Determination of permeability of oxygen through molten sodium sulfate using an oxygen concentration cell with solid electrolyte

Solid State Ionics 3/4 (1981) 627-630 Nor th-Ho lland Publishing Company

D E T E R M I N A T I O N OF P E R M E A B I L I T Y OF O X Y G E N T H R O U G H M O L T E N SODIUM SULFATE USING AN O X Y G E N C O N C E N T R A T I O N C E L L W I T H S O L I D E L E C T R O L Y T E K. N A G A T A , M. S U S A and K.S. G O T O Department of Metallurgical Engineering, Tokyo Institute of Technology, Tokyo, Japan The permeability of oxygen through pure molten Na2SO4 with or without added NaCI, Fe203, NiO or Cr203 was measured in steady state by an oxygen concentration cell with ZrO2.CaO as solid electrolyte. The permeability of oxygen in pure Na2SO4 was 1.1 x 10-8 mol cm -~ s i at 1000°C and was increased by NaCI or metallic ions.

I. Introduction Nickel-based superalloys are rapidly oxidized at high t e m p e r a t u r e when their surfaces are covered by molten sodium sulfate [1]. The cause is considered to be the dissolution of the protective films into the sodium sulfate, followed by oxidation of the superalloys by oxygen permeating through the sulfate which may contain NaCI or some metallic ions such as nickel, iron, chromium, etc.). This p h e n o m e n o n is called "accelerated oxidation" or "hot corrosion". Some mechanisms for the oxygen transport in the sulfate have been proposed such as the diffusion of neutral oxygen molecules [2, 3], the counterdiffusion of SO42- and MO~- ions [4, 5] and diffusion of O 2-, M 2+ and M 3+ ions [3, 5] according to the conditions of hot corrosion. The purpose of the present work was to measure the permeability of oxygen through molten sodium sulfate containing NaCI, Fe203, NiO o r C r 2 0 3 and to clarify the mechanism of the permeation.

2. Experimental Fig. 1 shows the schematic diagram of an apparatus for measuring the permeability of oxygen through molten sodium sulfate. The molten sulfate was suspended by its surface tension in a group of fine alumina capillaries at the end of an alumina shield tube in order to

avoid convection. The inside diameter of the shield tube was 5.0 m m in which 12 capillaries (inside and outside diameter and length of 0.50.8, 1.0-1.2 and 5.0 mm, respectively) were wedged and further supported by a fine platinum net. The t e m p e r a t u r e of the sulfate was measured by a P t - P t 1 3 % R h thermocouple in contact with its surface. Oxygen gas was passed into the shield tube and deoxidized argon gas was passed to the outside of the tube. Oxygen p e r m e a t e d through the molten sulfate from the inside to the outside and was then carried by argon gas to a chamber containing an oxygen concentration cell in a separate furnace. The oxygen concentration cell consisted of a closedend tube of calcia-stabilized zirconia as solid electrolyte and platinum wires as electrodes.

02

;?

Zv

i

"l

--

F.........

f

:1

: 11 J i l

' ":-PureA Fig. i. E x p e r i m e n t a l apparatus.

0167-2738/81/0000-0000/$02.50 O North-Holland Publishing C o m p a n y

628

K. Nagata et al. / Permeability of oxygen through molten sodium sulfate

1100 t -6.5 I

Pure oxygen gas was passed to the inside of the electrolyte tube which acted as reference electrode. The temperature of the cell was kept constant at 800°C. The permeability of oxygen in the steady state can be calculated by the equation: P = 4.06 x 10-5 V(L/A)p~ exp(46420E/T),

1050 I

1000 9_50 901081~;41°c) 1 o pure NG2S04 ~ 2°/o NGCl- No2S04 n NiO - No2S04

co L I\

X~r~ '~,°\

(1)

where P is the permeability of oxygen (mol cm -1 s-Z), V is the flow rate of argon gas ( c m 3 s - I ) under 25°C and 1 atm and L and A are the effective length (cm) and area (cm 2) of the molten sulfate in the capillaries, respectively. P~f2 is the partial pressure of oxygen of 1 atm at the reference electrode, E is the electromotive force (volt) and T is the absolute temperature of the oxygen concentration cell. The electromotive force changed with the flow rate of argon gas, but the permeability of oxygen calculated from eq. (1) was independent of the flow rate. It was also independent on the flow rate of oxygen inside the shield tube. Thus, the flow rates of the argon and oxygen gas were fixed at 500 c m 3 min -] and 50 c m 3 rain -], respectively. After the outside and inside of the shield tube had been flushed by deoxidized argon gas, the argon gas inside the shield tube was switched to oxygen gas. The partial pressure of oxygen in the carrier gas gradually increased and reached a constant value. The constant electromotive force of the cell then gave the premeability of oxygen in the steady state according to eq. (1). Its reversibility was confirmed by changing the temperature. Pure sodium sulfate powder and its mixture with NaC1, Fe203, NiO or Cr203 were melted in alumina crucibles at 950°C in air and sucked up into the capillaries in the shield tube. After the permeability measurements, the sodium sulfates were dissolved into hot water in order to determine the concentration of each component by chemical analysis.

N

E

\o o o \ O \

o\

0: -7.5

n

o~\

o o v.~

o

-8.0

~o

& \~, "u

o

rff

~ ~ ~ ~ ~ )

-8.5 I

7.5

I

8.0 1 /TxlO" (K-')

_

I

8.5

Fig. 2. R e l a t i o n b e t w e e n the l o g a r i t h m of o x y g e n perm e a b i l i t y and the r e c i p r o c a l of a b s o l u t e t e m p e r a t u r e for pure m o l t e n Na2SO4, Na2SO4-2% NaC1 a n d Na2SO46.7 p p m NiO.

reciprocal of absolute temperature for various systems of the pure molten sodium sulfate with or without NaC1 or metallic ions. The permeabilities were expressed as a function of temperature by least-square fits and may be given as follows: for pure molten N a 2 S O 4 (970-1100°C), P = (9.893 --- 21.35) x 107 x exp[-(93.27 _+6.56)(kcal

mol-l)/RT];

(2)

for molten Na2SO4-2% NaCI (920-1070°C), P = (3.099 -+ 4.340) × 105 x e x p [ - (75.46 --+3.51)(kcal

mol-1)/RT];

3. Results

for molten NazSO4-30.5 ppm Fe203 (8901060°C),

Figs. 2 and 3 show the relation between the logarithm of the permeability of oxygen and the

P = (181.6 --- 95.9) × exp[-(54.52 - 1.32)(kcal

mol-1)/RT] ;

(3)

(4)

K. Nagata et al. / Permeability of oxygen through molten sodium sulfate

.90

~oso,

lOOO

9so

-6.5

9oo, ~.,

°C)

o Fe~C)3- Na2504 z, Cr203-Na2SO,

x'X~o

C~-7.S

Q.

-8.0

-8.S 7).5

the addition of F e 2 0 3 , NiO o r C r 2 0 3 to the sulfate increased it by a factor of about eight, seven or four, respectively, at 1000°C.

4. Discussion

\

-7.0

629

810

815

I I T x I O 4 (K-') Fig. 3. Relation between the logarithm of oxygen permeability and the reciprocal of absolute temperature for Na2SO4-30.5 ppm FeeO3 and Na:SO4-87.5 ppm Cr203.

for molten Na2SO4-6.7 ppm NiO (920-1050°C), P = (831.4 ± 395.8) × exp[-(59.02 ± 1.20)(kcal

mol-1)/RT ;

(5)

for molten Na2SO4-87.5 ppm Cr203 (920-1060°C) P = (258.1 ± 293.9) × exp[-(57.30 ± 2.86)(kcal mol

1)/RT].

(6)

After the permeability measurements these molten sulfates were saturated by ~470 ppm alumina. Then, the alumina gradually dissolved into the molten sulfates in the capillaries which became saturated after 3 h. However, the permeability of oxygen was scarcely influenced by the dissolved alumina content in the sulfates. The permeability of oxygen in pure molten N a 2 S O 4 w a s 1.1 × 10-8 mol c m -1 S 1 at 1000°C. The addition of NaC1 to the molten Na2SO4 increased the permeability by a factor of four, and

Molten sodium sulfate is not particularly viscous, but it is reported [6] that convection can be neglected if the Rayleigh number is less than 1000. In the present case, convection may appear when the temperature difference between the wall of a capillary and the molten sulfate is kept at more than 15°C. Because the capillaries were carefully set in a uniform temperature region of the furnace, convection should be negligible in the present work. If the mechanism of permeation of oxygen through pure molten Na4SO4 is the diffusion of oxygen molecules, the permeability of oxygen can be estimated from the product of the diffusivity [7] and solubility [8] of oxygen at 1 atm to be 1.5 × 10-11 mol cm -~ s -1 at 1000°C in the steady state. The estimated activation energy is 31 kcal tool -1. These estimated values are much smaller than the measured ones. The reason for this difference seems due to the different atmosphere used in the present work and by Rapp et al. [7]. The latter workers used gas mixtures of 02 and SO3 to prevent molten Na2SO4 from decomposition. In the present case, it is suggested that the molten Na2SO4 is decomposed on its surface at low partial pressure of oxygen according to the reaction of 1 Na2SO4~ 2Na + ~O2 + SO3 and that this metallic sodium dissolved in the sulfate diffuses to another surface at higher partial pressure of oxygen and is oxidized. In these circumstances the apparent permeation of oxygen would be due to the diffusion of the metallic sodium. The enthalpy of the decomposition of Na2SO4, e.g. for Na~SO4 ~ Na20 + SO3, is ~70 kcal per mole of sodium [9]. The activation energy of diffusion of sodium in the sulfate is estimated to be ~10 kcal mo1-1 from that of electric conductivity of molten NaESO4 [10, p. 289] and that of the diffusivity of sodium in molten Na~CO3 [10, p. 351]. Thus, the apparent activation energy

630

K. Nagata et al. / Permeability of oxygen through molten sodium sulfate

would be ~ 8 0 kcal mo1-1 a n d this value agrees a p p r o x i m a t e l y with the m e a s u r e d one. A c c o r d ing to this hypothesis, the increase of the perm e a t i o n of oxygen in m o l t e n Na2SO4 caused by addition of NaCI may be d u e to the increase of the c o n c e n t r a t i o n of metallic s o d i u m since molten NaC1 is a good solvent for metallic s o d i u m [111. O n the o t h e r hand, the effect of a d d i t i o n of metallic ions to m o l t e n Na2SO4 suggests a different m e c h a n i s m for the p e r m e a t i o n of oxygen from that in the pure melt. Sasabe a n d K i n o s h i t a [12] r e p o r t e d that the addition of 0.2% Fe203 to the m o l t e n slag of C a O - S i O 2 A1203 increased the p e r m e a b i l i t y of oxygen by a factor of 10 m, b e c a u s e oxygen was c o n s i d e r e d to p e r m e a t e as oxygen ion, a n d this was accelerated by positive hole t r a n s p o r t . H e n c e oxygen in m o l t e n sulfate c o n t a i n i n g t r a n s i e n t metals is b e l i e v e d to p e r m e a t e by the same m e c h a n i s m as in the m o l t e n slag with Fe203.

References [1] J. Stringer, Ann. Rev. Mat. Sci. (1977) 477. [2] E.L Simons, G.V. Browning and H.A. Liebhafskey, Corrosion NAGE 11 (1955) 505. [3] M. Kawakami, K.S. Goto, R.A. Rapp and F. Kajiyama, Tetsu to Hagane 65 (1979) S1 I. [4] J.A. Goebel and F.S. Pettit, Met. Trans. 1 (197(/) 1943. [5] R.A. Rapp and K.S. Goto, Proceedings of the Symposium on Fused Salt, Oct., 1978. [6] S. Nagasaki, Introduction to experimental techniques for thermal analysis (Kagakugijutsusha, 1979) p. 174. [7] R.A. Rapp, R.C. John and D.A. Sores, private communication, April, 1978. [8] R.A. Rapp and R.E. Andersen, Proceedings of the ECS Meeting, Seattle, May, 1978. [9] D.R. Stull and H. Prophet, eds., JANAF Thermochemical Tables, 2nd Ed., NSRDS (1971). [10] G.J. Janz, Molten salts handbook (Academic Press, New York, 1%7). [1 I] M.A. Bredig, in: Molten salt chemistry, ed. M. Blander (Interscience, New York, 1964) p. 375. [12] M. Sasabe and Y. Kinoshita, Tetsu to Hagane h5 (Iq79) 1727.