applied
surface science EISEWIER
Applied Surface Science
115(1997) 3.55-360
The CsCl- and CsNO,-induced high temperature oxidation of Nimonic-90 alloy at 1123 K M. Misbahul Amin Department
of Chemisty.
Facultyv
Received
of Science.
*
ofMuiduguri.
lJniwrsi@
17 June 1996; accepted 23 December
Maiduguri.
Borne-Stair.
Nigericr
1996
Abstract Nickel-base alloys find vital scope in space and nuclear applications due to their numerous favourable properties such as low density, high strength and corrosion resistance. The high temperature oxidation of Nimonic-90 (N-90) alloys has been studied in the absence or presence of CsCl and CsNO, salts at 1123 K for a period of 180 ks in atmospheric condition. The alloy is more severely attacked by CsCl than CsNO, due to formation of volatile chlorides. The tests included mass change monitoring, oxide scale analysis by X-ray diffraction, surface morphology examination by scanning electron microscopy (SEM) and detection of possible surface contaminants by energy dispersive X-ray analysis (EDAX). 0 1997 Elsevier Science B.V.
1. Introduction The corrosion resistance of nickel-base is due to a 20-40 A passive protective film which consists mainly of Cr,O,. It has been reported that the high temperature oxide film produced during operations has reduced corrosion resistance [ 11. The nickel-base alloys are widely used as high temperature oxidation resistant materials and a good amount of work has been carried out on Na,SO,-induced high temperature attack [2-61. Oxide spallation and cracking produced by CsCl exposure has been attributed to the formation of volatile metallic chlorides like NiCl, or CrCl I in locally reducing environments. Metallic chlorides transform to oxides, realeasing Cl, gas under oxidising atmosphere are reported [7-lo]. An extensive investigation of the alkali metal chlorides
* Tel.: + 234-76-232
and/or nitrates on hot corrosion of nickel-base alloys has been conducted [ 1 1- 161. The objective of this study is determining the high temperature oxidation rate and the mechanism of oxide formation in molten CsCl and CsNO, salts on N-90 alloy, oxidised at 1123 K for the period of 180 ks in atmospheric conditions. The investigations help to clarify the great importance of understanding the oxidation kinetics and the morphologies of the scales of N-90 alloy.
2. Experimental
150232655.
0169-4332/97/$17.00 0 1997 Elsevier Science B.V. All rights reserved. PII S0169-4332(97)00003-2
details
The present study has been carried out using commercially available N-90 alloy in plate having nominal composition (in wt%) Cr-2 1.OO; Co-2 1.OO; Fe- 1SO; Ti-3.00; C-O.13 and Ni-balance. The coupons of 2.0 X 1.5 X 0.5 cm were cut from plate
356
M. Misbahul Amin / Applied Sufuce
and were polished up to 800 grade Sic papers, lubricated with Vaseline oil, degreased with petroleum ether and methanol. After cleaning, the coupons were covered with a saturated solution of CsCl or CsNO,. Deposition of salt was carried out by spraying in a preheated condition [17]. In this way a well-controlled coating (about 5 pm thickness) and a uniform deposition of CsCl or CsNO, could be established. The coupon edges were coated to a lesser extent. Oxidation experiments were carried out in a vertical tubular furnace attached to a helical thermal balance, which has been described previously [ 181. High temperature oxidation studies of N-90 alloy were carried out in absence or presence of CsCl and CsNO, salts at 1123 K in atmospheric conditions. The mass change with time was recorded every 18 ks throughout the 180 ks run. After oxidation, the representative coupons were characterized by X-ray diffraction analysis. The various constituents were identified using a Phillips PW1730 X-ray diffractometer fitted Fe-, Co- or Cu-K,targets. Structural studies were performed using
Science I15 (I 997) 355-360
SEM-JEOL 840A. X-ray elemental line profiles analysis of various elements in the oxidised coupons were carried out using an EDAX 711B analyser.
3. Results The rates of oxidation of the N-90 alloy in the absence or presence of CsCl and CsNO,, oxidised at 1123 K for the period of 180 ks in atmospheric conditions are illustrated in Fig. 1. It is evident that the uncoated alloy shows a linear increase in oxidation rate with time. The CsCI-induced alloy exhibits mass gain upto 120 ks followed by a decrease in mass. Alloy coated with CsNO, appears to be mass losses upto 90 ks followed by slight increase in mass at longer time. It is observed that the maximum mass gain is seen in the case of uncoated alloy than CsCland CsNO,-induced alloy. Examination of surfaces of the uncoated coupons by SEM after oxidation revealed that they were relatively uniform (Fig. 2). The scales formed on the coupon were thinner, dense and fine-grained and
7
I 095 0
I 3
I 6
k Time
_
Fig. 1. The change in mass with time curves of N-90 alloy uncoated atmospheric conditions.
I
112 KS
X
15
-+
‘0
and coated with CsCl and CsNO,,
oxidised at
1123 K for 180 ks in
357
M. Misbahul Amin /Applied Surface Science 11.5 (1997) 355-360
Table 1 Constituents identified in the scales by X-ray diffraction analysis. oxidised at 1123 K for 180 ks duration in atmospheric conditions Salt coating Nimonic-90 Uncoated CSCI CsNO,
Fig. 2. SEM of N-90 alloy. oxidised at 1123 K for 180 ks (8 pm).
retained their integrity upto 180 ks of oxidation without any evidence of loss of oxide adhesion. The internal chromia scales are relatively compact and the outer layers are porous, containing cobalt oxide. In contrast, the scales formed on the CsCl-coated coupon are rough exhibiting a tendency to deform. wrinkle and undergo local detachment (Fig. 3). The inner layers of scales constitute chlorides of chromium and nickel incorporated with Cr,O, and NiO (Table 1). It is quite evident that during cooling of the molten fluxed mass, stresses develop resulting
Fig. 3. SEM of N-90 alloy coated with CsCl. oxidised (ID /.Lm).
at 1123 K
Constituents
identified
alloy
Cr,O,, NiO, Co,O, Cr,O,, Cr,O,.
CrCI,, NiO, Cs2NiOz, NiCl,. CsCl NiO. CszNiOz. COO. CS,C~O,
in the appearance of cracks and voids. This might be due to formation of intermediate species CrO,Cl,, some of which evaporate and condense on the wall of the reaction tube and some of which decompose and accumulate at the salt/alloy interface in the form of Cr,O,. The X-ray elemental line profiles analysis of the CsCl-induced alloy (Fig. 4) indicate the chromium, nickel and cobalt are enriched on the surface of oxides. No significant variation in chloride concentration is found. Due to formation of volatile chlo-
Fig. 4. Ni. Cr. Co and Cl X-ray profiles across a scale formed on C&-coated N-90 alloy. oxidised at 1123 K for 180 ks.
M. Misbahul Amin /Applied
358
Fig. 5. SEM of N-90 alloy coated with CsNO,, K for 180 ks (10 wrn).
oxidised at 1123
Surface Science 115 (1997) 355-360
rides such as NiCl,, CrCl,, some portions of the scales contain low concentration of chlorides as observed. Surface morphological examinations revealed that the scales formed on the CsNO,-induced alloy exhibited a porous copious scale (Fig. 5). The triplex scale contains Cr,O,, NiO and Co0 which are distributed. The scales contain thick layers of Cr,O, and NiO which are in predominant concentrations followed by a discontinuous layer of Co in the form of oxides. The existence of fluxing products and this has been further confirmed by the identification of products like Cs2Cr04 and Cs,NiO, in the scales by X-ray diffraction analysis. The Ni-, Cr- and Co-K, X-ray concentration profiles (Fig. 6) indicate the existence of triplex oxide scales with NiO in predominant concentrations in the outer layers and Cr,O, in the inner scales.
4. Discussion
--Cr _._.90 ,um sample
co
,I ,
, I I
Fig. 6. Ni, Cr. and Co X-ray profiles across a scale formed on CsNO,-coated N-90 alloy, oxidised at 1123 K for 180 ks.
Oxide scales formed on N-90 alloy in the presence of alkali metal halides and nitrates under investigation are heterogeneous and layered. The alkali metal salts have a deleterious effect on the protectivity of the scales and rapid decomposition of the alloy is noted [19,20]. The oxidation behaviour of N-90 alloy coated with CsCl and CsNO, and oxidised at 1123 K in atmospheric conditions has demonstrated that these coatings are effective in significantly reducing the extent of oxidation of the base alloy. This result is in agreement with other data [7-91 which show that additions of alkali metal halides or nitrates at a sufficient level can significantly reduce the extent of oxidation of high chromium-containing nickel-base alloys. The severe attack observed on the CsCl-coated alloy which has caused the formation of a volatile, CrO,Cl, by interaction of Cr,O, and CsCl and its subsequent conversions into Cs,CrO,. The oxide of nickel also reacts with CsCl to form NiCl,. It has been shown elsewhere [7] that the ease of formation of volatile chlorides seems to be the rate determining factor in the high temperature oxidation of the alloy.
M. Misbahul Amin /Applied
The reactions 2CsCl-t
Cr,O,
can be described
+ :O,
(1)
--) 6Cs,CrO, + 4CrC1,
NiO + NiCl, + Cs,O
(2)
(4)
The formation of volatile products such as CrO,CI? and Cs,CrO, and volatile halides, i.e. CrCl, and NiCl?, exert sufficient vapour pressure so as to break the passivation of oxides on the alloy. Once passive film breakdown occurs, the molten CsCl further attacks the alloy substrate forming metal halides. as described by the following reactions: Ni + 2CsCl + +O, + NiCI, + Cs,O
(5)
Ni + Cs,O + +O, + Cs,NiO,
(6)
2Cr + 6CsCI + i02 + 2CrC1, + 3Cs20
(7)
Cr + 2CsCl+
(8)
$0,
--f CrO,Cl,
+ Cs,O
Cr + Cs,O + *O, --j Cs,CrO,
(9)
It is concluded that the chlorides might be entrapped between the inner oxide layers of the alloy and get condensed on cooling and appear as distinct and discrete phase(s) in the scales. It is apparent that CsNO,, being an active oxidising agent, releases additional oxygen during fluxing reactions. The decomposition of CsNO, into CsNO, and then involvement of the latter in fluxing reactions results in the formation of molten fluxing products and evolution of nitrogenous gases. The consequences of fluxing and evolution of gases are the cracking and disruption of the scales. Availability of fused salt would come in contact with the fresh metal resulting in higher oxidation rates. The results obtained for the dissolution of passive oxides and alloy with the following mechanisms: 2CsN0,
+ 2CsN0,
4CsN02 + Cr,O,
+ 0,
+ :O,
(‘0) -+ 2Cs2Cr0, + 4N02
2CsN02 + NiO + +02 + Cs,NiO, 2CsN0,
(3)
NiO + Cs,O + Cs,NiO,
-+ 2Cs,CrO, + 2N2 + ;02
+ $0, -+ CrO,CI,
I2CsCI + SCr,O,
2CsCl+
4CsN02 + Cr,O,
as follows:
+ Cs,CrO,
359
Surface Science 115 (1997) 355-360
(1’)
(12) + 2N02
+ NiO + Cs,NiO, _ + N,_ + $0,
(13) (14)
Although the morphologies of the scales in both cases are similar. The scale cracks and voids initiated in the CsCI- and CsNO,-coated alloys appeared to be similar but the mode of propagation is different.
5. Conclusions Based on a comparison and analysis of the structure and properties of oxide scales of N-90 alloy in CsCl and CsNO, coatings, it is clear that the passive films can simultaneously deteriorate the alloy surface. During the reaction of molten CsCl with oxides scales, chlorides and fluxing products are formed and some volatile products like CrO,CI,, Cs,CrO, and chlorides such as NiCl, and CrCI, exert sufficient vapour pressure so as to break the protective scales. The CsCl-induced alloy involved fluxinggchloridination reaction. CsNO,-coated alloy appeared to have suffered mass losses initially due to evolution of nitrogenous gases and formation of fluxing products. The easy availability of oxygen indicates the higher oxidation rate and produces a positive effect on the fluxing rates of the alloy.
References [I] 0. Bianchi. A. Cequetti. F. Muzza and S. Torchio. Corros. Sci. 12 (1972) 495. [3] P.L. Daniel and R.A. Rapp, Adv. Corros. Sci. Technol. 5 ( 1976) 57. [3] C.G. MeCreath, Trans. Inst. Mar. Eng. 88 (1976) 145. [4] P.L. Grouse and C.M. Stander. J. Phys. Chem. Solids 49 (1988) 1149. [s] M. Skeldon. J.M. Calvert and D.G. Lees, Oxid. Met. 28 (1987) 109. [6] J.G. Smeggil and H.S. Bronstein, NASA-CR-l 35 348 NASA Scientific and Technical Information Branch, Washington. DC 20546 (1977) p. 166. [7] M. Misbahul Amin. Prakt. Met. 30 (1993) 239-247.
360
M. Misbahul Amin /Applied
[8] P. Hancock, in: Proc. Conf. High Temperature Metal Halide Chemistry, Eds. D.L. Hilderbrand and D.D. Cubicciotti (The Electrochem. Sot., Princeton, NJ, 1978) p. 646. [9] R.L Jones, Metal-Slag-Gas Reaction and Processes Symposium, Eds. Z.A. Foroulis and W.W. Smeltzer (The Electrochem. Sot., 1975) p. 764. [IO] O.F. Devereux, K.Y. Kim and K.S. Yeum, Corros. Sci. 23 (1983) 210. [l 11P. Reiger, Electrochemistry (Prentice-Hall, Englewood Cliffs, NJ, 1987) p. 292. [12] G. Picard, Proc. 5th Int. Conf. on Molten Salts, Las Vegas, Oct. 13-18, 1985, p. 426. [13] B. Tremillon and G. Picard, in: Proc. 1st Int. Symp. on Molten Salts, Kyoto, 1983, Ed. Y. It, p. 114. [14] F.S. Pettit and C.S Giggins, in: Proc. Hot Corrosion in Supperalloys II, Eds. C.T. Sims, N.S. Stoloff and W.C. Hagel (Wiley, NY, 1987) p. 327.
Surface Science 115 (1997) 355-360 [l5] S.R.J. Saunders and J.R. Nicholls, Thin Solid Films 119 (1984) 114. [16] Thermal Convection Loop Corrosion Tests of 31655 and IN-800 in Molten Nitrate Salts, Sandia Report 81-8210 ( 1982). [l7] A.U. Malik, M. Misbahul Amin and S. Ahmad, Trans. Jpn. Inst. Met. 25 (1984) 169. [I 81 M.J. Oraham and M. Cohen, J. Electrochem. Sot. 119 (I 972) 879. [19] F.J. KohI, C.A. Stearns and G.C. Fryburg, in: Proc. 4th US/UK Navy Conf. on Gas Turbine Materials in Marine Environment, Vol. II (US Naval Sea Command, Annapolis, MD. 1979) p. 565. [20] C.S. Giggins and F.S. Petit, Pratt and Whitney Aircraft Group Scientific Report, 1 June-30 Sept (1978) P.W.A. Rep. No. FR 11545.