determine the effects of

temperature on chromium.mass transfer from the Hastelloy N to the salt.

The salt was analyzed(before and after test to determine 1nmmr1ty pickup
The NaBF;—8 mole %‘N&F contained ‘approximately 400 ppm each of oxygen and
water before and after test. rThe_chromium.concentration_in the salt

. Increased while the iron decreased,'as'seen'in Table 5. No changes in

the concentration of‘nickel and molyhdenumuwere noted. Again, the

'-results suggest chromium oxidation by reduction of an iron fluoride

compound. Metallography of: Hastelloy N from.this test showed, after

\etching, no attack of the specrmen (Fig.-S)

Table 5. Analy31s of Salt for Impurities Before and After Test
o (Series II) -

 

--Asz'erage-Concentrationa (ppm)

 

 

Trpurity - As . Capsule  Capsule  Capsule  Capsule
Received ~~ 1 2 3 4
Cr 15 - 45 87 225 . 575
Fe 200 - 190 180 150 40
-0 460 430 - 450 . 412 550
“Hp0 - 320 - _ ;7400',' ' 530 480 . 490

 

SNickel and.molybdenumuwere not detected (( 10 ppm) in any
sam@les.

 

Fig. 8. Hastelloy N Exposed to NaBF4—8 mole % Na.F for 1200 hr at

760°C. Etched.with glyceria regis, 500x.
14

INTERPRETATION OF CORROSION RATES

If the chromium surface concentration remainsg constant (only true
if the salt contains an excess of iron) at any given point in the cap-
sule, the amount of chromium leaving a capsule in Series I is given by

the equationl?

M = 20(Co — C) V/Db/7 ()

where
MW = integral flux of Cr leaving metal, g/cm?
p = density of metal, g/cm’
Co = initial weight fraction of Cr
C_ = surface weight fraction of Cr
D = diffusion coefficient of Cr in alloy, cm?®/sec
t = exposure time, sec.
Using the following informetion
Salt in capsule = 282 g
Cr increase in salt = 55 ppm (by weight)
Capsule surface area = 142 cm?

to obtain /AW and then using

Co = 7.5%
¢, =0 (consistent with excess of Fe)
o = 8.8 g/em?
t = 6800 hr,
we obtain
Dagleg = 3-3 X 10-1° cm?/sec at 605°C.

This compares favorably with 3 x 1071° cm?/sec at 605°C extrapolated

from the range 700 to 850°C from datal?® of DeVan and Evans for chromium

 

12L. 8. Derken and R. W. Gurry, Physical Chemistry of Metals,
McGraw-Hill, New York, 1953.

1°§. R. Grimes, G. M. Watson, J. H. DeVan, and R. B. Evans, "Radio-
Tracer Techniques in the Study of Corrosion by Molten Fluorides,"
Pp. 559574 in Conference on the Use of Radioisotopes in the Physical
Sciences and Industry, September 6-17, 1960, Proceedings, Vol. III,
International Atomic Energy Agency, Vienna, 1962.

 

 

 
15

diffusion in Hastelloy N. From this comparison, we conclude that the
controlling rate mechanism in this experiment was the solid-state dif-
fusion of chromium in the alloy and that the oxidation potential of the
salt was established by the presence of iron fluoride.

Figure 9 shows the variation of chromium content with test temper-
ature for the experiments of Series II. Included for comparison are
chromium concentrations measured in loop tests after 1200 hr at various

H,0 impurity levels.'*”17 An Arrhenius-type relationship appears to

 

147, W. Koger and A. P. ILitman, MSR Program Semiann. Progr. Rept.

Feb. 29, 1968, ORNL-~4254, pp. 218-221.

157, W. Koger and A. P. Litman, MSR Program Semiann. Progr. Rept.
Aug. 31, 1968, ORNL-4344, pp. 257—264.

167, W. Koger and A. P. Litman, MSR Program Semiann. Progr. Rept.

Feb. 28, 1969, ORNL-4396, pp. 243—246.

175, W. Koger and A. P. Litman, Compatibility of Hastelloy N and
Croloy 9M with NaBF,-NaF-KBF, (90-4-6 mole %) Fluoroborate Salt,
ORNL-TM-2490 (April 1969).

ORNL—-DWG 69— 12238R
TEMPERATURE. (°C)

 

 

 

 

 

 

 

 

 

 

103 800 700 650 600 550 500 450 400
TN b ! '
I - ,‘4\@=22.0 kcal/mole |7 |
\. A— - |-
5 N - . ]
. N ,,_::.‘___‘(2000 ppm leo) ]
. \‘\ (4000 ppm HZO)

 

2 b TN

NS

<‘O\O ppm HZ0)
1 |

 

10

 

 

— :fiifi; : ;f";_-w

 

 

 

 

 

Cr CONCENTRATION IN FLUOROBORATE ({(ppm)

 

 

 

5 e T - —
- ] 1 ——— e &

,| ' STATIC CAPSULE TESTS (1200 hr)

s THERMAL CONVECTION LOOPS ]

----- CALCULATION BASED ON Cr DIFFUSION

: IN HASTELLOY N

10 ' : ' ‘ ‘

0.90 10 14 1.2 13 14 15

1000/r (o)

Fig. 9. Temperature Dependence of Chromium Concentration in Sodium
Fluoroborate Salt.
16

hold between 538 and 760°C. This would be expected if the rate of
chromium buildup in the salt were controlled by solid-state diffusion
of chromium to the capsule wall. Using diffusion data for chromium in
Hastelloy N obtained by DeVan and Evans®® and assuming the chromium
surface concentration to have been reduced to zero by the salt, we
predict a chromium buildup shown by the dashed line in Fig. 9. The
agreement between the slopes (activation energy) of the predicted and
experimental curves, although not exact, gives credence to the assump-
tion that chromium mass transfer to the salt is limited by solid-state

diffusion of chromium to the Hastelloy N surface.

CONCLUSIONS

1. Titanium-modified Hastelloy N specimens were exposed for
6800 hr to isothermal NaBF,-NaF salt mixtures containing only a few
hundred parts per million oxygen and water and varying in composition
from 4 to 8 mole % NaF. Specimens in the vapor region lost weight in
all but the 96 mole ¢ NaBF, mixture. Specimens fully immersed in the
salt gained weight in all but the 96 mole % NaBFs mixture. The weight
gains decreased as the NaBF, content of the salt increased.

2. No differences in the corrosion of specimens in the vapor
region were noted as a function of BF; overpressure to 400 psig.

3. In the main corrosion reaction, chromium from the alloy was
oxidized by iron fluoride impurities initially in the salt. The iron
that was reduced deposited on the specimens and capsule.

4. Metallographic observation of both standard Hastelloy N capsules
and titanium-modified Hastelloy N specimens showed very little attack,
although a thin discontinuous deposit rich in iron was observed at the
surface of specimens exposed to vapor.

5. A diffusion coefficient for chromium in Hastelloy N calculated
from the rate of chromium removal from standard Hastelloy N by
NaBF,~8 mole % NaF held between 538 and 760°C agreed well with an
extrapolation of published results. Thus, solid-state diffusion of
chromium appeared to be the rate-controlling mechanism for corrosion

in these tests.
"

17

6. Under the condition of these experiments, the fluoroborate
salt mixtures were compatible with the Hastelloy N alloys. By virtue
of other experiments we conclude that increases in water as impurity

increase the chromium uptake and the corrosion.

ACKNOW LEDGMENT

We wish to acknowledge E. J. Lawrence for his expert supervision
of the fabrication, operation, and disassembly of the test capsules and
F. D. Harvey for his handling of the test specimens. We are also
indebted to H. E. McCoy, Jr., J. H. DeVan, and C. E. Sessions for con-
structive review of the manuscript.

Special thanks are extended to H. R. Gaddis and Helen Mateer of
the General Metallography Group, Harris Dunn and others in the Analytical
Chemistry Division, Graphic Arts, and the Metals and Ceramics Division

Reports Office for invaluable assistance.
1-3.
4—5,

6—15.
16.
17.
18.
19.
20.
21.
22.
23.
24,
25.
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
36.
37.
38.
39.
40.
41.
42.
43,
4te,
45,
46.

9394,
95.

96.

97.

19

INTERNAL DISTRIBUTION

Central Research Library 47,
ORNL — Y-12 Technical Library 48.

Document Reference Section 49,
Laboratory Records Department 50-52.
lLaboratory Records, ORNL RC 53.
ORNL Patent Office 54,
MSRP Director's Office (Y-12) 55.
G. M. Adamson, Jr. 56—65.
J. L. Anderson 66.
R. F. Apple 6'7.
C. F. Baes 68.
S. E. Beall 69.
E. S. Bettis 70.
F. F. Blankenship 71.
E. G. Bohlmann 72.
R. B. Briggs 73.
S. Cantor T
0. B. Cavin 75.
Nancy C. Cole , 76.
W. H. Cook 77.
J. L. Crowley 78.
F. L. Culler 79.
J. H. DeVan 80.
J. R. DiStefano 81.
S. J. Ditto ' 82.
W. P. Eatherly 83.
J. I. Federer 84.
D. E. Ferguson g5.
L. M. Ferris g6.
J. H Frye, Jr. g7.
R. E. Gehlbach 88.
L. 0. Gilpatrick 89.
W. R. Grimes 90.
A. G. Grindell 1.
F. D. Harvey 92.

EXTERNAL DISTRIBUTICN

mEUYrPQEIORESUE RS YR

= G2

ORNL-TM-2978

Harms
Haubenreich
Helms

Hill
Huntley
nouye

Kasten
Koger
Lawrence
Lindauer
Tundin
MacPherson
Martin
McCoy, Jr.
McHargue
Meyer

. Moulton
Patriarca

M. Perry

L EwE RO

RAHEIE Yo RS

L

E. Sessions

H. Shaffer

M. Slaughter
N. Smith

A. Sundberg
R. Tallackson
E. Thoma

B. Trauger

M. Watson

R. Weir, dJr.
C. White

e Young

L. Youngblood

HHRPREE R R HO

al

D. F. Cope, RDT, SSR, AEC, Oak Ridge National lLaboratory
A. R. DeGrazia, Division of Reactor Development and Technology,

AEC, Washington

David Elias, Division of Reactor Development and Technology,

AFC, Washington

J. E. Fox, Division of Reactor Development and Technology, AEC,

Washington
20

98, Norton Haberman, Division of Reactor Development and
Technology, AEC, Washington
99. Kermit Laughon, RDT, OSR, AEC, Oak Ridge Operations
100-109. A. P. Litman, Division of Space Nuclear Systems, AEC,
Washington
110-111. T. W. McIntosh, Division of Reactor Development and
Technology, AEC, Washington
112. D. R. Riley, Division of Reactor Development and Technology,
AEC, Washington
113. M. Shaw, Division of Reactor Development and Technology,
AEC, Washington
114. J. M. Simmons, Division of Reactor Development and Technology,
AFEC, Washington
115, Iaboratory and University Division, AEC, Oak Ridge Operations
116-130. Dpivision of Technical Information Extension